(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 "Race, ethnicity, and participation in the arts : patterns of participation by Hispanics, Whites, and African-Americans in selected activities from the 1982 and 1985 surveys of public participation in the arts"

Race, Ethnicity 
" and Participation 



in the Arts 



Patterns of Participation by Hispanics, Whites, and 
African-Americans in Selected Activities from the 

1 982 and 1 985 Surveys of Public 
Participation in the Arts 



Paul DiMaggio 
Francie Ostrower 



NATIONAL 
ENDOWMENT 



FOR S0 



THE 



Research Division Report #25 ARTS 



Race, Ethnicity, and 
Participation in tiie Arts 



Race, Ethnicity, and 
Participation in the Arts 

Patterns of Participation by Hispanics, 

Whites, and African-Americans in 

Selected Activities from tine 1982 

and 1985 Surveys of Public 

Participation in the Arts 



Paul DiMaggio Francie Ostrower 

Princeton University Harvard University 



Research Division Report # 25 
National Endowment for the Arts 



Seven Locks Press 
Washington, D.C. 



RacCy Ethnicity, and Participation in the Arts is Report #25 in a series on 
matters of interest to the arts community commissioned by the Research 
Division of the National Endowment for the Arts. 



Library of Congress Cataloging-in-Publication Data 

DiMaggio, Paul. 

Race, ethnicity, and participation in the arts : patterns of 
participation by Hispanics, Whites, and African- Americans in 
selected activities from the 1982 and 1985 surveys of public 
participation in the arts / Paul DiMaggio, Francie Ostrower. 

p. cm. — (Research Division report / National Endowment for 
the Arts ; #25) 

ISBN 0-929765-03-6 

1. Arts surveys — United States. 2. Arts audiences — United States. 
3. Ethnic arts — United States. 4. Minorities in art. I. Ostrower, 
Francie. II. Title. III. Series: Research Division report 
(National Endowment for the Arts. Research Division) ; 25. 
NX220.D56 1992 

700' .973* 09048— dc20 92-3 1145 

CIP 



Manufactured in the United States of America 

Seven Locks Press 
Washington, D.C. 



Contents 

EXECUTIVE SUMMARY vii 

Differences in Participation viii 

Demand for More Participation xi 

Additional Findings from November/December 1982 xiii 

1 : RACE, ETHNICITY, AND PARTICIPATION IN THE ARTS 1 

Defining Our Terms 2 

2: RATES OF PARTICIPATION BY RACE AND ETHNICITY 10 

The Core Activities 10 
Differences in Core Participation 

Among Hispanic Ethnic Groups 16 

Participation in Other Artistic Activities 16 

Use of the Media for Exposure to the Arts 21 

Musical Preferences 27 

Childhood Experience in the Arts 30 

Summary 36 

3: NET EFFECTS OF RACE AND ETHNICITY ON 

PARTICIPATION IN SPPA CORE ACTIVITIES 39 

Explaining Racial and Ethnic Differences 40 

Racial/Ethnic Effects Net of Sociodemographic Differences 44 

Differences in Predictors of Participation by Race/Ethnicity 49 

4: RACIAL/ETHNIC DIFFERENCES IN UNSATED 

DEMAND FOR PARTICIPATION 56 

Demand for Greater Participation 56 

Why People Don't Attend More 64 

Conclusions 68 

5: EVIDENCE ON RACIAL AND ETHNIC DIFFERENCES 
IN PARTICIPATION FROM THE NOVEMBER/DECEMBER 

1982 SUBS AMPLE 71 

Measures 72 

Intergroup Differences in Socialization, Taste, 77 

and Participation Scales 

Race, Ethnicity, and Parental Guidance 77 

Race, Ethnicity, Musical Taste, and Television Arts Viewing 80 



Contents 

Race, Ethnicity, and Artistic Participation 80 

Summary of Findings Thus Far 84 
Differences in Models Predicting Artistic Socialization, 

Taste, and Participation by Race 86 
Do Intergroup Differences Vary by Gender, Educational 

Attainment, or Age? 92 

Summary 98 

6: CONCLUSIONS 101 

Do Rates of Participation Vary? 102 

Does Participation Vary Net of Sociodemographic Factors? 103 

Does Demand for Artistic Participation Vary? 106 

Summary Conclusions 109 

Further Research 110 

NOTES 113 

APPENDDC TABLES 1 23 

About the Authors 201 



VI 



EXECUTIVE SUMMARY 



This report uses data from the 1982 and 1985 Surveys of Public 
Participation in the Arts (SPPA) to describe and explain differences 
in patterns of participation in selected artistic activities by African- 
American, Hispanic, and non-Hispanic white respondents. The surveys 
permit generalization to national populations of whites and African- 
Americans because the SPPAs were designed to be nationally representative 
of the U.S. population with respect to age, gender, and race. Because the 
sample was not designed to be representative with respect to Hispanic origin 
or other ethnic categories, conclusions about the participation of Hispanic- 
Americans must be more tentative. Asian-Americans were identifiable only 
in the 1985 data, but too few were included in the sample to permit 
generalization about this group. Native Americans were not identified sepa- 
rately, thus making analysis of their participation impossible. 

Data on socioeconomic and demographic background and on participa- 
tion in ten "core" activities were collected from all respondents in both years. 
The core activities were: attending jazz, classical music, opera, musical 
theater, theater, and ballet performances; visiting art museums or exhibits; 
reading works of imaginative literature; playing a musical instrument in 
public; and dancing or singing or acting on stage. The SPPAs also asked 
subsamples of respondents each year about: (1) participation in "other" 
activities, including visiting historical or science museums or monuments, 
reading poetry, taking art lessons, painting or drawing, engaging in various 
craft activities, and working backstage in the performing arts; (2) consump- 
tion of arts programming on television, radio, or sound recording; (3) desire 
for additional participation in the core activities and reasons for not partici- 
pating more; (4) socialization into the arts as children and specific kinds of 
art lessons taken throughout the respondent's life; (5) and attitudes towards 
1 3 kinds of music. 

Descriptive statistics on the core questions were derived from analyses 
of the full samples for both years; descriptive statistics on the other questions 
were derived from analyses of subsamples of whom these questions were 



VII 



Race, Ethnicity, and Participation in the Arts 

asked for both years; and multivariate analyses employing data from two or 
more of the intermittently asked questions are based on data from November 
and December 1982, the only months during which the same respondents 
were asked all of the questions. 



Differences in Participation 

SPPA Core Activities. With the exception of attendance at jazz concerts, for 
which African-American rates of participation exceeded those of whites or 
Hispanics, white respondents participated more in all of the core activities 
than did African-American or Hispanic respondents. Most absolute differ- 
ences between groups with respect to core activities were relatively small, 
with spreads ranging from one-tenth of 1 percent (Hispanic ballet attendance 
in 1982) to, at most, almost 24 percent (Hispanic fiction reading in 1982) 
between minority groups and the non-Hispanic white majority. Most abso- 
lute percentage differences were low in large part because, except for reading 
literature, relatively few members of any group participated in core activities. 

If one looks not at gross differences between the percentages of groups 
participating, but rather at the ratio of percentages participating for different 
groups (odds ratios), for some activities the differences in rates of participa- 
tion for whites, on the one hand, and African-Americans and Hispanics, on 
the other, were sizable. For example, in both years whites were more than 
twice as likely as African-Americans to report attending a classical music 
concert, an opera performance, a musical theater performance, a play, or a 
ballet. Non-Hispanic whites were also more than twice as likely as Ameri- 
cans of Hispanic origin to report attending a play (in both years) and in 1985 
attending a classical music concert or an opera performance.^ 

Rates of public performance on musical instruments or by singing, 
dancing, or acting were lower than those for attendance at arts events for 
members of all groups. Differences between whites and other groups were 
smaller for these art-producing activities than for most core art-consuming 
activities. 

Other activities. Whites were substantially more likely to visit museums or 
exhibits than Hispanics, who were somewhat more likely to do so than 
African- Americans. Differences between African- Americans' and whites' 
rates were substantial for visiting history or science museums, historical 
monuments, and arts or craft fairs. White respondents were also substantially 
more likely than others to engage in needlework crafts, and much more likely 



VIII 



Summary 

than African- Americans to participate in other crafts activities. By contrast, 
whites were only somewhat more likely than others to have read or listened 
to poetry, taken art lessons, or engaged in printing and drawing, photography, 
or filmmaking. Some of these participation rates for African-Americans and 
Hispanics fluctuated between 1982 and 1985. (For example, Hispanics in 
1982 were more likely to report creative writing than whites, whereas they 
were less likely than African-Americans or whites to indicate participating 
in this activity in 1985.) 

Evidence from the core and other activities indicates that minority group 
members were less likely to attend cultural institutions, relative to whites, 
than to be found among amateur creative artists. Nonetheless, the tendency 
of white Americans to participate at higher rates than others manifested itself 
in responses to most of these questions. The exception of jazz, for which 
African-American attendance rates were well above those of whites or 
Hispanics, indicates that these differences are genre specific, and that inter- 
group patterns of difference should not be generalized beyond the activities 
about which the SPPA asked. 

Use of the Media for Arts Consumption. More people encountered the arts 
covered in the SPPA through the media than in live settings. The proportion- 
ate gap between white and minority attendance was smaller in consumption 
of the arts through the media than in live attendance. In other words, although 
members of all groups were more likely to watch the arts covered by the core 
questions than to attend them, this tendency was more pronounced in the case 
of minority group members than in the case of whites. 

People who watched an arts program on television were more likely than 
others to attend comparable live events. Viewing and attending tended to be 
more closely associated for African-Americans and Hispanics than for 
non-Hispanic whites. Smaller intergroup differences for viewers than for 
non viewers were evident in both 1982 and 1985 for Hispanic respondents 
with respect to classical music, musical theater, ballet, and art, and for 
African-American respondents with respect to opera and musical theater. 

Musical Preference. Respondents were asked if they enjoyed listening to 
each of 13 musical genres: classical, opera, jazz, show tunes, big band, 
soul/rhythm and blues, rock, country western, easy listening, folk, bluegrass, 
hymns/gospels, and barbershop. Their responses indicated significant differ- 
ences associated with race or ethnicity within the context of a national 
musical culture dominated by commercially produced genres. African- 
Americans were particularly likely to report enjoying such forms as jazz. 



IX 



Race, Ethnicity, and Participation in the Arts 

blues, and gospel, which have deep roots in the African- American experi- 
ence, whereas white and Hispanic respondents were more likely to choose 
country western, easy listening, and rock. But even commercial genres Uke 
rhythm and blues or country westem, which are associated historically with 
specific racial or ethnic groups, appear to have become part of a national 
musical culture. Thus, approximately one in four whites liked jazz and blues, 
and an equal proportion of African-Americans enjoyed country westem. 
Preferences were neither sharply segmented by race nor indicative of an 
undifferentiated mass culture in which racial and ethnic differences have 
atrophied. 

Few respondents in any group reported enjoying opera, although sub- 
stantial minorities liked classical music, jazz, and show tunes. Although 
whites were considerably more likely than African-Americans to report 
enjoying classical music, Hispanics were almost as favorable in 1982 and 
more likely to report enjoying classical music than whites in 1985. 

Parental guidance. White respondents were considerably more likely than 
either African- American or Hispanic respondents to report that their parents 
took them to art museums or listened to classical music when they were 
children. Whites were only somewhat more likely than African- Americans, 
who were more likely than Hispanics, to report that their parents took them 
to plays, dance performances, or classical music concerts, or that parents 
encouraged them to read when they were young. 

Lessons and classes. African-Americans and whites were almost equally 
likely to report having taken many kinds of classes in the arts during the high 
school years, whereas whites were more likely to report taking classes before 
and/or after high school. By contrast, Hispanic-Americans were less likely 
than whites or African -Americans to report taking many kinds of arts classes 
when they were young. Differences were particularly marked with respect 
to music lessons and music appreciation courses.^ 

Net Differences Between African-Americans, Hispanics, and Whites. To 
what extent were differences in participation rates in the core activities the 
result of differences in the socioeconomic standing and demographic char- 
acteristics of African- Americans, Hispanics, and non-Hispanic whites? Lo- 
gistic regression analyses were used to predict participation in core activities, 
with attention to the effects of group membership (African-American and 
Hispanic as compared to white), controlling for age, gender, educational 
attainment, occupation, family income, marital status, and SMSA residence.-^ 



Summary 

Even with these controls for sociodemographic factors, whites were signifi- 
cantly more likely than African-Americans to participate in most of the core 
consumption activities, but not in attending jazz concerts (for which Afri- 
can-Americans were significantly more likely to participate). Whites were 
also significantly more likely to participate by performing on a musical 
instrument or as actors, singers, or dancers. 

With respect to the consumption activities, a substantial portion (but, 
with one exception, less than half) of the gross difference (i.e., without 
controls) in participation rates between African-American and white Ameri- 
cans stemmed from sociodemographic, especially socioeconomic, differ- 
ences between the races. When one looks not at probabilities of participation 
in specific activities but at a measure of the range of performing arts 
attendance (excluding jazz) in which respondents participated, more of the 
gross difference between African-Americans and whites is explained by 
sociodemographic factors. 

Although these interracial differences survived the inclusion of numer- 
ous controls, they are small relative to differences associated with other 
determinants of participation. With respect to all of the activities for which 
being African-American was associated with significantly lower participa- 
tion (relative to whites), the direct effect of race is dwarfed by the impact of 
educational attainment and (except for reading in 1982) exceeded by the 
effect of family income. Similarly, once other sociodemographic factors are 
taken into account, participation rates of African-Americans and whites are 
more similar than are rates for men and women for all such activities except 
visiting art exhibitions. 

Although gross rates of participation in the core activities were similar 
for Hispanic and African- American respondents, larger proportions of the 
differences between Hispanics and whites than between African-Americans 
and whites stemmed from intergroup differences in sociodemographic at- 
tributes. Controlling for socioeconomic and demographic factors left signifi- 
cant differences between whites and Hispanics in both years only for reading 
and attendance at musical and dramatic theatrical performances — the only 
ones of the 10 core activities for which command of the English language is 
often essential. 



Demand for More Participation 

Some respondents in each year were shown a card listing the core arts 
attendance activities and told: 



Xi 



Race, Ethnicity, and Participation in the Arts 

"Few people can do everything they would like to do. But if you could 
do any of the things listed on this card as often as you wanted, which ones 
would you do more often than you have during the last 12 months?" 

Those respondents who said they would like to have attended a given kind 
of performance or exhibition more than they had in the past year were then 
asked to indicate what prevented them from participating more. 

The percentage of respondents in each group who had not participated 
in each activity but who reported that they wanted to do so was added to the 
percentage who reported having participated to estimate a potential partici- 
pation rate. These potential participation rates were much greater than actual 
participation rates for all groups. Except for white attendance (in 1982 and 
1985) and Hispanic attendance (in 1985) at classical music concerts, and 
white and Hispanic visits to art museums and galleries (in both years), 
potential rates were at least twice the actual rates of attendance and, in many 
cases, were much greater. 

Demand for participation in the seven core consumption activities ap- 
peared to be cultivated by attendance. People who already attended were 
much more likely to want to attend more than were people who had not. 
Thus, although there was much apparent unsated demand for these activities, 
most of it came from attenders rather than nonattenders. Because, with the 
exception of jazz performances, whites were more likely to attend than 
African-Americans or Hispanics, unsated demand appeared to be greater 
among whites. Moreover, nonattenders from groups that had the highest 
attendance rates (African-Americans for jazz, whites for everything else) 
were more likely than nonattenders from other groups to want to attend. 
Consequently, if everyone had done what they said they wanted to do, the 
absolute margins in participation rates between whites and everyone else 
would have been wider. (For the exceptional activity, jazz, the gap between 
African-Americans and others would have widened.) For most activities, 
however, the ratios of white to both African-American and Hispanic rates 
would have declined. 

This could be interpreted as meaning that eliminating barriers to atten- 
dance would exacerbate intergroup differences in participation in the SPPA 
core activities (if one focuses on gross differences) or at best moderate only 
some differences, and these only slightly (if one focuses on ratios). This 
conclusion is questionable, however, on at least two grounds. First, the most 
important barriers to participation may be those that influence demand, not 
those that influence the ability of persons to satisfy demand they already 
have. Second, respondents to the SPPA "want-more" questions may have 



XII 



Summary 

responded on the basis of taken-for-granted understandings about the costs 
associated with getting more of what they wanted, thus artificially suppress- 
ing demand among groups facing greater barriers. 

For members of all groups, cost and lack of time were the most important 
reasons given for nonparticipation. With respect to most activities, white 
respondents were more likely to give time as a reason than cost, and Hispanic 
respondents were more likely to cite cost than time. In 1982, African- Ameri- 
can respondents were somewhat more likely to mention cost than time for 
most activities, whereas in 1985 they were somewhat more likely to cite time 
than cost. Lack of availabihty was frequently cited by whites and lack of 
accessibility was often mentioned by Hispanics. African- American respon- 
dents frequently mentioned these and also cited inadequate transportation as 
an impediment to attendance more than whites and, for most activities, more 
than Hispanics. For most activities, Hispanics were more likely than Afri- 
can-Americans or whites to cite child care problems as reasons for not 
attending. Fear of crime, being handicapped or having health problems, poor 
quality of available activities, inadequate publicity, work-related reasons, or 
inconvenient times of performance were rarely cited as reasons for not 
participating by respondents in any group. 



Additional Findings from November/December 1982 

Because all respondents to whom the SPPA was administered in Novem- 
ber and December 1982 were asked all the questions, this subsample is 
useful for investigating a broader range of questions than could be addressed 
using data from the total 1982 or 1985 samples. 

Net differences in parental guidance/ lessons. Two scales were created: one 
to count the number of kinds of parental arts socialization each respondent 
reported receiving as a child, and one to count the number of kinds of arts 
lessons or classes taken by age 17. (Because the survey failed to distinguish 
between lessons at home and those at school, these are treated separately 
from parental guidance.) Although African-American and Hispanic respon- 
dents received fewer parental-guidance experiences and took fewer arts-re- 
lated classes or lessons in their youth than whites, these differences were 
entirely a result of the fact that African-American and Hispanic respondents 
had parents who had received fewer years of formal education than did the 
parents of white respondents. Controlling for parental education, African- 
American and Hispanic parents gave their children more kinds of home 



XIII 



Race, Ethnicity, and Participation in tlie Arts 

guidance experiences than did comparable white parents, and no differences 
remained in the number of kinds of lessons. 

Taste for art music and related genres. Factor analysis isolated a cluster of 
musical genres (classical and chamber music, opera, show tunes, big band, 
and easy listening music) which were summed into an additive scale. White 
respondents scored significantly higher than African-Americans and Hispan- 
ics on this scale. Controls for sociodemographic factors reduced the sizable 
African-American/white difference by almost half, but a modest significant 
difference remained. Sociodemographic controls eUminated all of the differ- 
ences between Hispanics and whites. 

Viewing arts programs on television. A scale was created as a simple count 
of the number of kinds of arts programs that each respondent reported having 
watched on television. White respondents reported viewing slightly but 
significantly more kinds of televised arts programs than African-Americans 
or Hispanics, but these small differences were entirely the result of sociode- 
mographic differences between whites and the other groups. 

Participation scales. Factor analysis of combined responses to the SPPA's 
core and other participation questions generated five scales consisting of 
participation items reflecting, respectively: performing-arts attendance (in- 
cluding and excluding jazz); visual and literary consumption activities; 
performing-arts production activities; and visual and literary production 
activities. Regression analysis was used to examine the effects of race and 
ethnicity on these scales, controlling for sociodemographic characteristics, 
parental guidance and lessons, and artistic taste and interest as reflected by 
the art-music and tele vision- vie wing scales. The results added further evi- 
dence that one cannot generalize about the effects of race or ethnicity on 
cultural participation per se. Hispanic-Americans attend fewer public arts 
consumption activities than whites (both performing and visually oriented), 
but this difference was almost entirely the result of the fact that white 
Americans had more years of education, higher incomes, and higher status 
occupations. Hispanic respondents participated in no fewer art-producing 
activities (either performing or visual) than white respondents and, with both 
sociodemographic factors and guidance/lessons controlled, they participated 
in these art-producing activities significantly more than did comparable 
whites. 

There is no statistically significant difference between African- Ameri- 



XIV 



Summary 

can and white respondents with respect to participating onstage or backstage 
in performing-arts events, but African-Americans scored significantly lower 
than whites on the other scales. Sociodemographic differences, however, 
accounted for approximately 80 percent of the significant difference be- 
tween African-Americans and white Americans in the number of kinds of 
performing-arts events attended with jazz excluded, and all of the difference 
with jazz included. The remaining gaps were not statistically significant. 

Controlling for sociodemographic differences eliminated approximately 
40 percent of the differences between white and African-American respon- 
dents in the visual/literary consumption and production scales, although 
small, but statistically significant differences remained. The remaining sig- 
nificant difference in production was attributable to differences between 
African-Americans and whites in childhood socialization in the arts (both at 
home and through lessons and classes); whereas the differences in consump- 
tion remained significant even after including the full range of controls. 

Separate predictive models for African- Americans , Hispanics, and whites. 
Data on each group were separated in order to see if the factors predicting 
outcome measures were similar or different for the three groups. For the most 
part, artistic socialization, taste, and participation measures were predicted 
by the same variables for African-Americans and Hispanics as for whites. 
Two exceptions were notable, however. 

First, the effects of age on parental socialization, taste for art music and 
related genres, and arts television watching were greater for whites than for 
African- Americans. With parental education controlled, white parents of 
young respondents offered fewer arts socialization experiences than compa- 
rably educated white parents of older respondents, whereas African- Ameri- 
can parents of younger respondents offered more than comparable 
African-American parents of older respondents, suggesting that a conver- 
gence is occurring. Similarly, controlling for other sociodemographic fac- 
tors, tastes for art music and viewing arts programs on television increased 
with age for whites, but not for African- Americans and Hispanics. (These 
differences were significant except for white/Hispanic viewing of arts pro- 
grams on television.) Although these results might mean that white Ameri- 
cans' tastes change more with age than those of African-Americans or 
Hispanics, they may also indicate a convergence of all groups over time with 
respect to tastes for art music, and convergence between African-Americans 
and white Americans in watching arts programs on television. These find- 
ings are consistent with inspection of means by race and age: intergroup 



XV 



Race, Ethnicity, and Participation in the Arts 

differences in socialization and lessons, taste for art music, and arts televi- 
sion watching were smaller among younger respondents than for older 
respondents. 

Second, education had a significantly stronger effect on viewing arts on 
television and on all of the participation scales except for performing-arts 
production activities for whites than for African-Americans, although in 
most cases education was a significantly positive predictor for both groups. 
Moreover, the effects of taking lessons or classes in the arts were less 
pronounced for African-Americans than for other groups, although these 
differences were not statistically significant. Watching the arts on television 
was also less strongly predictive of attendance for African- Americans, and 
the differences between African-Americans and Hispanics were significant 
with respect to visual-arts consumption and production activities. In other 
words, the analyses provided tentative evidence that formal education, both 
general and arts-specific, was less strongly related to interest and participa- 
tion in the arts for African-Americans than for other groups. 

Change over time. For most participation activities, gaps between white 
and minority populations were greater for older than for younger respon- 
dents. Most of the decline in intergroup differences appeared to be the result 
of changes in the sociodemographic profiles of African- American, Hispanic, 
and white Americans. Sharp increases in the educational attainment of 
African-Americans and Hispanics have narrowed the gap in participation in 
the arts by younger men and women. 



XVI 



Chapter 1 

RACE, ETHNICITY, AND 
PARTICIPATION IN THE ARTS 



Since the creation of the first modem museums and orchestras after the 
Civil War, many Americans have regarded the arts as a public good. 
Even during the Great Depression, a number of institutions, including 
the Carnegie Corporation of New York and the Works Progress Administra- 
tion of the federal government, supported the extension of the visual arts and 
"good" music to communities that had little access to them.' After World 
War II, especially from the 1960s on, attention turned specifically to making 
the arts available to groups believed to have been culturally isolated. 

With the emergence of government and the large foundations as patrons 
of the arts, attention to minority participation became widespread. The shift 
of concern from the amount of artistic activity to the distribution of oppor- 
tunity to participate in such activity stemmed from at least three separate 
factors. 

First, due largely to the civil rights movement, the 1960s witnessed 
increased attention to the problems of the least well-off Americans and to the 
equitable distribution of such public goods as educational opportunity. 
Second, the traditionally dominant role of individual patrons in financing the 
arts was complemented by support from large institutions, especially private 
foundations and federal and state government agencies, which were com- 
pelled by their charter purposes to take a broad view of the public good. 
Third, with the expansion in the number and activity of arts organizations 
nationally in the 1970s, inequality of access to the arts came to be perceived 
less as a matter of regional disparity (at least among metropolitan areas) than 
of differences among groups within regions. 

It is the purpose of this report to examine the participation of racial and 
ethnic minorities in certain arts activities, primarily as audience members 
and to a lesser degree as amateur producers of art. With the completion of 
the 1 982 and 1985 Surveys of PubUc Participation in the Arts (SPPAs), more 
reliable data became available. Previous studies of attendance of racial and 
ethnic minorities in the arts were either narrow in scope or of dubious 



Race, Ethnicity, and Participation in the Arts 

technical quality. The SPPAs were undertaken by the U.S. Bureau of the 
Census as part of the National Crime Survey at the request of the National 
Endowment for the Arts. Responses from 1 7,254 persons in 1 982 and 1 3,675 
in 1985 were weighted (by age, gender, and race) to be representative of all 
noninstitutionalized Americans 1 8 years of age or older. The advantages of 
the SPPA data over data from earlier surveys include its national scope and 
representativeness, careful question design and pretesting, closely super- 
vised survey administration (usually in person rather than over the tele- 
phone), the broad scope of the questions asked, and the large number of 
respondents. Consequently, the SPPAs permit researchers and policy makers 
to pose more interesting questions and to generalize more confidently than 
we have been able to do in the past.^ > 

The SPPA included eight kinds of questions about cultural participation. 
The first set of questions {core activities) asked respondents to report on 
whether or not they had engaged in each of 10 kinds of activities during the 
previous year and, if so, how often they had done so during the previous 
month. ^ The second set of questions {barriers) asked respondents which of 
the core activities they would like to participate in more than they do now, 
and what factors prevent them from doing so. The third set of questions 
{socialization) asked respondents about the extent to which their parents 
encouraged certain kinds of participation in the arts and whether (and if so, 
when) they had taken several kinds of classes or lessons in the arts. The 
fourth set (not analyzed in this report) asked respondents about their partici- 
pation in a range of non-arts activities. The fifth set {location) asked those 
who responded affirmatively to one or more of the core questions where their 
participation had taken place. The sixth set of questions {musical preference) 
asked respondents whether or not they liked each of several genres of music, 
and which they liked best. The seventh {other participation) asked whether 
or not respondents had participated in several cultural activities that were not 
included among the core questions. The last set {media) asked respondents 
whether they had watched or listened to several kinds of arts presentations 
on television, radio, records, or tapes. All respondents in both years were 
asked the core-activities questions, whereas only a portion (approximately 
one third in 1982 and one sixth in 1985) were asked the others.'^ 

Defining Our Terms 

The task of this report is less straightforward than it may appear. To 
explore the extent to which members of racial and ethnic minority groups 
are underrepresented as participants in the arts, it is necessary to define such 



Race, Ethnicity, and Participation in the Arts 

terms as "racial and ethnic minority," "underrepresentation," and "participa- 
tion in the arts." 

Different definitions of the terms entail different definitions of the 
problem and, in some cases, different implicit values as to what is desirable. 
Different definitions may also yield different conclusions. In the sections 
that follow, we explain how and why we define our terms as we do, and 
speculate about the possible consequences of our choices. These explana- 
tions provide warnings that may help the reader interpret our results. 

Racial or ethnic minorities. The categories "race" and "ethnicity" are socially 
constructed, not scientifically given. The ways and the extent to which 
differences associated with racial or national origin are perceived as impor- 
tant bases for social cohesion, exclusion, and individual identity vary con- 
siderably among societies and across historical eras.^ In the United States, 
race is treated as a social fact, and most respondents to surveys have little 
trouble designaUng themselves as African-American (or Black), white, Asian 
or Pacific Islander, or American Indian (or Native American).^ These dis- 
tinctions are reflected in relatively low rates of intermarriage among mem- 
bers of different racial groups, so defined. Because the SPPA sample was 
designed to be nationally representative as to race, we began with the racial 
categories available to us in that survey: in 1982, White, African-American, 
and Other; and, in 1985, White, African -American, Asian, and Other.^ 

With respect to ethnicity, the situation is more complicated. The SPPA 
data provide a single ethnic code for each respondent, but many Americans 
are of mixed national origin. When asked to designate their ethnicity, they 
may have difficulty doing so, and if they are compelled to do so, their 
responses may be only partially accurate. Moreover, the SPPAs were not 
designed to be representative with respect to ethnicity, as they were with 
respect to race. Therefore, we do not know whether patterns of participation 
in the arts found among respondents to the SPPA surveys are typical of (or 
generalizable to) members of their ethnic groups. These considerations, and 
a comparison of SPPA responses with those on the 1980 decennial census, 
which allowed for multiple national origins, led us to conclude that the 
ethnicity data in the SPPA were generally not suitable for our analyses. 

At the same time, however, we concluded that the SPPA data would be 
useful for examining arts participation among Hispanic-Americans, includ- 
ing those respondents whose ethnicity was coded as Mexican, Puerto Rican, 
Cuban, Central or South American, or Other Spanish. The proportions of 
respondents who reported their ethnic origin as Mexican, Puerto Rican, or 
Cuban are comparable to those reported in the United States census, and few 



Race, Ethnicity, and Participation in the Arts 

Table 1-1 

Comparison of National Origin Estimates 

from 1 980 Census and 1 982 and 1 985 SPPAs 

Origin SPPA82 SPPA85 Census Single ii^ultiple 



German 


8.9 


8.0 


26.1 


36.5 


63.5 


Italian 


3.7 


3.8 


6.5 


56.5 


43.5 


Irish 


5.0 


4.6 


21.3 


25.7 


74.3 


French 


1.6 


1.9 


6.9 


23.8 


76.2 


Polish 


1.9 


1.8 


4.5 


46.3 


53.7 


Russian 


1.0 


0.8 


1.5 


49.6 


50.4 


English 


5.5 


5.4 


26.3 


47.9 


52.1 


Scottish 


0.9 


0.9 


5.3 


11.7 


88.3 


Welsh 


0.3 


0.2 


0.9 


18.5 


81.5 


Mexican 


3.5 


4.4 


4.1 


90.9 


9.1 


Puerto Rican 


0.7 


0.6 


0.8 


88.0 


12.0 


Cuban 


0.4 


0.2 


0.3 


83.7 


16.3 


Central/South 












American 


0.5 


0.8 


NA 


NA 


NA 


Other Spanish 


0.6 


0.6 


NA 


NA 


NA 


Afro-American 


10.3 


10.3 


11.1 


97.9 


2.1 


Other 


55.3 


55.7 


NA 


NA 


NA 



Note: Rightmost two columns report percentage of respondents to 1 980 census in each national origin 
group who reported single and multiple national origins, respectively. Only those national origins coded 
in SPPA are included. Because respondents to the 1980 census could give multiple responses, the 
census columns sum to more than 100 percent. All percentages from SPPA are weighted by race, 
age, and gender, and missing data (1 .95 percent for 1982, 2.55 percent for 1985) are omitted from 
base. Source for census data is Bureau of the Census, 1980 Census of Population, Ancestry of the 
Population by State: 1980. Supplementary Report PC80-S1-10, April, 1983 (Table 2). 



census respondents in these categories reported multiple ethnic origins. 
Given these findings, and the fact that Hispanic-Americans represent an 
important set of ethnic minority groups in the United States, we felt war- 
ranted in distinguishing between Americans of Hispanic descent and other 
Americans in our analyses.^ 

Almost all Hispanic respondents to the 1982 and 1985 SPPA (99 and 97 
percent, respectively) reported their race as white, and the absolute numbers 
of those who did not were far too small to permit separate analysis.^ 
Therefore, we did not divide the Hispanic respondents by race. 



Race, Ethnicity, and Participation in tlie Arts 

On the basis of these decisions, we concentrate in this report on compar- 
ing the responses to the SPPAs of four groups, three racial and one ethnic: 
White Americans (not of Hispanic descent), African-Americans (not of 
Hispanic descent), Asian-Americans (not of Hispanic descent), and His- 
panic-Americans. (Data permitting the separation of responses from Asian- 
Americans were available for 1985 only.) Although the SPPA surveys were 
designed and weighted to be representative of the racial composition of the 
American population, no such representativeness is guaranteed for the His- 
panic ethnic category. 

Finally, it should be recognized that although we focus on "Hispanic- 
Americans," this category includes members of ethnic groups that differ 
from one another in many respects. '^ Similarly, the African-American and 
white racial categories include members of diverse ethnic backgrounds. 
Indeed, it is possible that variations in behavior within these groups are 
greater than those between them. Our use of broad categories, despite the 
potential differences they may obscure, is based on the coding categories and 
numbers of respondents available to us in these data. We hope that in the 
future, surveys with a larger number of respondents will be conducted so that 
more refined ethnic categories can be employed, allowing researchers to 
discover differences and similarities in arts participation that cannot be 
identified using the categorization scheme currently available. 



Table 1-2 

Comparison of Estimates for Race and Hispanic Origin Between 

1982 and 1985 SPPAs and 1980 Census 





SPPA82 


SPPA85 


1980 Census 


White 


87.1 


87.2 


85.0 


Black 


10.6 


10.8 


10.5 


American Indian 


NA 


0.2 


0.5 


Asian 


NA 


1.6 


1.5 


Other 


2.3 


0.1 


2.5 


Hispanic Origin 


5.6 


6.7 


5.5 



Note: Individuals 1 8 and over only. SPPA figures based on data weighted for race, age, and gender. 
Census figures from Bureau of the Census, 1980 Census of Population, General Population 
Characteristics: U.S. Summary, PC80-1-B1 (tables 43 and 44). Census figures for "Other" calculated 
by subtracting sum of other racial categories from 1 00 percent. 



Race, Ethnicity, and Participation in the Arts 

Underrepresentation. The term "underrepresentation" is pejorative, indicat- 
ing a state of affairs that is unjust. Because the term is value-laden and 
because it has several meanings, we shall avoid it in the narrative of this 
report. Nonetheless, because a concern with "underrepresentation" underlies 
the analyses we undertake, it is necessary to discuss the issue. 

In one sense, members of a racial or ethnic group can be described as 
underrepresented relative to some other group if they participate less fre- 
quently in some activity. We can assess the degree of underrepresentation, 
thus defined, by comparing the rates of participation by different groups. If 
24 percent of Group A reports attending arts and crafts fairs, for example, 
but just 12 percent of Group B, the members of Group B are underrepre- 
sented. We investigate underrepresentation by race and ethnicity in this 
sense in Chapter 2. If one is concemed with equality of result — i.e., if one 
feels that equalizing participation in the arts is itself a legitimate goal of 
public policy — then such differences among groups are a concem in their 
own right. 

By contrast, public policy in the United States has often been concemed 
not with equality of result but with equality of opportunity. From the 
perspective of equality of opportunity, it is less important that members of 
different groups all participate to the same degree than that persons are not 
disadvantaged, by virtue of their racial or ethnic origin, in attempting to 
share a public good. American society tolerates all sorts of inequality, so this 
argumicnt goes, opposing as odious only inequality that results directly from 
statuses such as race or gender into which one is bom. Thus what are 
important are not differences in rates of participation' by members of differ- 
ent groups, but rather differences in opportunities to participate that are a 
consequence of rather than simply associated with, membership in a racial 
or ethnic minority group. 

In this view, the appropriate measure of underrepresentation is the 
existence of a negative effect of racial or ethnic group membership on rates 
of artistic participation, net the influence of people's other characteristics. 
To return to our previous example, imagine that members of Group A attend 
arts and crafts fairs less than members of Group B not because they are 
excluded on the basis of race but because they have less of other charac- 
teristics {e.g., education or money) that are associated with participation. If 
we control for these other characteristics, we can estimate the net effect of 
racial or ethnic origin. This we do, using logistic regression analysis, in 
Chapter 3. 

There is another reason one might wish to look at the "net effect" of race 
or ethnicity on artistic participation rather than the simple association of the 



Race, Ethnicity, and Participation in the Arts 

two. Measures of association, like those in the tables presented in Chapter 2, 
tell us what degree of inequality exists, but they do not tell us why it exists. 
Inspecting the factors that account for such variation in participation, as we 
do in Chapter 3, enables us to assess what would have to change to reduce 
the inequality we see. For example, if differences in the artistic participation 
of different racial or ethnic groups were simply a result of differences in the 
length of time members of different groups stay in school, then equalizing 
educational opportunity would suffice to equalize artistic participation. If 
not, then other programs would be required. 

The factors that lead to participation in the arts may not be the same for 
all groups. If one is concerned with increasing racial or ethnic minority 
participation, then it is important to understand the factors that account for 
participation by members of these groups, and how these factors may differ 
from those predicting participation by members of the majority. In Chapter 

3, we present results of separate analyses for white, African- American, and 
Hispanic respondents to the SPPAs, to explore the possibihty that participa- 
tion in the arts stems from different origins in each group. 

The notion of "underrepresentation" implies that participation in the arts 
is a public good that, like education or political influence, almost anyone 
would find attractive. By contrast, most of us think of our artistic participa- 
tion (or lack thereof) in individualistic, voluntaristic terms. Differences in 
artistic participation, either gross or net, may result from the exclusion of 
some groups from artistic opportunities (either through active discrimina- 
tion, of the kind commonly exercised against African- Americans in the past, 
or through more subtle, perhaps unintended, social pressures that make 
members of minority groups feel unwelcome or uncomfortable at certain 
artistic events). Or they may simply reflect intergroup differences in taste or 
preferences. The SPPA data do not provide such clear accounts of the extent 
to which racial or ethnic differences in participation represent exclusion or 
differences in taste as they do of the extent to which such differences exist. 
But they do permit us to hazard some guesses, which we shall do in Chapter 

4. Note, however, that many arts advocates may not regard such evidence as 
relevant to public policy. In their view, participation in the arts is a good 
thing, and people who do not want more of it may simply have been deprived 
of opportunities that would have awakened them to its virtues. 

Finally, to the extent that underrepresentation (however defined) is a 
concern, it is important to know what subgroups are most underrepresented 
and whether underrepresentation is increasing or decreasing. In Chapter 5, 
we shall present the results of analyses comparing the extent of racial and 
ethnic differences in artistic participation among men and women and 



Race, Ethnicity, and Participation in the Arts 

among Americans of different ages and educational levels. In that chapter, 
we shall also use a special subsample of the 1982 SPPA that enables us to 
explore the impact of childhood experiences and indices of musical taste and 
artistic interest on several kinds of participation, controlling for socio- 
economic factors. 

In summary, "underrepresentation" may mean at least three different 
things: 1) differences in the extent to which members of different groups 
participate; 2) differences in the extent of participation of members of some 
groups compared with members of other groups who are in other respects 
similar to them; or 3) differences in the extent of participation by members 
of different groups attributable to differences in access rather than to 
differences in taste. Each definition entails a different view of art and of the 
nature of a just society, and assessing underrepresentation according to each 
definition requires a different methodological approach. Rather than choose 
one, we address each definition, investigating the first two rather thoroughly 
and the third as much as limited data permit. . 

Artistic participation. No two people define "art" in exactly the same way. 
Some emphasize the most prestigious expressions of "high culture," others 
include such modem forms as jazz or film and "folk arts" or "crafts," and 
still others admit the full range of "the popular arts" distributed by the 
commercial media. People also differ in their estimation of the kinds of 
participation that are most important: some believe that any exposure to art 
is desirable, others value personal encounters more highly than exposure to 
the arts through broadcast or reproduction, and still others maintain that the 
best measure of our cultural well-being is the extent to which people create 
and perform art themselves. 

We call attention to these issues because the extent to which we find 
intergroup differences in participation will depend on where we look, on the 
kinds of art forms and participation about which measures are available in 
our data. Fortunately the 1982 and 1985 SPPAs took a relatively inclusive 
view both of the arts and of participation. The core questions, for example, 
asked respondents about jazz, as well as classical music and opera; and asked 
respondents whether they played a musical instrument (in any kind of 
musical presentation) or acted, sang, or danced onstage, as well as whether 
they watched others do so. Respondents were asked whether they had taken 
classes in photography as well as in painting; and in crafts as well as fine arts 
and art or music appreciation. And one question asked respondents whether 
they enjoyed each of a wide range of musical genres, from country westem 
to chamber music. Moreover, people were asked about the arts they watched 



8 



Race, Ethnicity, and Participation in the Arts 

on television or listened to on radio or recordings, as well as those they 
witnessed live. 

Nonetheless, the SPPA questions tend to reflect both its sponsorship by 
the National Endowment for the Arts, and what is probably a loose consen- 
sus among educated Americans as to what forms of artistic participation 
matter most. Thus the survey focuses predominantly upon the arts that are 
within the domain of the Arts Endowment. These include the kinds of 
performances or presentations that are sponsored by nonprofit cultural 
organizations or public television stations rather than those that are produced 
by commercial media conglomerates. They also focus on the kinds of 
performances that are more likely to take place on a public stage than in 
one's living room or on the street. 

Notwithstanding the legitimacy of these emphases, the choice of activi- 
ties and the way in which questions are worded may have influenced the 
patterns that emerged from the data. For one thing, the SPPA simply did not 
ask specifically about certain activities: e.g., break dancing, graffiti art, clog 
dancing, mariachi music, many kinds of ethnic dance and song, or televised 
soap operas. Consequently, we cannot know if racial and ethnic patterns of 
participation in such activities are different from patterns in the activities 
about which the respondents were asked. It follows that the data cannot yield 
grand generalizations about racial and ethnic differences in "artistic partici- 
pation," in the broadest definition of that term. To the extent that the SPPA 
focuses upon activities that are favored by white college graduates, it 
probably overestimates the extent of the difference between the artistic 
participation of white Americans and of everyone else. 

On the other hand, certain questions were phrased so broadly that the 
absense of racial and ethnic differences in response would not permit us to 
infer that the practices of white, Hispanic-, Asian-, and African-Americans 
were the same. For example, the fact that similar proportions of two groups 
attend live plays does not mean that they are attending the same kinds of 
dramatic productions. It is important, therefore, to remember that our find- 
ings pertain to a circumscribed, although relatively broad, set of artistic 
practices. 



Chapter 2 

RATES OF PARTICIPATION BY 
RACE AND ETHNICITY 



n reporting the rates at which members of three racial and one ethnic group 
participate in several artistic activities, we shall compare responses to 
questions about different kinds of artistic activities to see if the pattern of 
intergroup differences can give us hints as to the sources of racial and ethnic 
variation. These comparisons allow us to document differences in participa- 
tion, but not to explain them. Differences may result 

• from patterns of racial or ethnic exclusion, 

• from differences in taste that are associated with race or ethnicity, or 

• from factors such as educational attainment or occupational status 
that are associated with both race and participation in the arts. 

The Core Activities 

We begin by looking at responses to questions about participation in 10 
core activities about which respondents were asked each month in which the 
surveys were administered. Responses to these questions, weighted by age, 
race, and gender, appear in Table 2-1 . 

Respondents were asked whether they had participated in each activity 
during the previous year, and how many times they had participated during 
the previous month. Because participation rates during the previous month 
were low for all groups, we focus here on whether or not respondents 
reported engaging in each activity during the year before the survey was 
administered.' The way in which this question was phrased means that we 
do not know whether members of different groups who answered affirm- 
atively differed in their frequency of participation over the course of the year. 
The respondent who attended a single play during the previous year, for 
example, is treated no differently than one who attended 20. 



10 



Rates of Participation by Race and Etiinicity 

We report the percentage of the members of different groups who 
participate. This is very different from the percentage of visits or attendance 
for which members of each group account. Previous research indicates that 
a relatively few people account for a large proportion of visits to museums 
and attendance at performing-arts events because they go very frequently.^ 
If one's primary interest is in these high attenders, the data reviewed here are 
of limited value. On the other hand, earlier studies and SPPA evidence on 
participation during the previous month suggest that high attenders represent 
only a small minority of the American population. Thus the data are suitable 
for comparing rates of attendance of groups within that population. 

We can make several generalizations about responses to core questions 
for 1982 and 1985. First, the absolute differences between groups with 
respect to core activities are relatively small, with spreads of from one-tenth 
of 1 percent (Hispanic ballet attendance in 1982) to, at most, almost 24 
percent (Hispanic fiction reading in 1982) between minority groups and the 
white majority. For the most part, absolute percentage differences are low 
because relatively few members of any group participate in the core activi- 
ties (aside from reading literature). For example, the largest percentage of 
any group that attended opera was the 4.58 percent of Asian-Americans in 
1985. The highest rate of visiting art galleries and museums was 26.02 
percent (again for Asian-Americans in 1985). Participation rates for other 
activities were intermediate. 

Taking just those groups for which data are available for both years 
(whites, African- Americans, and Hispanics), we see that participation rates 
were higher for whites in both years for all but attendance at jazz concerts. 
Taking not the absolute differences but rather the odds ratios, we see that for 
some activities these differences were sizable. For example, in both years, 
whites were more than twice as likely as African-Americans to report 
attending a classical music concert, an opera performance, a musical theater 
performance, a play, or a ballet. In 1985, they were more than twice as likely 
to report visiting an art gallery or museum. Non-Hispanic whites were also 
more than twice as likely as Americans of Hispanic origin to report attending 
a play (in both years) and in 1985 attending a classical music concert or an 
opera performance.^ Hispanic respondents reported rates of attendance com- 
parable to those of whites at jazz concerts, operas, and ballet in 1982, 
although the gap widened slightly in 1985. 

Inspecting ratios of the probability of participation in many core activi- 
ties for white respondents to the probability of participation for minority 
respondents makes the differences between groups look large. We can tum 
the measure around and look, instead, at the ratio of the probability that 



11 



Race, Ethnicity, and Participation in the Arts 

Table 2-1 

Participation in Core Arts Activities 

by Race/Ethnicity 



White 



African- 
American 

Hispanic 



Asian 



White 

African- 
American 

Hispanic 
Asian 



Attend jazz 


Attend classical 


1 


concert 


1 


concert 


1982 


1985 


1982 


1985 


9.13 


9.48 


14.42 


14.31 


13,890 


10,861 


13,909 


10,875 


15.64 


13.08 


. 6.67 


6.39 


1,654 


1,384 


1,656 


1,384 


8.27 


6.55 


7.87 


6.77 


940 


788 


941 


789 




7.81 




16.50 




232 




232 




Attend 


\ 


i/lsitArt 




Ballet 




Exhibit 


1982 


1985 - 


1982 


1985 


4.64 


4.72 


23.94 


24.14 


13,913 


10,878 


13,905 


10,872 


1.78 


2.14 


12.47 


10.71 


1,657 


1,385 


1,656 


1,385 


4.54 


3.21 


16.22 


18.18 


941 


790 


941 


790 




6.22 




26.02 




232 




232 



12 



Rates of Participation by Race and Ethnicity 

Table 2-1 (Continued) 

Participation in Core Arts Activities 

by Race/Ethnicity 





Attend opera 
performance 


1 


Attend 
musical 


i 


\ttend 
play 






1982 


1985 


1982 


1985 


1982 




1985 


White 


3.33 
13,901 


2.97 
10,861 


20.67 
13,908 


18.60 
10,873 


13.44 
13,899 




13.10 
10,869 


African- 
American 


1.36 
1.654 


1.43 
1.384 


10.10 
1,656 


8.45 
1,384 


5.82 
1,655 




6.09 
1,383 


Hispanic 


2.52 
940 


0.78 
788 


10.96 
940 


9.52 
789 


5.47 

941 




6.41 
788 


Asian 




4.58 
232 




13.89 
231 






8.87 
232 




Perform on 

musical 

instrument 


Perform: 

act/sing/ 

dance 


1 


Read 
fiction 






1982 


1985 


1982 


1985 


1982 




1985 


White 


4.01 
13,916 


2.98 
10,879 


4.68 
13,916 


4.27 
10,879 


60.19 
13,868 




59.66 
10,852 


African- 
American 


3.35 
1.658 


1.72 
1,385 


4.87 

1,658 


3.49 
1,385 


42.41 
1,651 




43.34 
1,381 


Hispanic 


3.11 
941 


2.03 
790 


2.85 
941 


2.63 
790 


36.45 
938 




41.46 
788 


Asian 




3.82 
232 




4.00 
232 






53.73 
230 



Note: First line to right of racial ethnic category refers to weighted percentage of group engaging in 
activity at least once during 12 months preceding survey. Second line refers to unweighted number 
of respondents. In 1982, Asian-Americans were included in an "Other" racial category. 



13 



Race, Ethnicity, and Participation in the Arts 

minority respondents do not participate in an activity to the probability that 
whites do not participate. Viewed this way, the same intergroup differences 
seem much smaller. For example, in 1982 whites were well over twice as 
likely as African-Americans (a big difference) to have attended an opera, but 
African-Americans were only 2 percent more likely than whites (a tiny 
difference) to have abstained from opera attendance. Similarly, in 1985, 
white respondents were twice as likely as Hispanic respondents to have 
attended a classical music performance, but Hispanics were only 9 percent 
more likely than whites not to have attended. 

Responses of Asian- Americans (available only in 1985) resemble those 
of white Americans. (So few Asian respondents were included in the survey 
that we cannot be as confident that the percentages reflect actual population 
distributions.) Asian respondents were somewhat more Hkely than whites to 
attend classical music concerts, operas, and ballet performances, and to visit 
art galleries and museums. They were less likely than whites to attend jazz 
concerts, musical theater, and stage plays, but white rates of attendance were 
in all cases less than 50 percent greater than those of Asians. 

Rates of public performance were lower than those for attendance at arts 
events for members of all groups, and these rates varied less among the 
groups than those for attendance. African-Americans were more likely than 
others to report singing, acting, or dancing onstage in 1982 (but not in 1985), 
and Asian- Americans were more likely to report performing publicly on 
musical instruments than any other group in 1985. Hispanics were somewhat 
less likely than others to report performing onstage, but the absolute margin 
of difference is very small. 

We may infer from this that artistic participation is more equal among 
ethnic groups with respect to performing than with respect to watching other 
people perform. We would also note, however, that the wording of the 
performance questions was somewhat broader than that of the attendance 
questions, covering any kind of public performance in any musical, dra- 
matic, or dance style. Respondents who answered the music performance 
question affirmatively were then asked if they performed classical music and 
if they performed jazz. One-fourth of the white musical performers and one- 
fifth of the Hispanic reported that they had played classical music, compared 
to one in ten of the African-American musical performers in 1982. (In 1985, 
the comparable figures were 71 percent for Asians, 32 percent for whites, 24 
percent for Hispanics and 19 percent for African-Americans.) In 1982, more 
than one-fourth of the Hispanics and more than one-fifth of the whites who 
reported performing on an instrument in public said that they had played 
jazz, compared to just 16 percent of the African-American instrumentalists. 



14 



Rates of Participation by Race and Ethnicity 

(In 1985, the comparable proportions were 36 percent for Hispanics, 26 
percent for whites, 17 percent for African-Americans, with no Asian report- 
ing a public performance of jazz.) 

Instrumentalists who performed in public but played neither classical 
music nor jazz presumably were playing either folk or ethnic music or some 
form of commercial popular music. Thus we can infer that in 1982, 77 
percent of African-American instrumentalists, compared to 63 percent of 
whites and 70 percent of Hispanics, played exclusively commercial popular 
or folk/ethnic music. (In 1985, the figures were 71 percent for African- 
Americans, 55 percent for whites, 57 percent for Hispanics, and 29 percent 
for Asians.) Had the question been restricted to classical music and jazz, as 
were many questions in the SPPA, ratios of white to African-American 
participation rates would have been higher, and Hispanic participation rates 
would have exceeded those of African-Americans. These results serve well 
to illustrate how the size and direction of intergroup differences are influ- 
enced by the ways in which "the arts" are defined. 

Although the music performance question is unusual in permitting infer- 
ence about participation in popular commercial art forms, other options do 
vary in the extent to which they focus upon conventionally prestigious "high 
culture" activities. Are racial and ethnic differences greater with respect to 
well-established forms of high culture and smaller for more contemporary 
or commercial forms? The gaps between whites and Asians, on the one hand, 
and Hispanics and African-Americans, on the other, are striking with respect 
to classical music and, to a lesser extent, opera and art exhibits. But the gap 
between whites, on one hand, and African-Americans and Hispanics on the 
other, is also sizable with respect to stage plays and musical theater presen- 
tations, often seen as more popular events; and differences between whites 
and Hispanics in ballet are relatively modest. Thus responses to the core 
questions defy generalization on this issue. 

Although African- Americans report the lowest participation levels with 
respect to most of the attendance activities, they report the highest rates of 
attendance at jazz concerts. Jazz is notable because it is the single art form 
included in the core questions that has emerged out of the African-American 
experience; and, although jazz attendance is a minority pursuit among 
African-Americans and jazz has established itself within American music 
more broadly, African-Americans artists are still especially prominent as 
composers and musicians. (The jazz audience is predominantly white, but 
only because there are so many more white Americans than African-Ameri- 
cans.'^) African-Americans were more than 60 percent more Hkely than 
whites to report attending a jazz concert in 1982, and more than 35 percent 



15 



Race, Ethnicity, and Participation in the Arts 



more likely to do so in 1985. This anomalous finding is important, for it 
shows that the gap between African-Americans and whites with respect to 
other kinds of attendance does not reflect a generalized indisposition towards 
attendance at the performing arts within the African-American subsample. 



Differences in Core Participation Among Hispanic 

Ethnic Groups 

The Hispanic subgroups vary with respect to demographic attributes 
such as age and formal educational attainment. These subgroups also vary 
substantially in their participation in core activities of the SPPA. The 1982 
SPPA included data from 425 Mexican- Americans, but only 98 Puerto 
Ricans, 47 Cuban-Americans, and 143 Hispanics with other national back- 
grounds. Evidence from the 1985 survey is even more limited, with 382 
Mexican- American respondents, but just 62 Puerto Ricans, only 22 Cuban- 
Americans, and 131 in the "other Hispanic" category. 

In both 1982 and 1985, respondents in the "other" category participated 
in most activities more than members of the named Hispanic ethnic groups, 
and, especially in 1985, their pattern of participation was similar to that of 
non-Hispanic whites. Puerto Rican and Mexican-American respondents 
tended to report lower levels of participation, although the former were 
relatively more active in 1985 than in 1982. (See Table 2-2.) 

These differences are both suggestive and consistent with what we know 
about differences in age and educational attainment among these segments 
of the heterogeneous Hispanic- American community. But without systema- 
tic sampUng of larger numbers of Hispanic respondents, we can treat these 
findings only as bases for hypotheses to be explored in future research. 



Participation in Other Artistic Activities 

An additional set of questions was asked of approximately one-third of 
the respondents in 1982 and one-sixth in 1985. Several of these questions 
focused on visual and craft arts, and on activities that involved "making" art 
rather than "consuming" it. This set of questions also asked respondents if 
they had read or listened to poetry, visited historical or science museums or 
visited historical monuments. 

Weighted participation rates by race/ethnicity are reported in Table 2-3. 



16 



Rates of Participation by Race and Etiinicity 

Table 2-2 

Participation in Core Activities by Hispanic-Origin 

Ethnic Groups, 1982 and 1985 





Mexican 


Puerto Rican 


Cuban 


Other 




1982 


1985 


1982 


1985 


1982 


1985 


1982 


1985 


Ja77 


6.40 


6.45 


2.02 


11.04 


5.74 


4.41 


8.78 


8.27 


Classical 


7.57 


5.03 


2.89 


6.41 


8.36 


0.00 


9.65 


11.46 


Opera 


1.44 


0.24 


2.40 


2.03 


8.86 


3.85 


5.95 


0.91 


Musical 


10.91 


6.53 


5.84 


17.53 


12.46 


6.23 


12.62 


21.05 


Play 


3.62 


4.83 


3.80 


6.57 


7.35 


3.85 


11.97 


13.29 


Ballet 


4.35 


2.08 


1.79 


1.27 


7.27 


0.00 


7.75 


7.22 


AH Ex. 


13.44 


15.43 


16.84 


18.35 


18.40 


12.07 


19.82 


29.49 


Peilorm: 


2.09 


2.50 


2.05 


0.00 


8.97 


0.00 


0.65 


2.55 


Music 


















Perform: 


2.23 


2.12 


3.99 


0.00 


4.42 


0.00 


0.82 


1.46 


Sing, etc. 


















Read 


29.22 


37.70 


34.72 


39.63 


29.31 


42.65 


48.93 


57.42 


N 


425 


382 


98 


62 


47 


22 


143 


131 



Note: N for Mexican-Americans in 1982 is 424 for attending operas and musicals. N for Mexican- 
Americans in 1985 is 380 for jazz and opera, 381 for plays. N for Other in 1985 is 130 for plays. Ns 
are unweighted, percentages are weighted. 



Whites were substantially more likely to participate in the attendance activi- 
ties and creative writing than Hispanics, who were somewhat more likely to 
do so than African-Americans. For visiting history or science museums and 
historical monuments, differences in African- American and white rates were 
substantial. In 1985, for example, 26 percent of white respondents, but just 
1 1 percent of African -Americans, attended a science or history museum. 
More than 40 percent of the white Americans, but just 18 percent of the 
African -Americans, visited a historical monument. 

White respondents were more likely than African-Americans or Hispan- 
ics to report attending arts or craft fairs in the previous year: 43 percent of 
whites in 1982 and 45 percent in 1985, compared to 27 and 26 percent for 



17 



Race, Ethnicity, and Participation in the Arts 

Table 2-3 

Participation in Other Arts Activities 

by Race/Ethnicity 





Visit 


non- 


Visit hist. 




art musem 


monument 




1982 


1985 


1982 


1985 


White 


. 24.06 


26.08 


40.19 


40.09 




3,461 


1,860 


3,462 


1,858 


African- 


13.20 


11.23 


21.68 


17.50 


American 


416 


249 


417 


248 


Hispanic 


21.09 


16.26 


26.99 


23.95 




186 


144 


185 


144 


Asian 




16.57 




31.06 






37 




37 




Work with 




Weave, 




pottery 


1 


crochet 




1982 


1985 


1982 


1985 


White 


13.29 


12.68 


33.58 


30.67 




3,463 


1,857 


3,463 


1,857 


African- 


6.93 


5.42 


22.97 


15.58 


American 


417 


249 


417 


249 


Hispanic 


8.82 


8.95 


22.20 


16.48 




186 


144 


186 


144 


Asian 




4.14 




38.74 






37 




37 



18 



Rates of Participation by Race and Ethnicity 

Table 2-3 (Continued) 

Participation in Other Arts Activities 

by Race/Ethnicity 





Read/listen 
poetry 


Arts/craft 
fair 


Take art 
lessons 






1982 


1985 


1982 




1985 


1982 


1985 


White 


20.66 
3,461 


20.37 
1,854 


43.18 
3.462 




44.56 
1,857 


11.12 
3.462 


10.98 
1,859 


African- 
American 


15.12 
417 


14.16 
248 


17.14 
417 




15.41 
249 


8.08 
417 


7.03 
249 


Hispanic 


16.83 
186 


14.83 
142 


26.50 
186 




26.03 
142 


10.60 
186 


6.92 
143 


Asian 




20.17 
37 






43.71 
37 




15.56 
37 




Creative 
writing 


Photography, 
film 


Paint or 
draw 






1982 


1985 


1982 




1985 


1982 


1985 


White 


6.70 
3,463 


6.73 
1,857 


11.05 
3,461 




10.67 
1,856 


10.34 
3,463 


9.54 
1,859 


African- 
American 


5.72 
416 


4.56 
249 


8.01 
417 




8.54 
248 


7.59 
417 


5.04 
249 


Hispanic 


6.97 
186 


4.00 
144 


7.91 
185 




5.23 
144 


8.80 
186 


9.01 
143 


Asian 




3.02 
37 






7.71 
37 




7.70 
37 



Note: First line to right of racial/ethnic category refers to weighted percentage of group engaging in 
activity at least once during 12 months preceding survey. Second line refers to unweighted number 
of respondents. Asian-Americans who were included in an "Other" racial category are not included 
here for 1982. 



19 



Race, Ethnicity, and Participation in the Arts 

Hispanics and just 17 and 15 percent for African- Americans. By contrast, 
whites were only somewhat more likely to have read or listened to poetry. 

Data were available from only 37 Asian-Americans, too few for confi- 
dent statistical inference, so participation rates for this group are suggestive 
at best. Asian respondents attended art and craft fairs and read or listened to 
poetry at roughly the same rates as whites. Their attendance at science and 
history museums was comparable to that of Hispanics, whereas their rate of 
visiting historical monuments fell between the white and Hispanic levels. 

The percentages of the African- American, white, and Hispanic groups 
that attended science or history museums — and the differences in those 
rates — were similar to pattems for attendance at art galleries and museums. 
This suggests that the latter differences have as much to do with museum 
visiting per se as with the content or exhibits of art museums. 

Other activities covered in this section of the SPPA were creative 
pastimes that individuals could pursue in private: taking lessons in writing, 
music, arts, dance, or crafts; working with pottery, ceramics, jewelry, 
leather, or metal; practicing a needlecraft (weaving, sewing, or others); 
creative writing; photography, film, or video "as an artistic activity"; and 
painting, drawing, sculpting, or printmaking. (Respondents were also asked 
if they had worked backstage at musical or other kinds of performances, but 
so few had that we do not report the results here.) 

Most intergroup differences with respect to these creative activities were 
strikingly small. In 1985, for example, 1 1 percent of the white respondents, 
compared to 7 percent of both African-Americans and Hispanics, reported 
taking art lessons; 7 percent of the whites, 5 percent of the African- Ameri- 
cans, and 4 percent of the Hispanics engaged in creative writing; 10 percent 
of the whites, 5 percent of the African-Americans, and 9 percent of the 
Hispanics created in the visual and plastic arts. In 1982 Hispanics were 
somewhat more likely to report creative writing than whites. Whites were 
twice as likely as Hispanics to report art lessons, photography, and film 
making in 1985, but not in 1982. Asian-American responses were high to 
moderate compared to those of other groups. 

Differences were greater for creative activity in the craft arts, both 
needlecrafts and other crafts. Whites were 50 percent more likely than both 
African-Americans and Hispanics in 1982, and almost twice as likely in 
1985, to report sewing, weaving, or similar activities. They were almost 
twice as likely in 1982 and more than twice as likely in 1985 as African- 
Americans, and about 50 percent more likely than Hispanics in both years to 
report working with pottery, ceramics, or comparable materials. This pattem 
suggests that with respect to making the visual arts, rates of minority 



20 



Rates of Participation by Race and Ethnicity 

participation relative to those of the white majority are no higher, and in fact 
may be lower, for the craft arts than for more prestigious creative activities 
like drawing, photography, or painting. The data as a whole indicate that 
minority group members are less likely to attend cultural institutions, rela- 
tive to whites, than to be found in the ranks of amateur creative artists. 
Nonetheless, the tendency of white Americans to participate at higher rates 
than others manifests itself in responses to most of these questions. 



Use of the Media for Exposure to the Arts 

A subset of respondents to the SPPAs (approximately one-fourth in 1982 
and approximately one-sixth in 1985) were asked if they had, during the past 
year, seen or heard a jazz performance, a classical music performance, an 
opera, a musical theater production, a stage play, a ballet performance, or a 
visual arts program on television and, where appropriate, radio or sound 
recording. These questions are of particular interest for two reasons. First, 
policy makers have viewed the media, especially television, as an important 
means of increasing exposure to art forms that have benefitted from public 
subsidy. To the extent that participation by minorities in consuming the arts 
via the media is greater relative to the non-Hispanic white majority than their 
attendance at live performances, many would regard such an apparent 
equalization of one kind of artistic opportunity as another benefit of pro- 
grams that promote the arts on television and radio. 

Second, a comparison of differences in the use of media arts programs 
by different groups with those intergroup differences that emerge when we 
look at attendance at live events and exhibitions may provide clues as to the 
origins of the latter differences. Nearly all American families own television 
sets, and nearly all television sets receive one or more public television 
stations, which tend to broadcast fine arts programming. As such, consump- 
tion of the arts on television or radio is costless, except in time. Roughly 
speaking, if intergroup differences are simply a matter of taste, we should 
not expect them to be much reduced when we compare viewing a kind of art 
on television to attending the same activity in person. If they are reduced, 
this suggests that lower levels of live attendance may reflect not simply 
differences in taste, but differences in the resources necessary to attend or in 
the comfort felt in the places where live performances and exhibitions are 
held. 

The results (Table 2-4) are striking in two respects. First, more people 
encountered the arts about which the SPPA asked through the media than in 



21 



Race, Ethnicity, and Participation in tlie Arts 

Table 2-4 
Use of Media for Arts Consumption by Race/Ethnicity 





Watch jazz 
on TV 


Listen jazz 
on radio 




1982 


1985 


1982 


1985 


White 


16.91 
3,288 


14.97 
1,705 


15.78 
3,281 


15.47 
1,699 


African- 
American 


27.95 
366 


37.94 
186 


36.01 
366 


32.42 
186 


Hispanic 


16.06 
203 


15.05 
123 


17.45 
201 


19.43 
124 


Asian 




24.85 
37 




35.90 
36 




Ciassical 
records 




Opera 
on TV 




1982 


1985 


1982 


1985 


White 


23.54 
3,264 


22.49 
1,698 


12.45 
3,288 


12.78 
1,710 


African- 
American 


13.24 
362 


15.48 
186 


9.32 
366 


9.97 
187 


Hispanic 


15.58 
200 


11.03' 
123 


9.71 
203 


12.70 
123 


Asian 




46.63 
36 

Musical 
theater 
on TV 


( 


25.16 
37 

l\/lusical 
theater 
3n radio 




1982 


1985 


1982 


1985 


White 


21.04 
3,279 


17.43 
1,707 


4.22 
3,275 


4.92 
1,700 


African- 
American 


17.21 
366 


17.82 
186 


4.44 
366 


2.80 
186 


Hispanic 


17.83 
203 


17.67 
124 


4.09 
201 


6.50 
124 


Asian 




40.12 
37 




14.10 
37. 



22 



Rates of Participation by Race and Ethnicity 

Table 2-4 (Continued) 
Use of Media for Arts Consumption by Race/Ethnicity 





Listen jazz 
records 


Classical 
music on TV 


Classical 
radio 




1982 




1985 


1982 




1985 


1982 


1985 


White 


18.42 
3,260 




17.03 
1,695 


26.04 
3,287 




24.89 
1,709 


20.49 
3,276 


22.01 
1,703 


African- 
American 


36.62 
361 




36.45 
186 


15.68 
366 




21.88 
187 


15.40 
364 


17.49 
187 


Hispanic 


18.76 
201 




16.18 
124 


21.66 
203 




18.59 
124 


19.90 
201 


18.05 
124 


Asian 






19.71 
36 






41.86 
37 




38.51 
36 




Opera on 
radio 


1 


Opera 
records 








1982 




1985 


1982 




1985 




White 


7.35 
3,269 




7.03 
1,702 


8.23 
3,281 




8.19 
1,709 




African- 
American 


5.32 
363 




4.36 
187 


3.94 
366 




4.31 
186 






Hispanic 


5.26 
201 




6.17 
124 


3.18 
202 




3.03 
124 






Asian 






11.04 
36 






14.10 
37 










Musical 
theater 
records 




Stage 

play 

on TV 




Stage 

play 

on radio 


1 




1982 




1985 


1982 




1985 


1982 


1985 


White 


9.53 
3,271 




8.24 
1,697 


27.86 
3,284 




22.91 
1,707 


3.90 
3,272 


3.93 
1,695 


African- 
American 


1.89 
365 




5.07 
185 


18.21 
366 




18.60 
186 


2.67 
361 


3.90 
183 


Hispanic 


3.40 
200 




3.00 
124 


14.58 
203 




15.38 
124 


6.54 
199 


3.34 
123 


Asian 






19.39 
37 






25.45 
37 


— - 


0.00 
37 



23 



Race, Ethnicity, and Participation in the Arts 

Table 2-4 (Continued) 
Use of Media for Arts Consumption by Race/Ethnicity 







Ballet 






Art 








on TV 






on TV 






1982 




1985 


1982 




1985 


White 


16.98 




15.00 


23.74 




26.75 




3,278 




1,707 


3,275 




1,706 


African- 


10.34 




15.66 


19.48 




23.62 


American 


365 




187 


366 




187 


Hispanic 


15.09 




16.58 


16.37 




18.40 




203 




123 


202 




124 


Asian 






40.92 






38.02 








37 






37 



Note: First line to right of racial/ethnic category refers to weighted percentage of group engaging in 
activity at least once during 12 months preceding survey. Second line refers to unweighted number 
of respondents. In 1982, the "Other" category included Asian-Americans, whereas in 1985 it did not. 
For the media questions, which were asked only two months of 1985, only two respondents were in 
the "Other" category, too few to warrant reporting results. 



live settings. Persons in every racial or ethnic group in each year were more 
likely to see a jazz concert, a classical music performance, an opera, a stage 
play, or a ballet on television (and in the case of the first three, to hear such 
an event on radio or recording) than to attend a live event. This tendency was 
less pronounced for musical theater (which, in 1985, a slightly larger propor- 
tion of the white sample reported seeing live than on television) and for the 
visual arts. (For the visual arts, white and Hispanic, but not African- Ameri- 
can, television viewing were roughly comparable to attendance at galleries 
or museums.) 

Second, the proportionate gap between white and minority attendance 
was smaller in consuming the arts through the media than in live attendance. 
The only exceptions were jazz, where African- Americans were even more 
likely than whites to report hearing jazz on television, radio, and record than 
in live performances; and the substantial gaps between Asians and all other 
groups in media-linked consumption of classical music, opera, musical 
theater, ballet, and the visual arts. (The evidence on Asian-Americans is 
intriguing, but inconclusive because the number of respondents [37] is so 
small.) In other words, although members of all groups were more likely to 



24 



Rates of Participation by Race and Ethnicity 

watch these art forms than to attend, this tendency was more pronounced for 
minority groups than for whites. 

Consider a few examples from 1985. That year, 24 percent of white 
respondents, compared to 6 percent of African-Americans and 7 percent of 
Hispanics, reported attending classical music concerts. By contrast, 25 
percent of the whites, compared to 22 percent of the African-Americans and 
19 percent of the Hispanics, reported watching classical music on television. 
In other words, in 1985 whites were twice as likely as Hispanics and more 
than twice as likely as African-Americans to attend a classical concert; but 
only 14 percent more apt than African- Americans and only 32 percent more 
likely than Hispanics to watch one on television. That same year, whites 
were more than twice as likely as African-Americans, and about 50 percent 
more likely than Hispanics to attend a ballet performance. By contrast, 
slightly larger proportions of the latter groups watched ballet on television 
than did whites. 

These findings are notable for two reasons. First, they tell us that the 
media, especially television, have done much to ensure that African-Ameri- 
cans and Hispanic-Americans are nearly as likely as whites to expose 
themselves to classical music, opera, musical theater, drama, ballet, and the 
visual arts. Second, they indicate that a substantial proportion of African- 
Americans and Hispanics who do not attend such events in person are 
sufficiently interested to watch them on television. This finding suggests the 
potential for developing minority audiences, and leads one to ask why 
African-American and Hispanic-Americans who view the arts on television 
do not attend them live. Do intergroup differences in attendance reflect 
differences in opportunity as well as taste? 

The implications of these data are inconclusive for four reasons. First, 
attending a live event requires more commitment than watching a similar 
program on television. To do the former one must spend time in transit, 
usually pay some money, and face embarrassment if one wishes to leave. By 
contrast, one can view the arts on television free of charge and without 
preparation, and leave a performance by changing the channel. We do not 
know how many respondents who reported watching opera on television, for 
example, did so intently or repeatedly and how many simply spent a few 
minutes watching an opera in between action adventure shows. People who 
watch fine arts events on television but not in person may have less interest 
than those who see them live, albeit more interest than persons who neither 
watch nor attend such activities. 

Second, it is necessary to disentangle watching televised arts programs 
from television viewing generally. African-Americans and Hispanics, on 



25 



Race, Ethnicity, and Participation in tlie Arts 

average, watch more television than whites; however, analyses described in 
Chapter 5 (Appendix Tables 5-7 and 5-14) indicate that their relatively high 
consumption of televised arts programming does not simply reflect a greater 
propensity to watch television. For the full sample, overall television watch- 
ing has a small, significant positive effect on the number of kinds of arts 
programs a respondents watched. But even with that measure controlled, 
African-Americans and Hispanics watch as much arts programming as 
comparable whites. Separate analyses on African-American, Hispanic, and 
white subsamples indicate that within each group, general television watch- 
ing has no significant impact on viewing arts television, with appropriate 
controls. 

Third, we do not know if the musical performances, plays, or dance 
presentations that people watch on television are similar to the ones they 
attend live. It may be that some African-Americans and Hispanics are more 
likely to watch arts events on television than to attend them because they 
prefer the specific programs on television to those available in their commu- 
nities. 

Finally, these data tell us nothing about the quality of the televised arts 
experience. Many would argue that television simply cannot capture the 
sound of a symphony hall or the texture and color of a work of art. Others 
would contend that attending an arts event represents a statement of social 
membership that solitary consumption cannot duplicate. We have no data 
that bear on these issues, which are matters of value rather than social science 
research. Nonetheless, if one holds to either of these views, a world in which 
African-Americans and Hispanic-Americans disproportionately experience 
the arts via media and white Americans experience them disproportionately 
in person does not seem equitable. By contrast, if one believes that it is good 
to have contact with the arts forms about which the media questions asked, 
and that they are as rewarding televised as live, or that watching such events 
on television will lead to attendance, these findings are encouraging.^ 

We explored these issues further by comparing the percentage of respon- 
dents in each group who watched a given kind of arts program on TV and 
who also attended comparable live events to the percentage of those who did 
not see such events on television but did attend in person. For all groups, 
people who watched an arts program on television were more likely than 
others to attend comparable live events (see Appendix Tables 2-2 and 2-3). 
In some cases the tendency was slight: for example, in 1982, Hispanic 
respondents who watched jazz programs on television were only 6 percent 
more likely than those who did not to go to live events. By contrast, Hispanic 
respondents in 1982 who watched classical music programs were more than 



26 



Rates of Participation by Race and Ethnicity 

10 times as likely to attend live classical music performances than were those 
who did not. 

A tendency for arts viewing and attending to be more closely associated 
for African-Americans and Hispanics than for whites (with smaller inter- 
group differences for viewers than for nonviewers) was evident in both 1982 
and 1985 for Hispanic respondents with respect to classical music, musical 
theater, ballet, and art, and for African-American respondents with respect 
to opera and musical theater. These findings may indicate that for these art 
forms television has served to develop an appetite for live attendance among 
new minority audiences. On the other hand, they could mean that African- 
American and Hispanic audiences of these events are more likely than whites 
to pursue their interest by watching them on television; or that watching 
these arts on television is more closely associated with other characteristics 
that lead to live attendance among African-Americans and Hispanics than 
among whites. Therefore, caution demands that these findings be regarded 
as no more than the basis for hypotheses to be pursued in future studies. 



Musical Preferences 

We have hypothesized that more specific definitions of art forms or 
genres vary more markedly with race and ethnicity than categories that are 
broadly defined. During certain months, the SPPA asked respondents if they 
liked to listen to any of a range of musical genres: classical music, opera, 
show tunes, jazz, soul/blues, big band, country western, bluegrass, rock, easy 
listening, folk, barbershop and hymns or gospel. The question did not 
include such genres as rap, salsa, mariachi, cajun, reggae, or polka, for which 
even greater racial or ethnic variation in taste might be expected. Although 
the question is not, strictly speaking, about participation, it provides an 
opportunity to investigate intergroup differences in taste for a wider range of 
musical genres than that about which the core or other participation ques- 
tions ask. 

Responses are described in Table 2-5. Intergroup differences are sum- 
marized by the correlations at the bottom of that table. Correlations between 
groups for 1982 are below the diagonal, for 1985 above it. Correlations on 
the diagonal describe the relationship between each group's own responses 
for 1982 and 1985. A correlation is a measure of association, in this case 
between the percentages of each group who reported liking each kind of 
music, that ranges from -1 .0 to +1 .0. If tastes were perfectly coincident, the 
correlation would be 1 .0. If they were totally opposed, it would be -1 .0. 



27 



Race, Ethnicity, and Participation in the Arts 

Table 2-5 

Percentage Reporting That They Enjoy Specific 

Musical Genres, by Race/Ethnicity, 1982 and 1985 





Whites 


African-Americans 


HIspanics 




1982 


1985 


1982 


1985 


1982 


1985 


Classical 


29.45 


27.13 


15.74 


12.61 


25.68 


31.32 


Opera 


10.41 


11.52 


5.74 


7.05 


5.51 


10.29 


Show tunes 


25.60 


26.77 


12.25 


12.32 


15.51 


2323 


Ja77 


24.52 


30.19 


43.23 


57.82 


26.67 


41.57 


Soul/'blues 


23.07 


28.87 


61.14 


72.45 


28.74 


34.80 


Big band 


35.69 


35.28 


18.53 


20.94 


23.91 


21.52 


Country 


63.68 


57.46 


24.65 


27.10 


49.26 


52.95 


Biuegrass 


28.27 


27.59 


5.07 


3.02 


9.51 


15.85 


Rock 


36.71 


43.17 


29.59 


32.29 


37.49 


51.01 


Easy listening 


52.39 


54.85 


24.93 


43.17 


40.30 


4629 


Folk 


28.00 


27.43 


8.72 


13.66 


18.01 


19.93 


Barbershop 


16.70 


17.53 


4.60 


2,88 


5.18 


7.67 


Hymns/gospel 


34.30 


38.30 


64.49 


65.05 


16.40 


26.61 


N 


4,518 


1,758 


532 


156 


. 277 


113 


Correlations 


White 


African- American 


Hispanic 


White 




.97 




41 




.84 


African-American 




.18 


, 


97 




.56 


Hispanic 




.86 


. 


,37 




.96 



Note: Pearson correlations. 1985 above diagonal. 1982 below. Diagonal=con'elat}on between 1982 
and 1985 for each group. Correlations subject to rounding error. Z-scores presented in Appendix 
Table 2-1. 



28 



Rates of Participation by Race and Ethnicity 

The diagonal correlations indicate that the musical tastes of each group 
were highly stable between 1982 and 1985. Correlations between the tastes 
of white and Hispanic respondents were also very high in both years. (Both 
whites and Hispanics favored country western music above any other genre, 
both liked easy listening music, and few in either group enjoyed opera.) 

In 1982, the correlation between African-American and white tastes was 
. 1 8, positive but nonetheless considerably weaker than any other association 
in the table. African-Americans were less likely than whites to report 
enjoying classical or chamber music, and whites were less likely than 
African-Americans to report enjoying jazz. The largest differences between 
the groups, however, were for soul/blues, country western, easy listening 
and hymns or gospel music. For example, more than 60 percent of African- 
American respondents, but fewer than one in four of the whites, reported 
liking soul or blues music. Less than one-fourth of the African-Americans 
but almost two-thirds of the whites enjoyed country western. More than half 
the whites but fewer than one in four African-Americans liked easy listening 
music. About one-third of the white respondents, but almost two-thirds of 
the African-American respondents, enjoyed hymns or gospel music. Sizable 
minorities of white respondents, but very few African-American respon- 
dents, reported enjoying folk or bluegrass music. Although whites were 80 
percent more likely than African-Americans to report that they liked opera, 
the two groups were similar in that few respondents, African-American or 
white, reported enjoying this form. In 1985, the African-American/white 
correlation (based on a smaller sample than in 1982) rose substantially to .41 . 
Most of the increase resulted from a marked rise in the proportion of 
African-American respondents who reported that they enjoyed easy listen- 
ing music, although there was some slight convergence in taste for opera, big 
band, country western, folk, and hymns or gospel music as well. 

Correlations between Hispanic and African-American tastes were mid- 
way between those between African-Americans and white preference, .37 in 
1982 and .56 in 1985. Like whites, Hispanics tended to enjoy country 
western and easy listening music and were less likely than African-Ameri- 
cans to report enjoying soul music or blues. Like African-Americans, they 
were less likely than whites to like big band music or, in 1982, show tunes 
and bluegrass. Hispanic respondents reported liking hymns or gospel music 
less than either African-Americans or whites. 

These results indicate notable differences associated with race or ethnic- 
ity in a national musical culture dominated by commercially produced 
genres. On the one hand, African-Americans are particularly supportive of 
forms like jazz, soul or blues, and gospel, all of which have deep roots in the 



29 



Race, Ethnicity, and Participation in the Arts 

African-American experience; and relatively uninvolved in such forms as 
bluegrass, barbershop, or, relative to others, country western music, associ- 
ated with white subcultures. But even genres associated with specific racial 
or ethnic communities appear to have permeated a national musical culture. 
Thus approximately one in four whites liked jazz and soul/blues, and an 
equal proportion of African-Americans enjoyed country western music. The 
findings lend support to images neither of racially segmented cultures nor of 
a homogeneous mass society where racial and ethnic differences have 
atrophied. Instead, we find differentiation without segmentation. 

African-Americans were less likely to report liking classical music than 
white or Hispanic respondents, and opera was enjoyed by only small minori- 
ties in any group. In 1985, Hispanics were more likely than whites to report 
enjoying classical music, ranking it sixth among the thirteen genres, higher 
than whites, who ranked it tenth, or African-Americans, for whom it ranked 
ninth. It is thus striking that in 1985 Hispanics were only one half as likely 
as whites to have reported attending classical music concerts. 



Childhood Experience in the Arts 

Advocates of arts education sometimes assert that appreciation of the arts 
must be cultivated from childhood if one is to understand and care about 
them as an adult. Sociologists sometimes refer to the familiarity with the fine 
arts with which educated parents endow their children as "cultural capital," 
analogous to bequests of financial capital as a means to ensure that one's 
children get ahead in life.^ If this emphasis on early artistic experiences is 
justified, then it is possible that intergroup differences in participation reflect 
differences in the way that children of these groups were socialized. 

Fortunately, the SPPAs asked a portion of the respondents (about one- 
third in 1982 and approximately one-sixth in 1985) about their childhood 
experiences with respect to a variety of art forms. Four questions concerned 
socialization by parents "when you were growing up." Respondents were 
asked if their parents "often, occasionally, or never" listened to classical 
music, took them to art museums or galleries, took them to plays, dance or 
classical music performances, or encouraged them to read books "which 
were not required for school or religious studies." Responses to these 
questions are presented in Table 2-6. 

People often have difficulty recalling events that happened in their 
distant past, and we all have some tendency to reconstruct our childhoods so 
as to make them consistent with our subsequent experience. We do not know 



30 



Rates of Participation by Race and Ethnicity 

Table 2-6 
Cultural Socialization in Family by Race/Ethnicity 





Parents listened 

classical 

music 


Parents took 

art museums/ 

galleries 


Parents took 

plays/dance/ 

classical 


Parents 

encouraged 

reading 




1982 


1985 


1982 


1985 


1982 


1985 


1982 


1985 


White 


32.64 
4,563 


34.19 
1,913 


35.33 
4,567 


36.64 
1,912 


33.05 
4,561 


33.59 
1,910 


40.36 
4,567 


39.08 
1,915 


African- 
American 


18.84 
507 


22.17 
197 


26.86 
508 


26.07 
198 


26.31 
511 


29.13 
194 


32.82 
511 


37.91 
199 


Hispanic 


16.56 
302 


25.05 
140 


22.76 
304 


27.09 
141 


20.36 
302 


23.16 
141 


22.26 
305 


20.06 
141 


Asian 




48.70 
39 




43.30 
39 




32.86 
39 




46.66 
39 



Note: First line to right of racial/ethnic category refers to weighted percentage of group reporting 
parents engaged in activity "occasionally or often" (for first three columns) or "often" (for "encouraged 
reading"). Second line refers to unweighted number of respondents. In 1982, Asian-Americans were 
in an "Other"category (not included). 



whether such distortions bias the responses affirmatively or negatively, or 
whether, by contrast, individual distortions more or less balance one another 
out. To the extent we are interested in comparisons between groups, we need 
be concerned less by absolute bias than by the possibility that responses from 
different racial or ethnic categories are flawed by different degrees (or 
directions) of biased recall. John Robinson and his colleagues have sug- 
gested that question ordering in the SPPA may have made childhood arts 
experiences more salient to respondents who had reported engaging in 
related arts activities. If this were the case, we would expect such tendencies 
to yield exaggerated differences between whites and members of other 
groups in the tables that follow.^ 

White respondents were most likely to report that their parents at least 
occasionally Hstened to classical music. In 1982, 33 percent of whites 
compared to 19 percent of African-Americans and 17 percent of Hispanics 



31 



Race, Ethnicity, and Participation in the Arts 

answered in this way. (In 1985, the figures were 34 percent for whites, 22 
percent for African-Americans, 25 percent for Hispanics, and 49 percent for 
Asians. Regrettably, the small number of Asian- American respondents pre- 
vents us from placing much stock in the latter arresting figure.^) These 
differences are comparable to those for attending classical music concerts 
and greater than those for viewing or listening to classical music programs 
on television or radio.^ 

In 1982, white respondents were 32 percent more likely than African- 
Americans to report that their parents took them occasionally or often to art 
museums or galleries when they were young. *^ (1985 results were similar.) 
In 1982, they were 55 percent more apt to report such experiences than were 
Hispanic respondents, whereas in 1985 Hispanics were more similar to 
African- American respondents. (More than 40 percent of the Asian respon- 
dents — compared to 37 percent of the whites — reported such early expe- 
rience in 1985.) If we compare these results to reports of visits to art 
museums in the past year (Table 2-1), we see that the gap between white and 
African-American respondents is somewhat greater than we would expect 
on the basis of these early socialization experiences, whereas the difference 
between white and Hispanic respondents is approximately the same. By 
contrast, the difference in the proportion of whites and African-Americans 
who report watching visual arts programming on television is somewhat less 
than we would expect on the basis of parental socialization. 

Because the question about going to performing arts events with one's 
parents was worded to include attendance at plays, dance, or classical music 
performances, it does not admit to straightforward comparison with any of 
the core participation questions. The responses are comparable to those for 
visits with parents to art museums and art galleries, with approximately one- 
third of the whites in both years, compared to 26 and 29 percent of the 
African-American respondents (1982 and 1985, respectively) and 20 and 23 
percent of the Hispanic group reporting that their parents at least occasion- 
ally took them to concerts and plays when they were young. ^^ 

In 1 982, 40 percent of the white respondents, compared to 33 percent of 
the African-American respondents and 22 percent of the Hispanics reported 
that their parents often encouraged them to do reading that was not required 
as part of school or religious instruction. (In 1985, with smaller samples, the 
figures were 39 percent, 38 percent, and 20 percent. Of the few Asian 
respondents, 47 percent reported such parental encouragement.) If we com- 
pare these responses to those for the core question on whether respondents 
had read novels, short stories, poetry, or plays during the previous year, we 
see that the proportionate gaps between African-American and white respon- 



32 



Rates of Participation by Race and Etiinicity 

dents are somewhat greater than one might expect on the basis of responses 
to the parental socialization question, whereas the differences between 
whites and Hispanics are somewhat less. 

The SPPAs also asked respondents if they had taken classes or lessons 
in voice or an instrument, art, acting, ballet, creative writing, craft arts, art 
appreciation, or music appreciation at various periods in their lives. Table 
2-7 reports the proportion that never took each kind of class, as well as the 
percentage of respondents who took their first class of each kind when they 
were under the age of 12, between the ages of 12 and 17, and older than 17. 

White respondents were more likely each year to report taking each kind 
of art class or lesson than were African-American or Hispanic respondents. 
Similarly, with just one exception, African-Americans were more apt to 
report having taken classes in each area in each year than Hispanics. ^^ As 
was the case for other questions asked in only one month of 1985, the 
number of Asian respondents was too small to yield conclusive results. 

Focusing upon 1982, for which the number of African- American and 
Hispanic respondents to these questions was substantially higher than in 
1985, the absolute gap between whites and African-Americans ranged from 
10 percent (50 percent of the whites compared with 40 percent of the 
African-Americans) for vocal or instrumental lessons, to less than 1 percent 
(22 percent of the whites and 21 percent of the African- Americans) for music 
appreciation courses. The ratio of white to African- American participation 
ranged from two to one (for ballet lessons, taken by 8 percent of the whites 
and just 4 percent of the African- Americans), to 1.03:1, for music apprecia- 
tion courses. Among class types taken by substantial minorities of all 
respondents, whites were 38 percent more likely than African- Americans to 
report classes in the visual arts, 33 percent more apt to report taking creative 
writing classes, 40 percent more likely to report classes in the craft arts, and 
31 percent more likely to report art appreciation classes. 

The proportion of Hispanic-Americans who indicated that they had 
taken classes or lessons was comparable to, although slightly lower than, the 
African-American percentage for the visual arts, acting, and ballet classes. 
Hispanics were just 62 percent as likely as whites and 82 percent as apt as 
African-Americans to report taking creative writing courses. For craft art 
courses the comparable figures were 59 percent and 83 percent. For art 
appreciation courses, they were 53 and 70 percent, respectively. Hispanics 
were especially unlikely to have taken music lessons or music appreciation 
courses. Only 22 percent of the Hispanics, compared to 40 percent of the 
African-American and 50 percent of the white respondents, reported taking 
vocal or instrumental classes or lessons. And only 9 percent, as compared 



33 



Race, Ethnicity, and Participation in thie Arts 

Table 2-7 

Age at First Class or Lesson In Selected Arts Subjects by 

Race/Ethnlclty, 1982 and 1985 





Age at 


Music 




Art 


Acting 


Ballet 




first class 


class 


class 


class 


class 




1982 


1985 


1982 


1985 


1982 


1985 


1982 


1985 


White 


Never 


49.51 


48.92 


74.55 


72.16 


90.05 


89.16 


92.00 


90.81 




<12 


26.30 


28.68 


3.00 


3.93 


1.07 


1.00 


5.57 


7.05 




12-17 


19.93 


18.61 


11,41 


12.16 


6.03 


6.73 


0.92 


0.89 




>17 


4.27 


3.78 


11,04 


11,75 


2.84 


2.56 


1.51 


1.25 


African- 


Never 


59.55 


62.96 


81.60 


82.77 


93.08 


91.53 


96.17 


97.04 


American 


<12 


13.81 


14.23 


2.44 


0.89 


1.02 


0.75 


1.73 


1.14 




12-17 


21.23 


17,31 


11.14 


11.98 


3.80 


6.85 


1.04 


1.40 




>17 


5.41 


5.51 


4.82 


4.36 


2.09 


0.88 


1.05 


0.43 


Hispanic 


Never 


77.65 


76.07 


82.97 


88.71 


92.69 


95.27 


96.56 


97.28 




<12 


6.38 


7.37 


2.48 


0.00 


0.60 


0.00 


1.85 


0.44 




12-17 


14.20 


11.24 


9.97 


7.74 


4.60 


3,79 


0.00 


1.41 




>17 


1.77 


5.32 


4.59 


3.55 


2.11 


0.94 


1.60 


0.87 


Asian 


Never 




59.88 




70.88 




88.04 




94.59 




<12 




15.04 




6.44 




0.00 




0.00 




12-17 




8.53 




11.76 




7.37 




2.75 




>17 




16.55 


- 


10.91 




4.59 




2.66 



with 21 percent of the African-American respondents and 22 percent of the 
whites, reported ever taking a course in music appreciation. 

The age at which persons first took classes or lessons varied by kind of 
lesson, with music and ballet lessons often taken during the elementary school 
years, and music and art appreciation often taken after the age of 17. Such 
patterns differed somewhat by race and ethnicity, however. For example, 26 
percent of the white respondents, but just 14 percent of the African-American 
respondents, reported taking voice or instrumental lessons before the age of 
12. By contrast, African- Americans were slightly more Hkely than whites to 
take such lessons during the high school years and after the age of 17. 
Similarly, African-Americans were somewhat more likely than whites to take 
music appreciation courses during the high school years, and somewhat less 
Hkely to take them before or after. This pattern of relatively equal African- 
American/white participation during the high school years, and less equal 
participation before and/or after high school tended also to be the case for 



34 



Rates of Participation by Race and Etiinicity 

Table 2-7 (Continued) 

Age at First Class or Lesson in Selected Arts Subjects by 

Race/Ethnicity, 1982 and 1985 





Creative 


Craft 


Art 


Music 


Number 




writing 


art 


appreciation 


appreciation 


1 


of 




class 


class 


class 


class 


respo 
1982 


indents 




1982 


1985 


1982 


1985 


1982 


1985 


1982 


1985 


1985 


White 


80.78 


78.94 


66.51 


62.44 


79.31 


78.50 


78.47 


78.27 


4,590 


1,923 




0.75 


1.07 


3.47 


4.25 


0.81 


0.76 


2.32 


1.95 








7.51 


9.21 


15.37 


17.95 


6.16 


6.50 


9.02 


9.39 








10.96 


10.78 


14.66 


15.36 


13.71 


14.24 


10.18 


10.39 






African- 


85.60 


87.90 


76.04 


71.79 


84.22 


83.17 


79.10 


83.03 


515 


199 


American 


1.05 


0.78 


3.57 


0.93 


0.68 


0.30 


1.73 


1.11 








6.21 


4.37 


13.00 


15.79 


6.02 


11.17 


11.01 


6.73 








7.14 


6.96 


7.38 


11.49 


9.08 


5.36 


8.16 


9.13 






Hispanic 


88.17 


95.91 


80.08 


84.31 


88.99 


93.06 


91.02 


93.02 


305 


143 




0.98 


0.00 


2.14 


2.22 


1.09 


0.00 


1.07 


0.44 








4.71 


1.64 


12.17 


8.99 


4.59 


4.00 


3.54 


3.22 








6.14 


2.45 


5.61 


4.49 


5.32 


2.94 


4.37 


3.32 






Asian 




89.78 




81.39 




88.89 




87.32 




39 






0.00 




0.00 




0.00 




4.56 










5.96 




6.45 




5.96 




8.13 










4.26 




12.15 




5.15 




0.00 







Note: Figures under class names refer to weighted percentage of group first engaging in activity at 
age indicated. Last two columns indicate unweighted number of respondents. (Ns for each group for 
each year were the same for all classes, except that N=4,589 for white respondents with respect to 
writing classes and craft art classes in 1982.) In 1982, Asian-Americans were coded in "Other" 
category (not included). 



other kinds of classes. Although the data are ambiguous because respondents 
were not asked where they took lessons or classes, the findings do suggest 
that U.S. secondary schools have tended to equalize access to arts training 
between white and African-American students. They do not seem to have 
done this for Hispanic-Americans, however. ^^ 

Whether such classes have had a long-term effect is another issue. If we 
assume that music appreciation courses focused on classical music, then the 
gaps between white and African-Americans in attending classical concerts 
and watching televised classical music programs are larger, and the differ- 



35 



Race, Ethnicity, and Participation in tlie Arts 

ences between African-Americans and Hispanics smaller (or in the opposite 
direction), than what one would predict on the basis of these responses. This 
could be the case if African-American students took different kinds of 
courses than whites, or if whites and Hispanics had more opportunities to 
develop a taste for classical music without taking classes, African- Ameri- 
cans are less likely than whites to visit art museums or galleries but more 
likely to watch visual-arts programs on television than we would expect 
from the rate at which they have taken art appreciation courses, whereas the 
white/Hispanic gap in the rate of visiting art exhibits is lower than the 
art-appreciation data would lead one to predict. The difference between the 
percentage of whites, on the one hand, and both African- Americans and 
Hispanics on the other, who report that they currently practice creative 
writing and painting or drawing is less than one would expect on the basis 
of differences in the proportion of these groups who have taken art or 
creative writing courses. ^"^ 



Summary 

The data reported above are too complex to summarize facilely, but one 
fact emerges clearly. African- Americans and Hispanics are statistically 
underrepresented, relative to whites, among those who attend fine-arts 
events, both performances and exhibitions. They also tend to be less likely 
than whites to participate in the fme arts by watching them on television and 
by engaging in amateur practice, but the differences are proportionately 
smaller than for most kinds of live attendance. White Americans are also 
more likely than African-Americans and much more likely than Hispanics 
to report that they have been socialized into the fine arts (and reading) by 
parents and by classes or lessons. With respect to core participation, the only 
set of questions for which there were a sizable number of Asian respondents, 
Asian-Americans were notable for their rates of attendance at classical 
concerts, art exhibits, and ballet and opera performances, all of which 
exceeded the white rate. 

Despite their superficial similarity in comparison to those of whites, the 
response patterns of African-Americans and Hispanics are distinct. Fewer 
Hispanics than African-Americans reported benefiting from most of the 
socialization experiences about which respondents were asked. Yet their 
rates of participation through watching the arts on media were similar to 
those of African- Americans (but higher for classical music and ballet), as 
was their participation in creative practice (with somewhat higher rates for 



36 



Rates of Participation by Race and Ethnicity 

most visual arts or crafts activities). Hispanics were also more likely than 
African-Americans to visit art exhibits. 

These patterns point to relatively low participation of African-Ameri- 
cans as fine-arts attenders (except for jazz concerts) and art exhibition 
visitors that cannot be explained by artistic socialization alone. Moreover, 
the fact that African- Americans attend jazz concerts at higher rates than 
whites (or Hispanics or Asians) indicates that they are not uninterested in 
performing-arts events per se. It is likely that had the SPPA questions 
emphasized artistic genres with closer historical links to the African- Ameri- 
can and Hispanic communities, artistic participation for these groups would 
have been as high or higher than that of whites. 

Nonetheless, we cannot assume that relatively low rates of African- 
American and Hispanic participation among attenders of fine-arts events 
simply reflect lower interest in or liking for such activities. The fact that the 
proportionate gap between white respondents, on the one hand, and African- 
American and (to a somewhat lesser extent) Hispanic respondents on the 
other, was greater for live attendance than for media participation indicates 
that there is interest in both the African-American and Hispanic populations 
in the fine arts that is not being manifested in attendance. Moreover, given 
the relatively small differences in the proportion of African-Americans and 
whites who take art and, especially, music appreciation courses, the low rates 
at which African-Americans attend classical music concerts, opera and ballet 
performances, and art exhibits are surprising. Add to this the greater statis- 
tical overrepresentation of African-Americans relative to whites for watch- 
ing jazz on television and listening to it on radio and recordings than for 
attendance at live concerts, and it appears that some factors other than taste 
may inhibit the attendance of African-Americans at live performing-arts 
events and art exhibits. 

If one believes that the kinds of arts participation about which the SPPA 
asked are so important that intergroup differences, of whatever origin, are 
unacceptable, then these findings are of grave concern. If one believes that 
such intergroup differences are unacceptable only if they reflect differences 
in opportunity, rather than differences in preferences, then these patterns 
raise cause for concern, but do not demonstrate conclusively that such 
concern is warranted. 

The reader should be aware, however, that almost all of the activities 
about which respondents were asked (except for reading novels, short 
stories, poetry, or plays) are ones in which only a minority of all respondents 
participated during the year prior to the survey. With respect to many 
activities (for example, attending opera or ballet performances or performing 



37 



Race, Ethnicity, and Participation in the Arts 

onstage), these minorities were very small ones. If one believes that the goal 
of policy should be to increase the number of minority Americans engaging 
in these activities rather than to make participation rates equal, this could be 
accomplished more effectively for most activities by doubling the current 
rates of participation of all groups than by bringing African-American and 
Hispanic rates up to white levels. 

These findings tell us that whites, African-Americans, Hispanics, and 
Asian-Americans participate in a wide range of artistic activities at unequal 
rates, but they do not tell us why these differences exist. If one believes that 
racial or ethnic differences of the sort identified here are only problematic if 
they seem to be explained by race or ethnicity, as opposed to being just 
associated with race or ethnicity, then these findings are not sufficient. In 
Chapter 3, we investigate the net effects of African- American and Hispanic 
origin on SPPA core participation rates among otherwise similar respon- 
dents, and address certain questions this chapter has posed but not answered. 



38 



Chapter 3 

NET EFFECTS OF RACE AND 
ETHNICITY ON PARTICIPATION IN 
SPPA CORE ACTIVITIES 



n Chapter 2, we observed persistent differences between the rates of 
participation of African- Americans, Hispanics, Asian-Americans, and 
whites in the artistic activities about which the SPPAs asked. Comparison 
of patterns of responses to different questions suggested that, with certain 
exceptions, differences by race were stronger for live attendance than for arts 
consumption through the media, stronger for live attendance than for art- 
producing activities, and stronger for performing "high culture" music than 
jazz or popular music. 

It is one thing to establish that racial or ethnic groups vary in the rates at 
which they participate in certain cultural activities. It is quite another to 
demonstrate that these differences result from race or ethnicity, rather than 
being by-products of other differences between such groups. The major goal 
of this chapter is to determine the extent to which differences among whites, 
African-Americans, and Hispanics stem from group membership itself, or 
originate in sociodemographic differences among these groups.^ In other 
words, we shall ask whether members of these groups would participate at 
different rates were they identical with respect to sociodemographic position 
as measured by the SPPA. 

These analyses are both of intrinsic interest and of interest for their 
relevance to public policy towards the arts. If one believes that racial or 
ethnic differences in participation are objectionable only if they flow directly 
from race or ethnicity, the results of this chapter will permit one to see to 
what extent this is the case. If one regards intergroup differences as lamen- 
table whatever their origin, the analyses in this chapter will provide clues as 
to how they might be modified. 

We cannot assume, however, that the factors that lead people to partici- 
pate in the arts are the same for members of different racial or ethnic groups. 



39 



Race, Ethnicity, and Participation in tlie Arts 

After exploring the net effects of race and ethnicity on participation, we 
analyze the sociodemographic determinants of participation in the "core" 
activities separately for each group. These separate analyses permit us to 
judge the extent to which the same factors account for variation within each 
group. The results, should they differ, may suggest that different kinds of 
programs are necessary to extend opportunities for participation to members 
of different groups. 

We restrict our analyses to the core participation questions because these 
activities are of particular interest and because they were asked throughout 
the survey periods, thus yielding large African- American and Hispanic 
subsamples.^ Because these questions cover only a limited range of activi- 
ties, the findings should not be generalized to other forms of participation in 
the arts. 



Explaining Raciai and Ethnic Differences 

In this section, we predict participation in each of the core activities as a 
function of race, ethnicity, and sociodemographic characteristics.^ For each 
core activity, we executed two predictive models: one including only racial 
or ethnic origin, and one including racial/ethnic origin and sociodemo- 
graphic measures. By comparing the size of coefficients estimating the 
influence of racial or ethnic group membership on participation with and 
without controls, we can estimate the percentage of intergroup differences 
for which sociodemographic differences account. 

Because the dependent variables — the participation measures — are 
binary, taking the value of " 1 " if the respondent did participate and "0" if he 
or she did not, we use a method designed for such variables, called logit or 
logistic regression analysis^ Race and ethnicity are included in the models 
as a series of dichotomous or "dummy" variables, taking the value of "1" 
when the respondent is a member of the group in question and "0" when he 
or she is not. To use dummy variables in this way, it is necessary to exclude 
a category. In these analyses, whites are the excluded category. Coefficients 
for other groups represent the impact of group membership on the prob- 
ability of participation (net the effects of other independent variables in the 
model) compared to the participation rates of comparable white respondents. 
For the 1982 data, we included "African- American," "Hispanic," and 
"Other" as racial/ethnic categories. (Because we do not know who is in the 
"Other" category, we do not report results for this group.) In models for 
1985, we excluded the very few "Other" respondents from the analyses and 



40 



Net Effects of Race and Ethnicity 

included "African-American," "Hispanic," and "Asian" as racial/ethnic vari- 
ables. 

In interpreting these results, we focus upon the coefficients comparing 
the net participation of each racial or ethnic group in the activity in question. 
By way of illustration, consider the section of Table 3-1 reporting the effect 
of being African-American on attending classical music concerts in 1982. 
(These results are reported under 1982 to the right of the rows labeled "A" 
under the column headed "attends classical concerts.") Column 1 reports the 
results of the model including only the racial/ethnic dummy variables. 
Column 2 reports the results of the model including the racial/ethnic dummy 
variables with sociodemographic controls. Under each column, the row 
labeled "b" reports the logistic regression coefficient indicating the net 
influence of being African-American on attending classical music concerts. 
The coefficient in column 1, for example, is -.845. Because there are no 
controls in the model reported in column 1 , this figure is comparable to the 
descriptive percentage results reported in Table 2-1 . (Its negative sign means 
that African-Americans are less likely to attend than the omitted group, i.e., 
whites.) Column 2 reports the effect of being African- American on attending 
classical music concerts, controlling for a wide range of sociodemographic 
differences between the white and African-American respondents. Because 
the coefficient is less than that in column 1 but nonetheless remains negative, 
it indicates that part, but not all, of the difference between African-Ameri- 
cans and whites is attributable to sociodemographic differences between the 
two groups. By dividing the coefficient in column 2 (-.566) by the coeffi- 
cient in column 1 and subtracting the result from unity, we can conclude that 
roughly 33 percent of the difference in rates of participation between Afri- 
can-Americans and whites resulted from measured sociodemographic dif- 
ferences between the two groups, whereas the remainder stems from other 
sources. 

We shall not discuss the standard error (the figure immediately under the 
logistic regression coefficient), which is of interest only to statistically 
sophisticated readers. Of more general interest is the alphabetical notation 
below that (in the row labeled "sig''). Probability theory tells us that when 
one uses a sample from a larger population, one gets some positive or 
negative coefficients simply by chance. The letters in the significance rows 
of Table 3-1 (keyed to an explanation at the end of the table) tell us how 
likely it is that a coefficient of a given magnitude would occur by chance. 
The letter "c" in the significance row of column 2 under classical music (for 
African-Americans in 1982) tells us that such an effect (-.566) would be 
estimated by chance fewer than 5 times out of 100,000. This is a very high 



41 



Race, Ethnicity, and Participation in the Arts 

Table 3-1 

Coefficients Representing Effects of African-American (A) and 

Hispanic (H) on Core Participation Items (1) with Race/Ethnicity 

Only and (2) witli Demographic Controls 

Jazz Classical 



1982 


1 


2 


1 


2 


A b 


.631 


.683 


-.845 


-.566 


se 


.071 


.084 


.097 


.110 


sig 
H b 


c 
-.090 


c, 
.075 


c 
-.667 


c 
-.071 


se 


.121 


.133 


.123 


.137 


sig 


NS 


NS 


c 


NS 


1985 


1 


2 


1 


2 


A b 


.381 


.453 


-.868 


-.557 


se 


.084 


.101 


.111 


.128 


sig 
H b 


c 
-.382 


c 

-.272 


c 
-.805 


d 
-.261 


se 


.140 


.156 


.137 


.153 


sig 


a 


NS 


c 


NS 




Ballet 






Art 


1982 


1 


2 


1 


2 


A b 


-.967 


-.781 


-.774 


-.617 


se 


.182 


.202 


.074 


.086 


sig 
H b 


c 

-.000 


b 
.511 


c 
-.486 


e 

-.039 


se 


.161 


.179 


.090 


.104 


sig 


NS 


a 


c 


NS 


1985 


1 


2 


1 


2 


A b 


-.799 


-.536 


-.948 


-.790 


se 


.187 


.205 


.088 


.102 


sig 
H b 


c 
-.384 


a 
.094 


c 
-.332 


c 
.101 


se 


.196 


.211 


.090 


.105 


sig 


a 


NS 


b 


NS 



42 



Net Effects of Race and Ethnicity 

Table 3-1 (Continued) 

Coefficients Representing Effects of African-American (A) and 

Hispanic (H) on Core Participation Items (1) with Race/Ethnicity 

Only and (2) with Demographic Controls 



Opera 



Musical 



Play 



1982 


1 


2 


1 


2 


1 


2 


A b 


-.900 


-.582 


-.825 


-.567 


-.903 


-.674 


se 


.208 


.240 


.081 


.093 


.103 


.118 


sig 


c 


a 


c 


c 


c 





H b 


-.275 


.356 


-.734 


-.314 


-.970 


-.505 


se 


.212 


.233 


.106 


.119 


.145 


.160 


sig 


NS 


NS 


c 


a 


c 


a 


1985 


1 


2 


1 


2 


1 


2 


A b 


-.715 


-.306 


-.884 


-.562 


-.815 


-.603 


se 


.228 


.249 


.098 


.111 


.113 


.133 


sig 


a 


NS 


c 





c 





H b 


-1.334 


-.832 


-.753 


-.339 


-.761 


-.359 


se 


.388 


.428 


.117 


.131 


.140 


.159 


sig 


b 


NS 


c 


a 


c 


a 




Instrument 


Act, 


sing 




Read 


1982 


1 


2 


1 


2 


1 


2 


A b 


-.194 


-.191 


.040 


.042 


-.701 


-.501 


se 


.137 


.147 


.116 


.127 


.051 


.062 


sig 


NS 


NS 


NS 


NS 





c 


H b 


-.271 


-.352 


-.518 


-.466 


-.951 


-.579 


se 


.191 


.217 


.199 


.212 


.070 


.084 


sig 


NS 


NS 


a 


a 


c 


c 


1985 


1 


2 


1 


2 


1 


2 


A b 


-.547 


-.317 


-.195 


-.043 


-.629 


-.456 


se 


.209 


.221 


.151 


.169 


.056 


.069 


sig 


a 


NS 


NS 


NS 


c 


c 


H b 


-.381 


-.103 


-.489 


-.293 


-.705 


-.284 


se 


.245 


.254 


.215 


.228 


.071 


.086 


sig 


NS 


NS 


a 


NS 


c 


b 



Note: b is the logistic regression coefficient; se is the standard error; sig refers to the level of statis- 
tical significance, where a=probability less than .05, b=probability less than .001, c=probability less 
than .00005, and NS=not significant. 



43 



Race, Ethnicity, and Participation in the Arts 

level of statistical significance and enables us to conclude that African- 
Americans really were less likely to attend classical music concerts than 
whites, as the negative regression coefficient indicates. 



Racial/Ethnic Effects Net of Sociodemographic 

Differences 

In chapter 2 we raised the possibility that differences in participation 
between whites, on the one hand, and African- Americans and Hispanic- 
Americans, on the other, might result simply from the fact that whites as a 
group are economically better-off. Because educational attainment and oc- 
cupational status are associated with patterns of leisure activity and interest 
in the arts, it seemed reasonable to expect that at least some of the differences 
we observed stemmed from sociodemographic differences between whites 
and members of other groups. 

We explored this possibility by including race/ethnicity in a predictive 
model that controlled for a wide range of socioeconomic and demographic 
characteristics. These characteristics included: gender, three categories of 
residence (central city, other SMSA, and outside an SMSA), age, education, 
income, seven broad categories of occupation (1982: professional and tech- 
nical; managerial and administrative; sales and clerical; craft, operative, 
service, farm, transport, laborers, private household, and armed forces; 
unknown; unemployed and retired; keeping house; and student; 1985: ex- 
ecutive, administrative, managerial; professional; technical, sales, and ad- 
ministrative support; craft, operative, service, farm, armed forces; unknown; 
unemployed and retired; keeping house; and student); and five categories of 
marital status (married, widowed, divorced, separated, and single).^ 

To what extent are differences in participation in the core activities 
attributable to differences among groups in the sociodemographic controls? 
No single generalization applies to African-American, Hispanic, and Asian 
respondents. 

Variation between African-Americans and whites. Differences in participa- 
tion between African-Americans and whites were partially attributable to 
sociodemographic differences between these two groups; but significant 
differences tended to persist even in the presence of sociodemographic 
controls. In both 1982 and 1985, African-Americans were significantly less 
likely than whites, even after controlling for sociodemographic factors, to 



44 



Net Effects of Race and Ethnicity 

attend classical music concerts, musical theater performances, plays, ballet 
performances, and art exhibitions, and significantly less likely to report 
reading novels, plays, poems, or short stories. In 1982, but not 1985, 
significant differences in opera attendance between African-Americans and 
whites remained after controls, as well. For reading and for attendance at 
classical music concerts, musical theater performances, plays, and art exhib- 
its, the differences, net sociodemographic factors, were highly significant. 
For these activities, there are small but persistent differences between Afri- 
can-Americans and whites that cannot be attributed to the different sociode- 
mographic characteristics of these two groups. 

Nonetheless, introducing sociodemographic controls did diminish the 
differences in both 1982 and 1985 with respect to each of the activities 
mentioned above. In 1982, between 20 percent (for ballet) and 36 percent 
(for opera) of the African-American/white differences were attributable to 
sociodemographic variation between African-Americans and whites. In 
1985, similar proportions of the African-American/white differential were 
attributable to sociodemographic variation (from 18 percent for art exhibi- 
tions to 37 percent for musical theater), with the exception that 57 percent of 
the variation in opera attendance was of sociodemographic origin. In other 
words, except for opera attendance, less than one-half, and in most cases 
closer to one-fourth, of the differences between African-American and white 
probabilities of participation in these activities stem from differences in the 
sociodemographic characteristics of these groups. 

It is instructive to consider the core activities — jazz concert attendance, 
public performance on a musical instrument, and acting, singing, or dancing 
in public — to which this generalization does not apply. In both years, 
African-Americans were significantly more likely than whites to report 
attending live jazz concerts, and controlling for sociodemographic charac- 
teristics merely increased their advantage, although modestly. In both years, 
whites were slightly more likely to report performing on a musical instru- 
ment in public. In 1982, adding sociodemographic controls yielded only a 
trivial reduction in the small and statistically insignificant difference. In 
1985, the gross difference was modestly significant; whereas, with sociode- 
mographic controls, it was not significant at all. For acting, singing, and 
dancing, neither the gross nor net difference between African-Americans 
and whites was significant in either year. 

The pattern that emerges is one of significant differences between 
African-American and white participation in the consumption of most high- 
cultural arts activities, both in gross terms and with sociodemographic 
characteristics controlled. With respect to these activities, a substantial 



45 



Race, Ethnicity, and Participation in tlie Arts 

portion, but (with one exception) less than half, of the difference results from 
variation between African-American and white Americans in sociodemo- 
graphic factors. By contrast, the greater propensity of African-Americans to 
attend jazz concerts — the one activity with historical ties to the African- 
American community — is actually accentuated when sociodemographic 
differences are controlled. Gross differences between African-Americans 
and whites with respect to acting or singing (including popular or commer- 
cial as well as fme-arts forms) are slight; in the one case in which such a 
difference is modestly significant, it becomes insignificant when sociode- 
mographic factors are taken into account. 

This pattern reinforces our conviction that one cannot generalize about 
racial differences in artistic participation, per se. We suspect that if more art 
forms with origins in the African-American experience had been included 
among the core activities, the results would reveal, as was the case for jazz, 
statistical underrepresentation of white Americans. 

It is with respect to attendance at live, noncommercial, high-cultural 
events, as well as attendance at musical theater and reading imaginative 
literature, that whites participate at significantly higher rates than African- 
Americans, even controlling for demographic differences between the two 
groups. In other words, African-Americans are less likely than whites of 
similar socioeconomic standing to engage in the public consumption of 
Euro- American "high culture" and related genres. 

Although these interracial differences are persistent and appear with 
respect to most of the forms of participation examined here, they are not 
large in magnitude relative to differences associated with other determinants 
of participation.^ With respect to all of the activities for which being Afri- 
can-American significantly lessens participation (relative to whites), the 
direct effect of race is dwarfed by the impact of educational attainment and 
(except for reading in 1982) exceeded by the effect of family income. 
Similarly, once other sociodemographic factors are taken into account, 
participation rates of African-Americans and whites are more similar than 
are rates for men and women for all such activities but visiting art exhibi- 
tions. They are also more similar than rates for inner-city dwellers and 
persons living outside of SMS As for all such activities but attending classical 
music performances and reading imaginative literature. Thus race is a far less 
important net predictor of participation in all activities in which African- 
Americans participate significantly less than whites than educational attain- 
ment. In most cases, race is also a weaker predictor than income, gender, or 
urban residence. Note, however, that African-Americans earn less money 
and have historically received fewer years of formal education than compa- 



46 



Net Effects of Race and Ethnicity 

rable whites.^ Therefore, in addition to its direct negative effect, being 
African-American exerts a small indirect negative effect on probability of 
participation in the core activities because African-Americans have, on 
average, lower incomes and fewer years of formal education than whites. 

Variation between whites and Hispanics. The results for Americans of 
Hispanic origin lend themselves less easily to generalization. For one thing, 
no single pattern characterized Hispanic participation in the core consump- 
tion activities. For another, the influence of Hispanic origin on participation 
in specific activities varied from year to year. Although the latter differences 
were not statistically significant, the relatively small size of the Hispanic 
subsamples and, more important, the fact that the survey was not designed 
to represent statistically the Hispanic population, make the differences 
between the 1982 and 1985 results difficult to interpret. 

In 1982, Hispanics were significantly less likely than whites to report 
reading novels and other imaginative works, attending classical music con- 
certs, art exhibits, plays, musical theater performances; or acting, singing, or 
dancing on stage. In 1985, they were significantly less likely than whites to 
report every activity but performing on an instrument in public. 

For most of these activities, however, large portions of the Hispanic/ 
white difference stem from differences in the sociodemographic composi- 
tion of the two groups. Entering sociodemographic controls into the 1982 
models, for example, eliminates 89 percent of the differential between 
whites and Hispanics in classical music attendance, 91 percent of the vari- 
ation in attending art exhibits, 57 percent in attending musical theater 
performances, 48 percent in attending stage plays and 39 percent in reading 
imaginative literature. Indeed, after controlling for these characteristics, 
rates of Hispanic participation are significantly lower than those of whites 
for no activities but attending musical theater performances and plays, and 
reading imaginative Hterature (in both years); performing on stage (in 1982); 
and attending opera (in 1985). In other words, these analyses demonstrate 
that Americans of Hispanic origin are about as likely as white Americans 
with similar sociodemographic characteristics to attend ballet, classical 
music, and jazz performances, to visit art exhibitions, and to perform on a 
musical instrument. Indeed, in 1982, Hispanic respondents were signifi- 
cantly more likely to attend ballet performances than sociodemographically 
comparable whites. 

Two differences are notable between patterns for Hispanic and African- 
American respondents. First, although participation rates are roughly com- 
parable for these two groups for most activities, larger proportions of the 



47 



Race, Ethnicity, and Participation in the Arts 

differences between Hispanics and whites than between African-Americans 
and whites stem from intergroup differences in sociodemographic attributes. 
By contrast, more of the differences between African-Americans and whites 
reflect differences between the races in tastes, access, or unmeasured char- 
acteristics not associated with the sociodemographic controls. What this 
means is that public policies or historical processes that made Hispanics 
more similar to whites with respect to such resources as educational attain- 
ment, occupation, or earnings would, as a by-product, minimize many 
differences in artistic participation as well. So would policies that increased 
the artistic participation of people with fewer educational, occupational, and 
financial resources, even if those policies were not directed specifically at 
Hispanic Americans. By contrast, even if African-Americans became more 
similar to white Americans in their sociodemographic characteristics and 
even if the link between such characteristics and participation were lessened, 
African-Americans could still be expected to participate slightly but signifi- 
cantly less than whites in several of the core activities. 

Hispanics and African-Americans also differ with respect to the activi- 
ties for which these generalizations do not hold. As we have seen, the 
statistical overrepresentation of whites relative to African-Americans ap- 
plies only to predominantly Euro- American consumption activities and not 
to jazz or public performance. By contrast, Hispanic Americans participated 
at lower rates than whites in both years, net sociodemographic differences, 
only in reading and in attendance at musical and dramatic theatrical perform- 
ances. Note that these three activities are the only core activities for which 
command of the English language is ordinarily essential. Whereas almost all 
African-Americans and white Americans are native English speakers, a 
substantial proportion of Hispanic Americans are not. Thus we surmise 
(although, lacking data on language we cannot be sure) that lower net rates 
of Hispanic participation in activities involving the printed and spoken word 
reflect the linguistic characteristics of the Hispanic population and the 
relatively low availability of performances and imaginative literature in 
Spanish. Were the availability of such materials increased, we would expect 
to see Hispanics participate in them at rates comparable to those of whites 
with similar sociodemographic attributes. 

Variation between Asian and white Americans. The 1985 SPPA (unlike its 
1982 counterpart) made it possible to distinguish between Asian- American 
and other respondents. Nonetheless, because there were so few Asian respon- 
dents (well under 2 percent of the total sample), we cannot report their 
behavior with much statistical confidence. In models with only race and 



48 



Net Effects of Race and Ethnicity 

ethnicity included, Asians were more likely than whites to participate in all 
the activities that do not rely on the spoken or printed word except attending 
jazz concerts, and less likely than whites to participate in those that do. 
Nonetheless, these differences were small and never statistically significant. 
Because Asian-Americans tend on average to have sociodemographic 
characteristics that are associated with participation in the core activities, 
entering sociodemographic controls actually decreased net Asian participa- 
tion relative to white participation. Thus Asian-Americans were signifi- 
cantly less likely than whites in similar sociodemographic circumstances to 
attend musical or dramatic stage presentations. Differences with respect to 
other core activities remained statistically insignificant. 



Differences in Predictors of Participation by 

Race/Ethnicity 

Built into the preceding analyses is the assumption that the same socio- 
demographic factors influence the participation of African-Americans, His- 
panics, and whites in the same ways and to the same extent. This is a useful 
simplifying assumption because it enables us to estimate net differences in 
participation. But if we are interested in understanding the factors that lead 
members of racial and ethnic minorities to participate in the arts activities 
about which the surveys asked, we must consider the possibility that differ- 
ent groups arrive at participation by different routes. After much analysis, 
we conclude that the sociodemographic predictors of artistic participation 
(as defined by the core variables) are not systematically different for Afri- 
can-Americans, Hispanics, and whites.^ 

Differences in predictors of participation for whites and African-Americans. 
For both African-Americans and whites, educational attainment tended to be 
the variable that most effectively distinguished participants from nonpartici- 
pants for most of the core participation measures. Each of the other inde- 
pendent variables was significant, although less so than education, in 
predicting most of the participation measures for whites. For African- Ameri- 
cans, SMS A residence, income, and, in 1982, occupation, were also signifi- 
cantly related to many core variables. Although they were less likely to be 
significant for African-Americans than for whites (in part because signifi- 
cance is a function of the number of cases and there were many more whites 
than African-Americans among the respondents to the survey), most predic- 



49 



Race, Ethnicity, and Participation in tlie Arts 

Table 3-2 

Significant Differences in Models Predicting 

Responses to Core Participation Questions for African-American, 

Hispanic, and White Subsamples 



ACTIVITY 



PREDICTOR 



Jazz WOMEN 

MARITAL 
EDUCATION 

Classical music EDUCATION 
OCCUPATION 



Opera 



SMSA 
OCCUPATION 



Musical theater WOMEN 



Plays 



Ballet 



EDUCATION 
WOMEN 

SMSA 
SMSA 



Significantly negative for African-Americans {1982 and 
1985). Not significant for whites in 1982 and signifi- 
cantly positive in 1985. 

Significantly positive for whites but not for African- 
Americans or Hispanics in 1982. 

Significantly positive for whites but not for Hispanics in 
1982. 

More significantly positive for whites than for Hispanics 
in 1982. 

Significantly positive for whites but not for Hispanics in 
1985. 

Extremely positive for African-Americans and Hispanics 
but not for whites in 1985. 
Extremely positive for Hispanics but not for African- 
Americans or whites in 1985. 

Significantly positive for whites but not for African- 
Americans in 1982. 

More significantly positive for Hispanics than for whites 
in 1985. 

Significantly positive for whites but not for African- 
Americans or Hispanics in 1982. 

More significantly positive for African-Americans than 
for whites in 1982. 

Extremely significantly positive for Hispanics but less so 
for whites and not significant for African-Americans in 
1982. Extremely significant (positive) for African- 
Americans but not for whites or Hispanics in 1 985. 



50 



Net Effects of Race and Ethnicity 

Table 3-2 (Continued) 

Significant Differences in Models Predicting Responses 

to Core Participation Questions for African-American, 

Hispanic, and White Subsamples 



ACTiViTY 


PREDICTOR 




Art 
museums 


WOMEN 


Significantly positive for whites but not for African- 
Americans in 1982. 




SMSA 


Significantly positive for African-Americans but less so 
for whites and not significant for Hispanics in 1 982. 




EDUCATION 


More significantly positive for whites than for Hispanics 
in 1985. 


Perform: 
instrument 


SMSA 


Extremely significantly positive for Hispanics but not for 
African-Americans or whites in 1982. 


Perform: 
act, sing, 
dance 


SMSA 


Extremely significantly positive for Hispanics but not for 
African-Americans or whites in 1982 and 1985. (Signifi- 
cantly negative for whites in 1985.) 




EDUCATION 


Significantly positive for whites but not for Hispanics in 
1982. 


Literature 
reading 


WOMEN 


More significantly positive for whites than for African- 
Americans and Hispanics in 1982 and African-Ameri- 
cans in 1985. 




SMSA 


More significantly positive for African-Americans than 
forwhitesin1985. 




AGE 


Significantly negative for African-Americans but not for 
whites in 1982. 




EDUCATION 


More significantly positive for whites than for African- 
Americans in 1982. 




OCCUPATION 


More significantly positive for African-Americans than 



forwhitesin1982. 

Note: Descriptive statements provided only for differences that are statistically significant. Similar 
differences that do not reach statistical significance are not noted in this table. 



51 



Race, Ethnicity, and Participation in the Arts 

tors took the same sign and, in many cases, were of the same order of 

magnitude for African- Americans and whites. 

In only a few cases were there significant differences in models predict- 
ing core participation activities for the two races. In 1982, most such 
differences reflected an apparently stronger sexual division of labor in the 
consumption of the arts among whites than among African- Americans. That 
year, white women were significantly more likely than white men to report 
having participated in all of the core activities except playing a musical 
instrument in public and attending jazz concerts.^ By contrast, African- 
American women were significantly more likely than African- American 
men only to read works of imaginative hterature and attend ballet perform- 
ances. Differences between African-Americans and whites in the impact of 
gender were statistically significant in 1982 with respect to attending jazz 
concerts, attending musicals, attending plays, visiting art museums and art 
galleries, and reading literature. In each of the first four cases, white women 
were more likely to engage in the activity than white men, but African- 
American men were more Hkely to do so than African-American women. 
Women of both races were more likely to read literature than men, but the 
differences were significantly greater for whites. 

For 1982, these differences were notable and persistent across different 
kinds of artistic participation. But in 1985, the effects of gender varied less 
markedly by race, except for attending jazz concerts and reading hterature, 
for which the gap widened. With respect to attending musicals and plays and 
visiting art exhibitions, however, African-American women joined white 
women in being more likely than men to participate, and the effects of gender 
became similar for the two groups. In summary, the pattern for the two years 
indicates that there may be more marked gender differences in artistic 
participation among whites than among African- Americans; but, except for 
reading Hterature and attending jazz concerts, the differences are not large or 
persistent. Nonetheless, the interaction of gender and race deserves further 
investigation. 

For some activities in each year, the positive impact of living in an 
SMS A was greater for African- Americans than for whites. In 1982, although 
both African-Americans and whites were significantly more hkely to attend 
plays and visit art galleries and museums if they hved in SMSAs, the 
advantage of SMSA dwellers was significantly greater for African-Ameri- 
cans. In 1985, the same was true for reading hterature. Also in 1985, SMSA 
residence was not significantly related to attending operas or ballet perform- 
ances for whites, but was overwhelmingly so for African- Americans. That 
year, all 20 African -Americans who reported going to opera performances. 



52 



Net Effects of Race and Ethnicity 

and 32 of 33 who attended ballet performances, lived in SMSAs. These 
differences between African- Americans and whites are notable, but because 
they were discernible for no activity in both 1982 and 1985, they must be 
regarded with caution. 

In 1982, although single and divorced whites and African-Americans 
were more likely to attend jazz performances than others, the advantage of 
single and divorced people was significantly greater for whites than for 
African- Americans. Although the pattern held in 1985, the difference was 
not significant. 

In 1982, although education and occupation were both very significantly 
associated with reading works of imaginative literature among both African- 
Americans and whites, the effects of education were significantly greater for 
whites and the effects of occupation were stronger for African-Americans. 
Moreover, whereas age was positively, but not quite significantly, associated 
with reading for whites, it was negatively and significantly related to reading 
for African- Americans. In 1985, these patterns held but none of the differ- 
ences were significant, although the difference for age was nearly so. The 
pattern suggests, but does not confirm, the hypothesis that there may be 
increasing interest in reading literature among younger cohorts of African- 
Americans that cannot be explained solely by reference to increases in 
African-American educational attainment. 

Taken together, the separate models for African-Americans and whites 
suggest that the same sociodemographic characteristics are related to most 
of the core participation measures in approximately the same way for 
members of each group. There is a greater tendency for white women to 
outparticipate white men than for African-American women to outpartici- 
pate African- American men, and a stronger tendency for residence outside 
an SMSA to depress African-American attendance at arts events more than 
it depresses white attendance. Few specific differences, however, were 
significant in both 1982 and 1985, leading us to offer these observations as 
hypotheses for further study rather than as conclusions. 

Differences in predictors of participation for whites and Hispanics, As was 
the case for African-Americans and whites, education was by far the stron- 
gest predictor of participation in most of the core activities for Hispanics in 
both 1982 and 1985. As was the case for African- Americans, most of the 
predictors took the same sign and many were of roughly the same magnitude 
for Hispanics as for whites, but, because of the far smaller number of 
Hispanic than of non-Hispanic white respondents, fewer were statistically 
significant. Given these relatively few, relatively weak, and very inconsistent 



53 



Race, Ethnicity, and Participation in the Arts 

results, we can only conclude that the factors accounting for participation in 
the core activities were similar for whites and Hispanics. 

Differences in predictors of participation for African-Americans and His- 
panics. There were no statistically significant differences in the models 
predicting participation in the core activities for African-Americans and 
Hispanics. 

Differences in predictors of participation for each group between 1982 and 
1985. Perhaps unsurprisingly, given the brief time between the two surveys, 
no statistically significant differences in predictors for African-Americans 
and Hispanics were observed. For whites, residence in an SMSA had a 
significantly stronger positive impact on visiting art galleries and museums, 
and a significantly stronger negative impact on playing an instrument in 
public in 1985 than in 1982. The positive impact of educational attainment 
on reading imaginative literature was slightly, but significantly, weaker in 
1985 than in 1982. Given the large size of the white subsample and the large 
number of coefficients, we place little stock in these differences. 

Summary. It was important to test whether the models for African-Ameri- 
cans, Hispanics, and whites indicated that the factors influencing participa- 
tion for these groups differed. If they had, such evidence might have 
suggested, first, that the social meaning of participation differed, on average, 
for members of these groups and, second, that public policies or sociodemo- 
graphic change would influence African-American, Hispanic, and white 
participation in systematically different ways. Moreover, if the differences 
were substantial, they might lead us to question our interpretations of the 
aggregated models described in the first part of this chapter. 

The findings of these analyses provide no compelling evidence of sys- 
tematic differences in factors leading African-Americans, Hispanics, and 
whites to participate in the core activities about which the SPPA asked. 
Significant differences were few, usually small in magnitude, and rarely 
persisted from one year to the other. It is possible that more differences 
would have been found had the selection of activities about which respon- 
dents were asked been broader. Note, however, that the variation present 
among the different core activities was sufficient to permit us to note 
systematic patterns of racial differences in rates of participation, whereas no 
such systematic differences were observed with respect to the predictors of 
participadon. It is also possible that a different set of predictor variables 
might have revealed significant differences not noted here. It is not obvious 



54 



Net Effects of Race and Ethnicity 

to us, however, what such additional predictors might be. Finally, were the 
African-American and Hispanic sample sizes larger, it is likely that more 
differences would have emerged as statistically significant. We believe there 
are important reasons to include more African-American and Hispanic (and 
Asian and Native American) respondents in the SPPA. But we regard the 
African-American and Hispanic sample sizes as adequate for detecting 
substantively meaningful differences in models predicting core participation 
items. In short, these analyses convince us that the sociodemographic char- 
acteristics accounting for most kinds of artistic participation are basically 
similar for African- American, Hispanic, and white Americans. In Chapter 5, 
we construct more detailed models predicting several additional dimensions 
of artistic participation. 



55 



Chapter 4 

RACIAL/ETHNIC DIFFERENCES IN 
UNSATED DEMAND FOR 
PARTICIPATION 



n chapter 2, we noted persistent differences in rates at which Asian, 
African- American, Hispanics, and white Americans participate in the core 
activities about which the SPPAs asked all respondents. In chapter 3, we 
asked to what extent these differences could be accounted for by sociodemo- 
graphic attributes with respect to which Asians, African- Americans, Hispan- 
ics, and whites also differ. In this chapter, we focus on the extent to which 
such differences reflect intergroup differences in demand for the arts as 
opposed to differential exposure to barriers to participation. We consider this 
question with respect to the seven core activities that involve attendance at 
arts events.' 

Approximately one-fourth of the 1982 respondents and one-sixth of the 
1985 respondents were shown a card listing the activities and told: 

"Few people can do everything they would like to do. But if you 
could do any of the things listed on this card as often as you wanted, 
which ones would you do more often than you have during the last 
12 months?" 

Those respondents who said they would like to have done a given 
activity more than they had in the past year were then asked to indicate which 
of several reasons were responsible for the fact that they had not participated 
more. 

Demand for Greater Participation 

Members of a group may participate in a given activity at a lower rate 
than members of another group for either of two reasons. They may do so 
because they enjoy or otherwise value the activity less. Or they may want to 



56 



Racial/Ethnic Differences in Unsated Demand 



engage in the activity as much as do members of the other group, but face 
obstacles to participation that the others do not. 

These two explanations have very different implications for public 
policy. If low participation results not from low demand but from barriers to 
participation, policy might equalize participation by eliminating the barriers. 
If low participation results not from barriers but from low demand, poHcies 
aimed at eliminating inequality must serve to increase demand and not 
simply to level barriers. 

We hesitate to interpret people's responses to questions about their 
desire for increased participation, for we are not sure what people mean 
when they say they "want" to attend arts events more than they do. Some 
people may deeply desire to attend more, but be unable to do so for well 
defined reasons. Others may wish to attend more, but lack the willingness 
to pay the cost in foregone opportunities to do things they value even more 
highly. Still others may mean that they wish they were the kind of person 
who liked the arts more than they do. We do not believe that everyone who 
reported wanting to participate in an activity more cared passionately about 
doing so. As long as the different meanings of "want" were not distributed 
by race and ethnicity in dramatically different ways, however, responses to 
this question may provide clues as to the extent that intergroup variation in 
attendance represents differences in demand or differences in opportunity. 
Nonetheless, without knowing more than we do about the subjective mean- 
ing of these responses, we are reluctant to regard them as any more than 
clues. 

Responses to the "want more" question are reported in Table 4-1 for each 
activity and for African -Americans, Hispanics, whites, and (in 1985) Asian- 
Americans. We assume that "wanting to do more" means something differ- 
ent for a person who already participates than it does for someone who does 
not. Consequently, we report results separately for attenders (respondents 
who engaged in the activity at least once during the previous 12 months) and 
nonattenders. Consistent with our focus on rates of participation (rather than 
levels of participation), we look more closely at the latter. 

Findings, It has been suggested that the arts are addictive. That is, whereas 
demand for most goods precedes and is sated by consumption, consumption 
of the arts is said to beget demand for more.^ If this is the case, it explains 
what John Robinson, in a report on the 1982 SPPA, has called the "more- 
more principle": the more activities in which respondents participate, the 
more likely they are to participate in still others.^ 

Our findings on unsated demand for the core attendance activities are 



57 



Race, Ethnicity, and Participation in tlie Arts 

Table 4-1 

Percentage of Attenders and Nonattenders Wanting to Do 

Each Activity More Than They Had in the Previous 12 {Months, 

by Race/Ethnicity 



Jazz 



Ciassicai 



Opera 



Musical 



Attenders 


1982 


1985 


1982 


1985 


1982 


1985 


1982 


1985 


Whites 


53.77 
394 


59.89 
168 


53.15 
568 


46.56 
279 


44.50 
111 


40.41 
48 


69.77 
882 


64.69 
380 


African- 
Americans 


67.74 
80 


50.12 
42 


42.79 
31 


41.62 
17 


26.95 
4 


25.18 
9 


50.03 
44 


56.85 
30 


Hispanics 


51.41 
28 


42.23 
10 


61.89 
17 


44.99 
8 . 


24.87 
4 


NA 



68.91 
34 


45.24 
11 


Asian- 
Americans 




NA 





42.41 
4 




0.00 
2 




30.61 
4 


Non- 
attenders 


1982 


1985 


1982 


1985 


1982 


1985 


1982 


1985 


Whites 


13.34 
3,995 


14.60 
1,643 


14.28 
3,824 


13.36 
1,534 


7.17 
4,278 


8.50 
1,762 


27.33 
3,510 


24.82 
1,430 


African- 
Americans 


24.08 
457 


22.30 
245 


10.34 
506 


7.82 
269 


4.88 
531 


3.42 
278 


16.35 
493 


14.87 
256 


Hispanics 


14.13 
273 


19.28 
138 


12.65 
284 


5.61 
140 


4.82 
297 


4.80 
148 


16.24 
267 


10.29 
137 


Asian- 
Americans 




6.77 
47 




5.29 
43 




1.98 
45 




14.01 
42 



consistent with Robinson's "more-more" dictum and with the addiction 
model of arts consumption. With only four exceptions (all cases in which 
only two or fewer respondents participated in the given activity), for every 
activity and every racial/ethnic group, attenders were more than twice as 
likely (and in most cases three or four times as likely) to want to participate 
more than were nonattenders. For example, in 1982, 54 percent of white jazz 
attenders, but just 13 percent of white nonattenders, reported wanting to 
attend jazz concerts more. That year, 43 percent of African- Americans who 
attended classical music concerts wanted to attend more, compared with just 



58 



Racial/Ethnic Differences in Unsated Demand 

TABLE 4-1 (Continued) 

Percentage of Attenders and Nonattenders Wanting to Do 

Each Activity More Than They Had in the Previous 12 Months, 

by Race/Ethnicity 

Plays Ballet Art 



Attenders 


1982 


1985 


1982 


1985 


1982 


1985 


Whites 


62.42 
575 


59.95 
245 


54.29 
182 


55.09 
80 


57.18 
1,057 


57.91 
475 


African- 
Americans 


55.38 
27 


48.06 
23 


31.83 
5 


44.33 
11 


50.48 
49 


57.26 
33 


Hispanics 


57.45 
18 


20.44 
4 


57.92 
9 


65.01 
3 


78.86 
52 


63.59 
28 


Asian- 
Americans 




NA 





0.00 
2 




43.23 
6 


Non- 
attenders 


1982 


1985 


1982 


1985 


1982 


1985 


Wiiites 


22.48 
3,816 


22.10 
1,565 


11.01 
4,211 


11.42 
1,733 


24.40 
3,334 


24.07 
1,337 


African- 
Americans 


9.45 
510 


8.31 
263 


6.20 
532 


5.74 
276 


17.80 
488 


20.20 
254 


Hispanics 


9.54 
283 


8.43 
144 


7.26 
292 


10.75 
145 


19.25 
249 


17.66 
120 


Asian- 
Americans 




6.43 
47 




6.04 
45 




11.60 
41 



Note: Percentages are weighted. Ns are unweighted. 



10 percent of African-American nonattenders. Almost 80 percent of Hispan- 
ics who visited art museums or galleries, but less than 20 percent of those 
who did not, wanted to go more often. 

The "more-more principle" also applied at the group level among non- 
attenders. That is, for each activity, except for Asian- Americans, nonattend- 
ing members of the racial or ethnic group that attended most were also more 
likely than members of other groups to want to attend. For example, more 



59 



Race, Ethnicity, and Participation in the Arts 

than 20 percent of African- Americans who did not attend jazz concerts 
wanted to do so in both survey years. By contrast, fewer than 15 percent of 
nonattending whites wished to attend. With respect to the other activities, 
which whites were more likely to attend, white nonattenders were more 
likely than other nonattenders to report wanting to participate. 

For some activities, the differences were small. For example, in 1982, 14 
percent of whites who had not attended classical music concerts wanted to 
do so, compared with 10 percent of such African-Americans and 13 percent 
of Hispanic nonattenders. In other cases, the differences were more sizable. 
In 1982, 22 percent of nonattending whites, but fewer than 10 percent of 
nonattending African-Americans and Hispanics wished to go to stage plays 
or "the theater." 

Asian-Americans were the exception to the more-more principle. Al- 
though they participated in most activities at rates either higher or only 
slightly lower than those of white Americans, the percentages reporting a 
desire to attend each activity more were lower than those for all or most other 
groups. Whatever the reason, the gap between self-reported aspiration and 
actual participation was smaller for Asian- American respondents than for 
members of other groups. 

The tendency of nonattenders from groups with relatively high rates of 
attendance to want to attend more than those from groups with lower 
attendance rates can be interpreted in either of two ways. To the extent that 
members of these groups tend to sociaHze disproportionately with others 
from those groups, nonattenders in groups with high attendance rates may 
come into more frequent contact with attenders than do members of other 
groups. On the one hand, this contact may engender a greater desire to try 
the activity in question. On the other, it may engender guilt about nonpar- 
ticipation, and consequently inflate what survey analysts refer to as "social- 
desirability bias" in their responses. The latter possibility is one more reason 
to interpret these data with caution. 

How would rates of participation change if everyone did what he or she 
wantedl Let us take the responses at face value and treat them as indicators 
of genuine unsated demand for the activities about which respondents were 
asked. If each nonattender who said that he or she wanted to participate were 
to do so, how would differences in participation by race and ethnicity be 
affected? 

The answers are presented in Table 4-2. For each activity, each survey 
year, and each racial/ethnic group, data are presented on the percentage 
reporting participation in the prior year; the percentage who did not partici- 



60 



Racial/Ethnic Differences in Unsated Demand 

Table 4-2 

Real Attendance Rates, Potential Increments, and Total Potential 

Attendance by Race and Ethnicity 



1982 



Jazz 



1985 





W 


A 


H 




W 


A 


H 


S 


Base 


9.13 


15.64 


8.27 




9.48 


13.08 


6.55 


7.81 


Increment 


12.30 


20.31 


12.13 




13.22 


19.38 


18.02 


6.24 


Potential 


21.43 


35.95 


20.40 




22.70 


32.46 


24.57 


14.05 






1982 


1 


Classical 


1 


1985 






W 


A 


H 




W 


A 


H 


8 


Base 


14.42 


6.67 


7.87 




14.31 


6.39 


6.77 


16.50 


Inaement 


122? 


9.65 


11.65 




11.45 


7.32 


5.23 


4.42 


Potential 


26.64 


16.32 


19.52 




25.76 


13.71 


12.00 


20.92 






1982 




Opera 




1985 






W 


A 


H 




W 


A 


H 


8 


Base 


3.33 


1.36 


2.52 




2.97 


1.43 


0.78 


4.58 


Increment 


6.93 


4.81 


4.70 




8.25 


3.37 


4.76 


1.89 


Potential 


10.26 


6.17 


7.22 




11.22 


4.80 


5.54 


5.47 






1982 




Musical 




1985 






W 


A 


H 




W 


A 


H 


8 


Base 


20.67 


10.10 


10.96 




18.60 


8.45 


9.52 


13.89 


Increment 


21.68 


14.70 


14.46 




20.20 


13.61 


9.31 


12.06 


Potential 


42.35 


24.80 


25.42 




38.80 


22.06 


18.53 


25.95 






1982 




Plays 




1985 






W 


A 


H 




W 


A 


H 


8 


Base 


13.44 


5.82 


5.47 




13.10 


6.09 


6.41 


8.87 


Increment 


19.46 


8.90 


9.02 




19.20 


7.80 


7.89 


5.86 


Potential 


32.90 


14.72 


14.49 




32.30 


13.89 


14.30 


14.73 






1982 




Ballet 




1985 






W 


A 


H 




W 


A 


H 


8 


Base 


4.64 


1.78 


4.54 




4.72 


2.14 


3.21 


6.22 


Increment 


10.50 


6.09 


6.93 




10.88 


5.62 


10.40 


5.66 


Potential 


15.14 


7.87 


11.47 




15.60 


7.76 


13.61 


11.88 






1982 




Art 




1985 






W 


A 


H 




W 


A 


H 


8 


Base 


23.94 


12.47 


16.22 




24.14 


10.71 


18.18 


26.02 


Increment 


18.56 


15.58 


16.12 




18.26 


18.04 


14.45 


8.58 


Potential 


42.50 


28.05 


32.34 




42.40 


28.75 


32.63 


34.60 



Note: Base rates from Table 2-1 . Increment=percentage of nonattenders who reported wanting to 
participate times complement of base. A=African-American; S=Asian- American; H=Hispanic-Ameri- 
can. 



61 



Race, Ethnicity, and Participation in the Arts 



pate but reported they wanted to; and the total "potential audience" compris- 
ing both groups. 

Potential participation rates, defined in this way, are much greater than 
the actual participation rates for all groups but Asian-Americans. Indeed, 
except for white attendance (in 1982 and 1985) and Hispanic attendance (in 
1985) at classical music concerts and white and Hispanic visits to art 
museums and galleries (in both years), potential rates are at least twice the 
actual rates of attendance. In many cases, the differences are much greater 
than that. In other words, fewer people participated in these activities than 
did not but said they would like to do so. 

Because nonattending members of groups with high participation rates 
are more likely to report wanting to attend than are nonattending members 
of other groups, the first effect of everyone doing what he or she reports 
wanting to would be to widen the absolute intergroup percentage difference 
in participation rates. In the case of jazz, the absolute difference between 
African-American participation rates and those of whites and Hispanics 
would double. In the case of the other six activities, the absolute difference 
between white rates and those of African-Americans and Hispanics would 
increase. (Again, Asian-Americans are the exception to the rule. Although 
their real participation rates in classical music concerts, opera, ballet, and art 
exhibits were higher than those for other groups, their potential participation 
rates were actually lower than those of whites for all of these and of 
Hispanics for opera attendance.) In other words, if all respondents did what 
they said they wanted to do, the absolute gap in participation rates between 
whites and everyone else would become wider. (The exceptional activity is 
jazz, for which the gap between African-Americans and everyone else would 
widen.) 

In chapter 2, we focused not on absolute differences in rates but on the 
ratio of the white rate to rates for other groups. In other words, we asked how 
much more likely whites were than African-Americans or Hispanics to 
participate in these activities. When we put the question this way, our results 
are mixed. For most activities — and in 1982 for all activities but jazz and 
ballet — the ratios of white to other potential attendance rates are lower than 
the ratios of white to other real attendance rates. For example, whites were 
more than twice as likely as African-Americans to attend musicals in 1982. 
If everyone who wanted had attended, they would have been only 1 .71 times 
as likely. Similarly, whites were nearly 50 percent more Hkely to visit art 
galleries or museums than Hispanic-Americans in 1982. If everyone who 
wanted had attended, their advantage would have declined to approximately 
30 percent. 



62 



Racial/Ethnic Differences in Unsated Demand 

With respect to several activities, however, ratios between white and 
other groups' potential participation rates are even higher than for real 
participation rates. This is true of the white/African- American ratios for 
opera attendance and theater-going in 1985 and of white/Hispanic ratios for 
opera and ballet attendance in 1982 and for classical music, musical theater, 
and straight theater attendance in 1985. For example, in 1985 white Ameri- 
cans were 2.04 times as likely as Hispanic-Americans to report theater 
attendance, whereas they were 2.26 times as likely to appear in the potential 
audience for stage plays. 

These results indicate that differences in participation in core activities 
do not result from barriers that disproportionately affect the ability of 
members of different groups from satisfying perceived demand. Instead they 
seem to reflect differences in the extent to which members of different racial 
and ethnic groups believe that they want to attend such arts events. For each 
of the seven activities about which they were asked (and with the notable 
exception of Asian-Americans), nonattenders of the groups whose members 
already participated at the highest rates were more likely than others to want 
to become participants. 

If African- American and Hispanic nonattenders had wanted to participate 
more than white nonattenders, this would have constituted strong evidence 
that intergroup differences reflected barriers to minority attendance and not 
differences in demand. Clearly these data do not point in that direction. It 
would be simplistic, however, to take these results as strong evidence that 
intergroup differences do not reflect differences in opportunity. 

First, the most effective barriers to participation may be those that 
influence demand, not those that influence the ability of persons to satisfy 
demand they already have. If, as the addiction theory mentioned at the 
beginning of this section suggests, taste for the arts is acquired through 
participation in the arts, then any barriers that prevent persons from partici- 
pating in the arts are likely to be reflected in lower demand from the persons 
excluded."^ 

Second, respondents to the SPPA "want-more" questions may have 
answered on the basis of pre-conscious understandings about the costs 
associated with getting more of what they wanted. If there are perceived 
higher costs to participation for minorities than for whites, differences in 
demand may reflect these costs. 

Third, it is possible, for the reasons discussed above, that social-desir- 
ability bias inflated the "want-more" responses of whites relative to those of 
African-Americans and Hispanics for those activities in which white Ameri- 
cans have the highest rates of participation. 



63 



Race, Ethnicity, and Participation in tlie Arts 

These are all hypotheses that should lead us to avoid hasty conclusions 
on the basis of these findings, but should not lead us to dismiss them either. 
The results of these analyses indicate that it would be simple-minded to think 
about intergroup differences in arts attendance in the same way we think 
about intergroup differences in education. In the case of education, we have 
much evidence that demand is similar among racial and ethnic groups, with 
everyone viewing education as a good thing that helps people get ahead.^ In 
the case of the arts, the evidence presented here indicates that attendance at 
the live events about which people were asked is not desired equally by 
members of all groups. The evidence indicates that demand varies by group 
and that if there are barriers, they work in large part by influencing demand 
for Hve attendance. 



Why People Don't Attend More 

Respondents who said they wanted to attend one of the seven core 
activities were given a list of possible reasons for not attending more than 
they did and were asked to check all those that applied. The reasons among 
which respondents could choose included: Tickets sold out. Cost, Not 
available. Feel uncomfortable. Don't have anyone to go with. Babysitter 
problems/Must care for children. Problem related to a handicap. Problem 
related to age/health, Too far to go, Transportation/Traffic/Parking prob- 
lems. Crime or fear of crime, Poor quality /Not very good. Prefer to watch 
TV, Don't have time, Procrastination/Lack of motivation, and Other.^ To 
simplify the analyses we coded together "Problem related to a handicap" and 
"Problem related to age/health." Similarly, we coded together "Procrastina- 
tion/Lack of motivation" and "Prefer to watch TV" because we regarded 
each of these as indicating exceptionally low levels of demand, so low as to 
suggest some inconsistency with the respondent's professed desire to attend 
more. 

We present the results in two forms. Appendix Tables 4-1 through 4-7 
list, for African- Americans, Hispanics, and whites for 1982 and 1985, the 
weighted percentage of "want-more" attenders and "want-more" nonattend- 
ers in each group in each year giving each reason, along with the unweighted 
numbers of respondents upon which results for each group are based.^ Table 
4-3 summarizes the information for nonattenders who reported wanting to 
attend — the group of most immediate interest here — by listing for whites, 
African-Americans, and Hispanics in each year the three reasons given by 
the largest numbers of respondents and the percentages (of the nonattenders 



64 



Racial/Ethnic Differences in Unsated Demand 



who wished to attend) giving each response. (In the two cases in which more 
than 20 percent of such respondents marked a fourth reason, that one is 
reported as well.) 

These findings are suggesiive at best. The results for African- Americans 
and Hispanic-Americans are based on very small numbers of respondents 
because these questions were asked during only some of the survey months 
and because relatively few nonattenders wished to become attenders.^ More- 
over, we find responses to these questions difficult to interpret. We can be 
reasonably certain that some of the reasons provided were hastily selected 
excuses offered under duress by respondents who may have expressed a 
casual wish to do something they had not done. We are certain that others 
reflect real barriers to attendance. There is no obvious way to tell the two 
apart. For example, some people who said they did not attend stage plays 
because they were given at sites too far away may have made no effort to 
find out whether plays were presented nearby. Others may have been 
suburbanites who think nothing of going downtown to visit a museum, but 
value stage plays less than other forms of recreation. Still others may live in 
rural areas of prairie states where the nearest theater is three hours away. 
Some respondents who gave "cost" or "don't have enough time" as reasons 
may be destitute or work 70-hour weeks to support large families: that is, 
they may be people with little or no discretionary money or time. Others may 
have more discretionary income or time, but choose to spend it on other 
things. For the latter, "cost" or "time" responses tell us not just about barriers 
but about the value that respondents place on the arts relative to other uses 
of their time and money. Without either direct or plausible indirect measures 
of value that respondents place on attendance at the core activities, responses 
to the barrier questions are virtually uninterpretable. 

Nonetheless, if we assume that the underlying valuation of arts atten- 
dance is the same for all three groups and if we remember to treat the data 
as merely suggestive, the results are interesting.^ For members of all groups, 
cost and lack of time were the most important reasons given for nonpartici- 
pation. With respect to most activities, white respondents were more likely 
to give time as a reason than cost, and Hispanic respondents were more likely 
to cite cost than time. In 1982, African-American respondents were some- 
what more likely to mention cost than time for most activities, whereas in 
1985 they were somewhat more hkely to cite time than cost. Lack of 
availability was frequently cited by whites, and a similar reason, that events 
were too far away, was often mentioned by Hispanics. African-American 
respondents frequently mentioned these and also cited transportation prob- 
lems as impediments to attendance more than whites and, for most activities. 



65 



Race, Ethnicity, and Participation in the Arts 

Table 4-3 

Leading Reasons Given for Nonattendance 

by Nonattenders Who Wished to Attend 

Jazz, 1982 

W Time (41 ), Cost (26), Not available (22) 
A Cost (45), Time {24), Transportation (14) 
H Cost (40), Time (37), Not available (14) 

Jazz, 1985 

W Time (45), Cost (29), Not available (23) 
A Time (41), Cost (39), Not available (13) 
H Cost (55), Lack motivation (31 ), Time (31 ), Child care (21 ) 

Classical, 1982 

W Time (39), Cost (28), Not available (23) 
A Cost (44), Time (35), Transportation (21 ) 
H Cost (48), Time (33), Too far to go (20) 

Classical, 1985 

W Time (35), Cost (30), Too far to go (25), Not available (24) 
A Time (48), Cost (24), Transportation (1 7) 
H Insufficient number of respondents 

Opera, 1982 

W Cost (35), Time (30), Not available (26) 
A Cost (39), Time (30), Too far to go (12) 
H Cost (68), Too far to go (36), Time (1 5) 

Opera, 1985 

W Cost (37), Time (33), Too far to go (26) 

A Time (61 ), Transportation (30), Too far to go (14) 

H Insufficient number of respondents 

Musical theater, 1982 

W Time (37), Cost (31 ), Not available (21 ) 
A Cost (47), Time (29), Lack motivation (12) 
H Cost (37), Time (33), Too far to go (29) 

Musical theater, 1985 

W Time (34), Cost (32), Too far to go (1 9) 
A Cost (43), Time (26), Too far to go (1 5) 
H Cost (53), Time (37), Child care (1 7) 



66 



Racial/Ethnic Differences in Unsated Demand 

Table 4-3 (Continued) 

Leading Reasons Given for Nonattendance 

by Nonattenders Who Wished to Attend 

Plays, 1982 

W Time (39), Cost (31), Too far to go (15) 
A Cost (24), Not available (20), Time (15) 
H Cost (44), Time (41), Lack motivation (12) 

Plays, 1985 

W Time (39), Cost (25), Not available (21 ) 
A Time (39), Cost (38), Transportation (14) 
H Cost (60), Time (51 ), Lack motivation (25) 

Ballet, 1982 

W Time (32), Cost (29), Not available (27) 
A Cost (43), Time (33), Not available (14) 
H Cost (46), Time (26), Too far to go (20) 

Ballet, 1985 

W Time (35), Cost (33), Too far to go (22) 
A Time (51), Cost (37), Fear of crime (12) 
H Cost (44), Time (28), Too far to go (1 6) 

Art museums and galleries, 1982 

W Time (40), Not available (25), Too far to go (20) 
A Time (31), Cost (23), Lack motivation (18) 
H Time (47), Child care (1 5), Transportation (13) 

Art museums and galleries, 1985 

W Time (39), Not available (24), Too far to go (21 ) 
A Time (53), Transportation (19), Cost (17) 
H Time (74), Lack motivation (34), Cost (30) 



Note: Figures in parentheses are weighted percentages of those nonattending respondents who 
wanted to attend who reported a given reason for not attending. Data summarized from Appendix 
Tables 4-1 through 4-7. 



67 



Race, Ethnicity, and Participation in the Arts 

more than Hispanics. For most activities, Hispanics were more likely than 
African-Americans or whites to cite child care problems as reasons for not 
attending. Fear of crime, handicap or health problems, poor quality, public- 
ity, work related reasons, or performance time did not loom large as reasons 
for many respondents in any group. 

In other words, whites tended to cite reasons indicative of an inadequate 
supply of activities more than members of other groups. By contrast, Afri- 
can-Americans and Hispanics were more likely than whites to mention 
problems like cost, transportation, and child care that are associated with 
insufficient financial resources. It follows from this that programs aimed at 
improving geographical access to the arts may disproportionately aid white 
Americans, whereas programs focused on financial access may be more 
likely to assist African-Americans and Hispanics. 

At the same time, however, most of these differences were either rela- 
tively weak or somewhat inconsistent from activity to activity or year to 
year. Overall, the reasons given by African-Americans, Hispanics, and 
whites who did not attend the core activities but would Hke to do so, were 
rather similar, and focused on cost, time, and availability. 



Conclusions 

Demand for participation in the seven core attendance activities appears 
to be cultivated by attendance. People who already attend are much more 
likely to want to attend more than are people who do not. Thus, although 
there is much apparent unsated demand for these activities, most of it comes 
from attenders rather than nonattenders. Because, with the exception of jazz 
performances, whites are more likely to attend than are African-Americans 
or Hispanics, unsated demand appears to be greater among whites than 
among members of these other two groups. 

If we look only at nonattenders, members of groups with higher atten- 
dance rates (African-Americans for jazz performance, whites for the other 
attendance activities) are more likely than others to say they want to attend. 
If we take professed desire for attendance at face value, then if all barriers to 
attendance were removed, the absolute differences in percentage participa- 
tion rates between the groups that participate most and those that participate 
least would increase. Because intergroup differences in desired participation 
among nonattenders are less, in most but not all cases, than are intergroup 
differences in actual participation, the ratios of white attendance to African- 
American and Hispanic attendance, respectively, would decline somewhat 



68 



Racial/Ethnic Differences in Unsated Demand 



for most, but not all, activities if everyone did what he or she said they 
wanted to do. 

Data on people's reasons for not attending are difficult to interpret and 
the numbers of African-American and Hispanic respondents are small. This 
weak evidence suggests that white, African-American, and Hispanic would- 
be attenders are all deterred most frequently by cost, lack of time, and limited 
availability. At the same time, whites are somewhat more likely to mention 
reasons related to limited availability than are members of other groups, 
whereas African-American and Hispanic respondents are more likely to 
mention reasons related to cost. Because, except for jazz, white nonattenders 
were more likely to report wanting to attend the events about which they 
were asked than were African-Americans or Hispanic nonattenders, and 
because most intergroup differences were relatively small or inconsistent, 
the evidence does not indicate that eliminating income-related barriers 
would quickly or markedly erode intergroup differences in participation. 

These findings may seem inconsistent with some of the results presented 
in earlier chapters. For example, we noted earlier that the differences in rates 
of participation between whites, on the one hand, and Hispanics and Afri- 
can-Americans, on the other, were less for watching the arts on television 
than for live attendance. This led us to suggest that African-Americans and 
Hispanics might be deterred from live attendance at the core activities (other 
than jazz) by something other than taste. Yet the "want-more" questions 
failed to uncover greater unsated demand for live attendance (except for 
jazz) among these groups than among white Americans. 

One reason for this may be that live attendance at an arts event requires 
a greater degree of commitment than watching a similar event on television. 
A second reason may be that demand for live attendance is influenced more 
by attributes of the attendance experience than by attributes of a program 
itself. A third reason is that people may consciously or unconsciously take 
account of barriers that raise their cost of attendance in responding to 
questions about unsated demand. The SPPA data do not permit us to 
determine which, if any, of these explanations is correct. 

Our results may also seem at odds with the logistic regression analyses 
that showed that the difference in participation rates between Hispanic- 
Americans and (non-Hispanic) white Americans was reduced to insignifi- 
cance when differences among groups in sociodemographic factors were 
taken into account. If this were the case, would we not expect to see high 
levels of unsated demand, explained by economic barriers, among Hispanic 
Americans? 

Not necessarily. Our regression analyses indicated only that Hispanic- 



69 



Race, Ethnicity, and Participation in tlie Arts 

Americans were similar in their participation in the core activities to white 
Americans with similar sociodemographic characteristics. It seems likely, on 
the basis of the data we have analyzed, that sociodemographic barriers work 
not just by making it more difficult for people who want to participate to do 
so, but also by influencing the extent to which people want to participate. 

Our conclusions in this chapter have been general and laced with quali- 
fications. We have relatively little confidence in the usefulness of the SPPA 
questions on the extent of and reasons for unsated demand for understanding 
intergroup differences in participation. Some of our reservations have to do 
with the small number of African-American, Hispanic, and Asian-American 
respondents upon which our analyses, especially of reasons for nonatten- 
dance, are based. We hope that future SPPAs will oversample African- 
American, Hispanic, Asian, and Native American respondents so that more 
detailed analysis will be possible. 

Most of our reservations, however, have to do with the questions them- 
selves, which seem to us to embody an unsophisticated view of human 
motivation. Although responses to these questions may be applicable to 
short-term marketing issues, we suspect that they tell us little about the 
complex processes that culminate in demand for attendance at arts events or 
about the long-term potential for increases in participation. 



70 



Chapter 5 

EVIDENCE ON RACIAL AND 
ETHNIC DIFFERENCES IN 
PARTICIPATION FROM THE 
NOVEMBER/DECEMBER 1982 
SUBSAMPLE 



M 



ost of the analyses reported in Chapters 2 and 3 drew on data from 
all respondents to the 1982 and 1985 SPPAs. Because there were 
so many respondents, these analyses were statistically powerful, 
and permitted confident generalization. 

At the same time, because most of the SPPA questions were asked only 
in certain months, reliance on these prevented us from exploring relation- 
ships among answers to the full range of questions the surveys included. In 
this chapter, we take advantage of the survey's breadth by using data 
collected in November and December 1982. In these months alone, respon- 
dents were asked all of the questions that appeared on the SPPA survey. 

There are two advantages to focusing on this subsample. First, we can 
go beyond the core items to examine participation in a broader range of 
artistic activities. We have already noted that intergroup differences vary for 
different kinds of arts participation. In this chapter we investigate such 
differences more thoroughly. 

Second, the November/December 1982 subsample permits us to explore 
the combined effects on participation of a broader range of explanatory 
variables by including them in the same models. In addition to the sociode- 
mographic factors investigated in Chapter 3, in this chapter we consider the 
influence on participation of childhood experience (including both parental 
guidance and classes or lessons), musical taste, and viewing arts programs 
on television. 

These advantages bear a cost: the decline of statistical power associated 



71 



Race, Ethnicity, and Participation in the Arts 

with the number of respondents being reduced from more than 15,000 to 
2,255. In particular, some of the following results are based on small 
numbers of African-American or Hispanic respondents. Thus, the effects of 
race or ethnicity must be larger than in analyses reported in earlier chapters 
if they are to reach statistical significance. Nonetheless, the sample size is 
sufficient to reveal intergroup differences that are substantively important. 
The November/December sample contained data on 2,255 respondents, 
of whom 1,908 were white, 230 were African- American, and 117 were of 
Hispanic origin.' (Respondents classified as "Other" were not included in 
these analyses.) Table 5-1 compares probabilities of participation by race for 
November and December in the 10 core activities to those for the year as a 
whole. The Hispanic-Americans included in the November/December sam- 
ple were much less likely to report attending classical music concerts, much 
more likely to report acting, singing, or dancing on stage, and somewhat 
more likely to report reading imaginative literature than the Hispanic sample 
for the year as a whole. African-American respondents for November/De- 
cember were somewhat less likely to report having visited art exhibits or 
having read imaginative literature than their counterparts during the rest of 
the year. Attendance rates at classical music concerts, opera performances, 
plays, ballet performances, and art exhibits were lower for all groups in 
November/December than in all of 1982. For the most part, however, 
differences in participation between African-Americans, whites, and His- 
panics are similar for the full and for the November/December samples. 



Measures 

The SPPA gathered many measures of artistic socialization and current 
participation. In Chapter 3, we focused exclusively on the core participation 
items. Because this chapter explores the full range of data available, econ- 
omy of presentation dictates that we use scales — omnibus measures com- 
prising several similar items in a single variable. 

As we have seen, different kinds of participation are associated with race 
and ethnicity in different ways. To develop scales of arts participation, we 
applied a statistical method called factor analysis to the core participation 
and other participation variables described in Chapter 2. 

Factor analysis permits one to identify families of variables that are 
strongly associated with one another. In the case of the participation mea- 
sures, it revealed the existence of four such clusters. (See Appendix Table 
5-1.) 



72 



Evidence on Racial and Ethnic Differences 

Table 5-1 
Percentage Participating in Core and Other Arts Activities by 
Race/Etlinicity, November/December and Full 1982 Samples 





Attend jazz 


Attend 


Attend 


opera 


Attend 


Attend 




concert 


classical concert 


performance 


musical 


play 




Full 


N/D 


Full 


N/D 


Full 


N/D 


Full 


N/D 


Full N/D 


White 


9.1 


8.8 


14.4 


11.7 


3.3 


1.5 


20.7 


19.7 


13.4 11.6 


Afr.-Am. 


15.6 


16.9 


6.7 


5.0 


1.4 


0.5 


10.1 


8.6 


5.8 4.9 


Hispanic 


8.2 


9.0 


7.9 


2.2 


2.5 


0.8 


11.0 


11.8 


5.5 3.9 




Attend 


Visit art 


Perform on 


Perform: 


Read 




ballet 


exhibit 


musical 


act/sing/ 


fiction 












instrument 


dance 






Full 


N/D 


Full 


N/D 


Full 


N/D 


Full 


N/D 


Full N/D 


White 


4.6 


3.8 


23.9 


23.3 


4.0 


3.8 


4.7 


4.2 


60.2 60.1 


Afr.-Am. 


1.8 


0.7 


12.5 


9.8 


3.4 


3.7 


4.9 


4.4 


42.4 38.4 


Hispanic 


4.5 


2.8 


16.2 


15.9 


3.1 


4.6 


2.9 


7.8 


36.5 42.5 



Note: Weighted percentage of group engaging in activity at least once during 12 months preceding 
survey. 



Performing- Arts Attendance'. The first six core participation mea- 
sures, all involving attendance at performing-arts presentations, 
loaded together on a single factor. These included (in descending 
order of the strength of the relationship of each to the others) attending 
plays, attending ballet, attending musical theater, attending classical 
music performances, attending opera, and attending jazz perform- 
ances. The resulting variable is an additive scale of the activities, 
ranging from to 6. 

Exhibit Visiting: The core activity, visiting an art gallery or museum, 
combined with items on the "other participation" list, forms a second 
factor. The first four activities in this scale — in descending order, 
visiting historic monuments, visiting art or craft fairs, visiting science 
or history museums, and visiting art exhibits — all involved atten- 
dance at exhibitions. The fifth and sixth items, reading novels and 
other imaginative literature and doing needlecrafts, were anomalous, 



73 



Race, Ethnicity, and Participation in the Arts 



having in common only that they do not involve the performing arts. 
This additive scale ranges from to 6. 

• Performing-Arts Activities: A third factor consists of four activities, 
two from the core Ust and two from the "other participation" items, 
each of which involves producing, rather than consuming, perform- 
ing-arts events. In descending order these activities, summed to an 
additive scale ranging from to 4, are acting, singing or dancing on 
stage, playing a musical instrument in public or on stage, working on 
a theatrical set, and working on a musical set. 

• Non-Performance Activities: A fourth factor comprises six activities 
involving the visual or literary arts, each oriented towards production 
rather than consumption. In descending order, these are painting or 
drawing; creative writing; taking art, writing, or music lessons; 
photography ; crafts (other than needlecrafts) ; and reading or listening 
to poetry. The additive scale ranges from to 6. 

These four scales represent four kinds of cultural participation, varying 
along two dimensions: performing arts versus visually oriented arts (visual 
and plastic arts, historical exhibits, literature), and arts consumption versus 
arts production. 

The first scale includes jazz, which African-Americans attended more 
frequently than whites, along with five other activities that white respon- 
dents are more likely to attend than African-Americans. Because race thus 
affects different parts of the scale in different ways, cancelling one another 
out to a degree, we created a fifth scale by eliminating jazz from the 
performing-arts attendance activities. Results for the attendance scales in- 
cluding and excluding jazz, respectively, are reported separately throughout. 

A focus of this chapter is on how participation in the arts during 
childhood (i.e., before age 18) may affect participation in the arts as an adult. 
We subjected the measures of parental guidance and childhood classes (see 
Chapter 2 for a description of these variables) to factor analysis (Appendix 
Table 5-2), from which emerged two scales: 

• Parental Guidance: A scale ranging from to 4, consisting of the 
following items, in descending order: parents took child to plays or 
concerts; parents listened to classical music; parents took child to art 
museums; and parents encouraged child to read. 

• Childhood Lessons: A scale ranging from to 8, consisting of items 
reporting lessons or classes before age 18 in the following areas, in 
descending order: visual-art production, art appreciation, writing, 



74 



Evidence on Racial and Ethnic Differences 

Table 5-2 

Means for Artistic Socialization, Musical Taste, TV Art Viewing, 

and Artistic Participation Scales by Race 



N 



Parental 



Lessons Art music TV arts 



White 


1,908 


1.134 


1.240 


1.509 


1.404 




Afr.-Am. 


230 


0.860 


0.864 


0.720 


1.082 




Hispanic 


117 


0.800 


0.667 


1.084 


1.027 






N 


Attend 


Attend* 


Exhibits 


Perform 


Do other 


White 


1,908 


0.571 


0.483 


2.288 


0.116 


0.762 


Afr.-Am. 


230 


0.365 


0.197 


1.203 


0.094 


0.449 


Hispanic 


117 


0.305 


0.214 


1.597 


0.166 


0.708 



*Excluding attendance at jazz performances. Means are weighted, Ns are unweighted. 



music appreciation, crafts, acting, playing an instrument or singing, 
and ballet. 

Throughout this report, we have speculated about the extent to which 
differences in participation reflect, on the one hand, barriers to participation 
and, on the other, differences in taste. In this chapter, we use two rough 
proxies for taste or interest in "high culture." The first is based on the 
question that asked respondents which of the following kinds of music they 
like to listen to: classical/chamber, opera, operetta/Broadway /musical/ show 
tunes, jazz, soul/blues/rhythm and blues, big band, country westem, blue- 
grass, rock, mood/easy listening, folk, barbershop, and hymns/gospel. Factor 
analysis (Appendix Table 5-3) yielded three factors, of which classical/ 
chamber, operetta/show tunes, and opera were associated strongly with the 
first, along with (at lower levels) big band and mood/easy listening music. 
(Jazz appeared on a distinct factor with soul/blues and rock; and a third factor 
included bluegrass, country westem, folk, barbershop, and hymns/gospel 
music.) From the components of the first factor, we constructed an additive 
scale, ranging in value from to 5, which we call art music. 

A final additive scale is TV arts, ranging from to 7, with 1 point for 



75 



Race, Ethnicity, and Participation in the Arts 

Table 5-3: Regression Analyses Predicting Scores on Parental 

Guidance Scale 



I.V. 


1a 


lb 


2a 


2b 


3a 


African-American 


-.073 
b 


-.092 
d 


-.073 
b 


-.099 
d 


.044 
a 


Hispanic 


-.071 
b 


-.078 
c 


-.076 
b 


-.085 
d 


.044 
a 


Fenfiale 






.119 
d 


.099 
d 


.116 
d 


Age 






-.114 
d 


-.156 
d 


.131 
d 


Father's education 










.336 
d 


Mother's education 










.276 
d 


d.f. 

R Squared 


1.750 
.008 


2,254 
.012 


1,750 
.033 


2.254 
.044 


1,750 
.271 



Note: Additive scale of number of kinds of family-based childhood artistic socialization activities 
respondents reported. Models labeled "a" are based on only those respondents for whom data on 
mother's and father's education were available. I.V.=lndependent variable. 



each kind of arts programming the respondent reported watching on televi- 
sion. Because such programs are available to most Americans free of charge, 
we regard this as a rough measure of interest in the arts, unaffected by 
barriers that may reduce attendance at live events or exhibitions. 

In addition to the measures described above, we use the same control 
variables introduced in Chapter 3, as well as three new ones. The latter 
include father's educational attainment in years; mother's educational attain- 
ment; and the number of hours the respondent reported watching television 
on an average day (Hours TV). Because data on father's or mother's 
education are missing for many cases, these variables are used only for 
analyses based on a special subsample. Hours of television is included as a 
control variable for analysis with TV arts. 



76 



Evidence on Racial and Etiinic Differences 

intergroup Differences in Socialization, Taste, and 

Participation Scaies 

Let us begin by considering intergroup differences in mean scores on the 
scales described above. (See Table 5-2.) Not surprisingly, the patterns mirror 
those noted in Chapter 2 with respect to the items of which these scales 
consist. White respondents reported more parental guidance experiences 
(1.13 compared to .86 and .80) than African- American or Hispanic respon- 
dents, respectively, as well as more kinds of classes or lessons (1.24) than 
African-Americans (.86) or, especially, Hispanic-Americans (.67). Whites 
reported liking more of the musical genres on the art music scale (1.51) than 
Hispanic (1.08) or, especially, African- American (.72) respondents. They 
also reported watching more kinds of televised arts programs (1.40) than 
African-American or Hispanic respondents (1.08 and 1.02). 

Whites also had higher scores than African-Americans and Hispanics on 
all the participation measures but performance activities. The differences 
were greatest with respect to the visually oriented consumption scale, for 
which the average for white respondents was 2.29, compared to 1.60 for 
Hispanic and 1.20 for African- Americans. Intergroup differences in other 
areas were more modest. Indeed, Hispanics participated in slightly more 
performance activities and almost as many nonperformance activities as 
whites. 

Although differences among groups are notable, especially with respect 
to consuming, rather than producing, art, even more striking is the modest 
degree of participation evident among any of these groups. Fewer than half 
the respondents from any group, for example, attended a performing-arts 
activity other than jazz, or participated in a performance, either onstage or 
backstage. Variation by race or ethnicity is limited, then, because whites, 
African- Americans, and Hispanic- Americans all reported low rates of par- 
ticipation. 



Race, Ethnicity, and Parental Guidance 

African- Americans and Hispanic- Americans report fewer parental-guid- 
ance experiences than do white Americans. Do these differences reflect 
differences in the degree to which African-American, Hispanic, and white 
parents value the arts? Or do they, instead, stem from differences in socio- 
economic opportunity related to race or ethnicity? 



77 



Race, Ethnicity, and Participation in the Arts 

To answer this question, we used multiple regression analysis, a method 
that lets one estimate net effects of race and ethnicity while holding other 
potential causal factors constant. In other words, the resulting coefficients 
describe differences between African-Americans and whites and between 
Hispanics and whites who are similar with respect to the variables for which 
we have controlled. Table 5-3 reports results of analyses predicting scores 
on the parental guidance scale, and Table 5-4 reports results of the analyses 
for lessons taken during childhood. Independent variables are arrayed verti- 
cally to the left of the page. Their statistical effects appear on the right, 
expressed as standardized coefficients, enabling us to compare the impacts 
of different predictors in a common metric. 

Each table reports results of three separate analyses or models, each 
containing different sets of variables. The pair of columns labelled la and lb 
reports the influence of being African- American or Hispanic (as compared 
to white, the omitted category), without controlling for any other factors. As 
such, they are comparable to Table 5-2. The second pair of columns, 2a and 
2b, reports results of models that included controls for gender and age. The 
column labelled 3 is based on a model that included controls for parental 
education. 

The first two models (1 and 2) are reported in two columns because the 
analyses were executed twice: once on the full November/December sample 
and once on a partial subsample, consisting of 1,751 cases from November 
and December that contained data on mother's and father's education. The 
latter data are somewhat biased because respondents who could not report 
their parents' educational level were disproportionately lower in socio- 
economic status than the sample as a whole. On the other hand, the subsam- 
ple includes information that is vital for understanding family influences.^ 

Columns 1 a and 1 b for tables 5-3 and 5-4 confirm that African-American 
and Hispanic respondents received significantly less childhood experience 
in the arts than their white counterparts. Columns 2a and 2b indicate that this 
difference remains constant (for parental guidance) or grows (for lessons and 
classes) after controlling for differences in gender composition and age 
among the three groups. 

The models reviewed thus far fail to take into account the fact that 
parents of African-Americans and Hispanic-Americans, on average, re- 
ceived considerably less formal education than parents of white Americans. 
When we control for mother's and father's education in model 3, two things 
become clear. First, parental education explains much more variation in 
parental guidance than does race or ethnicity. Second, African- American 
and Hispanic respondents received no less parental guidance than did white 



78 



Evidence on Racial and Etiinic Differences 

Table 5-4 

Regression Analyses Predicting Scores 

on Childhood Lessons Scale 



I.V. 


la 


lb 


2a 


2b 


3a 


African-American 


-.069 
b 


-.085 
d 


-.074 
c 


-.100 
d 


-.011 


Hispanic 


-.080 
c 


-.091 
d 


-.102 
d 


-.111 
d 


-.038 


Female 






.057 
a 


.051 
b 


.056 
b 


Age 






-.341 
d 


-.371 
d 


-.209 
d 


Father's education 










.168 
d 


Mother's education 










.159 
d 


d.f. 

R Squared 


1,750 
.009 


2,254 
.013 


1,750 
.125 


2,254 
.151 


1,750 
.192 



Note: Additive scale of number of kinds of lessons or classes respondent reported taking before the 
age of 18. Models labeled "a" are based on only those respondents for whom data on mother's and 
father's education were available. 



Americans of equivalent age with similarly educated parents. Indeed, both 
African-American and Hispanic respondents reported that their parents gave 
them slightly, but significantly, more kinds of exposure or encouragement 
than did parents of whites. Parental education had less influence on classes 
or lessons, which include those for which the schools as well as those for 
which the family are responsible. Nonetheless, once one controls for moth- 
er's and father's years of schooling, the effects of race and ethnicity on 
lessons are no longer significant. 



79 



Race, Ethnicity, and Participation in the Arts 

Race, Ethnicity, Musical Taste, 
and Television Arts Viewing 

We have seen that Hispanic and, especially, African- American respon- 
dents reported liking fewer of the genres on the art-music scale than whites 
and viewed somewhat fewer kinds of televised arts programs. Do race and 
ethnicity exert an independent influence on taste for art music or interest in 
the watching of arts programs on television? Or do differences stem entirely 
from intergroup variation in characteristics like socioeconomic status or 
artistic socialization that are related to artistic tastes or interests? 

With respect to scores on the art-music scale (which includes big bands 
and easy listening as well as classical music, opera, and musical theater), 
being African-American, but not being Hispanic, makes a differences (see 
Table 5-5). Without controls, both African-Americans and Hispanics report 
liking significantly fewer of these musical styles than whites. Controlling for 
sociodemographic factors eliminates the difference between whites and 
Hispanics, and accounts for almost half the difference between African- 
Americans and whites. Nonetheless, the remaining effect of race indicates 
that African-American and white musical tastes are significantly different. 
Controlling for childhood experience reduces the remaining African-Ameri- 
can/white margin by only 14 percent, and the difference remains statistically 
significant. 

Race is not a major factor, however, compared to other significant 
predictors of differences in art-music scores. The effect of age, for example, 
is almost four times that of race, the influence of educational attainment 
almost three times as great, and the effects of parental guidance and child- 
hood lessons twice as substantial. (See Appendix Table 5-5.) 

The small but significant tendency for African- Americans and Hispanics 
to report viewing fewer kinds of televised arts programs than whites is 
entirely the result of sociodemographic differences among these groups. In 
other words, if we take such viewing as a measure of interest in the arts, 
African-Americans and Hispanic-Americans display just as much interest as 
do whites who are similar in educational attainment, occupational status, 
income, and related characteristics.^ 

Race, Ethnicity, and Artistic Participation 

In this section we consider effects of race and ethnicity on five scales of 
participation: attending performing-arts events (jazz included); attendance at 
performing-arts events (jazz excluded); visiting museums, fairs, or exhibits. 



80 



Evidence on Racial and Ethnic Differences 

Table 5-5 

Effects of Race and Ethnicity on Art-Music Scale and Number of 

Kinds of Televised Arts Programs Viewed 

Art music TV arts 



Model 


1 


2 


3 


1 


2 


3 


African- 
American 


-.179 
d 


-.093 
d 


-.080 
d 


-.057 
b 


.017 


.028 


Hispanic 


-.067 
b 


.001 


.019 


-.047 
a 


.009 


.030 



Note: Standardized beta coefficients. a=p less than .05; b=p less than .01 ; c=p less than .001 ; d=p 
less than .0001 . Model 1 includes no control variables. Model 2 includes controls for gender, age, 
educational attainment, occupation (white-collar v. other), family income, marital status (single or 
divorced v. other), and residence in SMSA. Model 3 includes same controls as model 2 as well as 
controls for parental guidance and childhood lessons. Based on 2,255-person sample from Novem- 
ber/December 1982. 



reading literature and related activities; performing on stage or working 
backstage at performances; and producing visual, craft, or literary arts. 

Throughout this report we have emphasized that artistic participation is 
multi-dimensional. Because the participation scales used in this chapter vary 
along two dimensions (consuming/producing, performing arts/other arts), 
we can use them to pursue this point. The reader should remember, however, 
that even the broad array of activities included in the participation scales 
does not begin to exhaust the diversity of artistic activities in the contempo- 
rary United States. In particular, except for jazz, the SPPA did not ask people 
about art forms or activities with special links to African-American, His- 
panic, or other American racial or ethnic minority communities. 

Absent controls for other variables (Table 5-6, model 1 under each 
participation heading), African-Americans reported participating in fewer 
items than white Americans on each scale except performance activities. 
Hispanic respondents reported fewer consumption activities than non-His- 
panic whites, but not fewer production activities. None of the zero-order 
differences is very large, although the differences between African- Ameri- 



81 



Race, Ethnicity, and Participation in tlie Arts 

Table 5-6 
Effects of Race and Ethnicity on Arts Participation Scales 







Attend performances 






Attend performances ' 


Model: 


1 


2 


3 


4 


1 


2 


3 4 


African- 


-.067 


.015 


.026 


.031 


-.107 


-.022 


-.013 -.006 


American 


b 








d 






Hispanic 


-.061 
b 


-.002 


.012 


.001 


-.070 
c 


-.008 


.004 -.007 






Visual consumption 






Performance activities 


Model: 


1 


2 


3 


4 


1 


2 


3 4 


African- 


-.199 


-.115 


-.101 


-.094 


-.016 


-.004 


.003 .007 


American 


D 


d 


d 


d 








Hispanic 


-.088 
d 


-.024 


-.004 


-.016 


.025 


.038 


.049 .045 
a a 



Model: 



Other activities 



African- 


-.091 


-.056 


-.035, 


-.032 


American 


d 


b 






Hispanic 


-.011 


.016 


.043 
a 


.034 



^Second attendance scale does not include jazz. Note: Standardized beta coefficients. a=p less than 
.05; b=p less than .01 ; c=p less than .001 ; d=p less than .0001 . Model 1 includes no control variables. 
Model 2 includes controls for gender, age, educational attainment, occupation (white-collar v. other), 
family income, marital status (single or divorced v. other), and residence in SMSA. Model 3 includes 
same controls as model 2 as well as controls for parental guidance and childhood lessons. Model 4 
includes same controls as model 3 as well as controls for art-music scale, TV arts viewing, and hours 
spent watching all kinds of television on average day. Based on 2,255-person sample from Novem- 
ber/December 1982. 



82 



Evidence on Racial and Ethnic Differences 



cans and whites with respect to visually oriented consumption activities and, 
to a lesser extent, attending performances (excluding jazz) are moderate. 

When sociodemographic controls are entered into the predictive equa- 
tions (model 2), the negati\'e coefficients of being African-American on 
performance attendance disappear (with jazz included) or become insignifi- 
cant (with jazz excluded). Controlling for such factors as educational attain- 
ment, family income, having a white-collar occupation, and marital status 
eliminates all of the differences between African- Americans and whites on 
the performance-attendance scale that includes jazz, and almost 80 percent 
of the difference on the scale excluding jazz. Sociodemographic controls 
also reduce the effect of race on visually oriented consumption activities by 
more than 40 percent, and on visual-art, craft, and literary activities by 
almost as much, but the differences between whites and African-Americans 
remain statistically significant in these areas. 

Sociodemographic differences account for almost all of the difference 
between whites and Hispanics in performance attendance and more than 70 
percent of the gap in the exhibit-visiting scale. When these characteristics 
are taken into account, being Hispanic has no significant influence on any 
form of participation. 

The third model adds controls for childhood experience (both parental 
guidance and lessons and classes) to the sociodemographic measures. These 
additional controls reduce the remaining effect of being African-American 
on visiting exhibits by only 12 percent, leaving a small but statistically 
significant difference between otherwise similar African-Americans and 
whites. They reduce the African-American coefficient for nonperformance 
creative activities by almost 40 percent, to nonsignificance. 

Although the impact of being African- American on visiting exhibits is 
statistically significant, it is small relative to the influence of other predic- 
tors. For example, it is less than half the size of the effects of educational 
attainment, gender, and parental socialization, and well below the influence 
of lessons and classes during youth."^ 

We have already seen that whenever Hispanics had significantly lower 
scores on participation scales than whites, these differences were almost 
entirely the consequence of intergroup sociodemographic differences. With 
respect to the consumption scales — performance attendance and visually 
oriented activities — controlling for childhood experience makes no notable 
difference. With respect to the art-producing activities, both performance 
and nonperformance, when one controls for childhood experience in the arts, 
Hispanic Americans are involved in slightly, but significantly, more activi- 
ties than are whites. In other words, Hispanic respondents reported partici- 



83 



Race, Ethnicity, and Participation in the Arts 

pating in more artistic production activities than did white or African- Ameri- 
can respondents of similar socioeconomic status and with comparable so- 
cialization into the arts. 

With respect to nonperformance activities, the positive effect of being 
Hispanic is small relative to that of other predictors: about one-eighth as large 
as childhood lessons, less than one-third the effect of home socialization, less 
than half the size of educational attainment, and smaller than the effect of age, 
income, marital status, having a white-collar occupation, and living in an 
SMSA. By contrast, the coefficient for Hispanic origin, although small, is one 
of only four significant predictors of onstage or offstage performance activi- 
ties, and the largest demographic predictor other than income. 

The fourth model we investigated added three new control variables: the 
art-music scale, the TV arts viewing scale, and a measure of hours watched 
per day of all kinds of television. These additional controls did not materially 
alter the results of the earlier models, except insofar as they reduced the 
coefficient for being Hispanic as a predictor of participation in nonperform- 
ance production activities to insignificance. What this means is that more 
than one-fifth of the advantage associated with being Hispanic in nonper- 
formance production ("other activities") results from Hispanic respondents 
having musical tastes and viewing habits associated with this kind of partici- 
pation. 



Summary of Findings Thus Far 

The analyses reported above clarify certain issues raised in earlier chap- 
ters. In Chapter 2, we saw that African-American and Hispanic respondents 
received fewer forms of parental guidance in reading and the fine arts and 
took fewer arts-related classes or lessons at an early age than did whites. In 
this chapter, we have seen that these differences are entirely a result of the 
fact that African-American and Hispanic respondents had parents who had 
received fewer years of formal education than had the parents of white 
respondents. Controlling for parental education, African-American and His- 
panic parents gave their children significantly more parental guidance expe- 
riences than did comparable white parents. To the extent that the way one 
socializes one's children reflects the value one places on the arts, then 
African-American and Hispanic families appear to value the arts (and read- 
ing) as much as comparable white families. 

In Chapters 2 and 3 we raised the question of whether differences in 
participation between whites on the one hand, and African-Americans and 



84 



Evidence on Racial and Etiinic Differences 

Hispanics on the other, resulted from differences in opportunity or from 
differences in taste. In this chapter we have looked at two proxy indicators 
of taste for the fine arts. The first, a scale of the number of kinds of art music 
and related genres respondents said they enjoyed, is a fairly direct indicator 
of a narrow spectrum of taste. The second, a scale of the number of kinds of 
arts programs respondents reported viewing on television, is a more indirect 
indicator of interest in a broader range of the arts. 

If we treat television viewing of the arts as an indicator of interest in the 
arts, then we see that African-American and Hispanic-Americans are no less 
interested in the arts than are white Americans of similar socioeconomic 
status. The same is true of taste for classical and related forms of music for 
Hispanic- Americans. By contrast, African-Americans do report liking fewer 
kinds of art music (but recall that this scale includes big band, Broadway, 
and easy listening music, as well as classical) than whites, and only about 
half of the difference is explained by sociodemographic characteristics. 
However, the results reported in Table 5-6 for model 4 indicate that this 
small difference in taste cannot explain interracial differences in any of the 
arts participation scales. 

Whereas most of the core questions examined in Chapter 3 concemed 
attendance at live, high-culture, performing arts events, use of the "other 
participation" items in constructing the participation scales permitted us to 
distinguish among the determinants of different kinds of participation. The 
analyses further confirmed that one cannot generalize about the effects of 
race or ethnicity on cultural participation per se. Hispanic Americans at- 
tended fewer public arts activities than whites, but this difference is almost 
entirely the result of the fact that white Americans have more formal 
education, higher incomes, and higher status occupations. When these fac- 
tors are controlled, Hispanic-Americans participate in active art-making 
activities significantly more than do white Americans. 

African-American/white differences in participation also vary for differ- 
ent kinds of activities. There is no statistically significant difference between 
African-American and white respondents with respect to participating on- 
stage or backstage in performing-arts events. And the significant difference 
between African-Americans and white Americans in the number of kinds of 
performing-arts events attended stems almost entirely from differences be- 
tween African-Americans and whites in sociodemographic characteristics 
other than race.'^ Significant, although relatively small, differences between 
white and African-American respondents who are similar in sociodemo- 
graphic profile did appear with respect to the scales measuring visually 
oriented consumption activities and in the nonperformance creative activity 



85 



Race, Ethnicity, and Participation in tlie Arts 

scale. The latter difference was attributable to differences between African- 
Americans and whites in childhood experience, whereas the former persisted 
even after controls for childhood experience, musical taste, and televised arts 
viewing were entered into the model. 

One advantage of multiple regression analysis over logistic regression 
analysis (the method used in Chapter 3) is that it enables one to compare the 
relative influence of different predictive factors using a common metric. The 
analyses reported above indicate that even in those relatively few cases in 
which race or ethnicity affect artistic outcomes after controlling for inter- 
group sociodemographic differences, those effects are usually dwarfed by 
those of childhood experience and educational attainment, and exceeded by 
other measures of socioeconomic status. 

In other words, at least for the range of participation measures about 
which the SPPA surveys asked, most differences among white, African- 
American, and Hispanic respondents result from differences in the sociode- 
mographic attributes of members of these groups. Where differences in 
participation other than those for which such factors account are found, they 
vary among kinds of participation. African- Americans report receiving more 
kinds of parental guidance into the arts, like art music and related genres less 
(but like jazz more), visit fewer kinds of public exhibitions, and engage in 
fewer arts, crafts, and literary creative activities than whites who are compa- 
rable with respect to sociodemographic characteristics. Hispanic- Americans 
report benefitting from more kinds of parental guidance and participate in 
more active art-producing activities (both performance and nonperform- 
ance) than white Americans who are comparable sociodemographically and 
with respect to childhood experience. Such net differences, where they are 
present, are in most cases small relative to other predictors. 

Differences in l\Aodels Predicting Artistic 
Socialization, Taste, and Participation by Race 

Do the same factors predict outcomes for African- Americans, Hispanics, 
and whites, or do members of these groups follow separate paths to artistic 
participation? Differences in the predictors of participation are relevant both 
to understanding intergroup differences in the extent of participation and to 
evaluating the likely effects of programs and policies aimed at reducing such 
differences. 

In this section we investigate differences in the predictors of socializa- 
tion, taste, and participation by applying the same predictive models 
described in the previous section (excluding the dichotomous African- 



86 



Evidence on Racial and Ethnic Differences 



American and Hispanic variables, of course) separately to respondents from 
each group. For the childhood experience variables (parental guidance and 
childhood lessons) these analyses employ the subsample with data on 
mother's and father's educational attainment and also include gender and 
age. For art music and TV arts viewing, the full November/December 
subsample is used for two separate models: with sociodemographic predic- 
tors, and with both sociodemographic and socialization variables included. 
For the participation scales (attending performances with and without jazz, 
visiting exhibits, performance activities, and nonperformance activities), 
three models are run using the full November/December subsample: with 
sociodemographic predictors only; with sociodemographic and childhood 
experience independent variables; and with sociodemographic, childhood 
socialization, and taste proxy measures all included. 

Table 5-7 reports all instances where predictors for two or more groups 
are significantly different across comparable models. (The full models are 
reported in Appendix Tables 5-13 through 5-20.) Most significant differ- 
ences are between whites and African-Americans or between whites and 
Hispanics. In part, this is an artifact of sample size. Because the number of 
white respondents is much greater than the number of African-American or 
Hispanic respondents, differences between whites and other groups are more 
likely to be statistically significant than gaps between Hispanics and Afri- 
can-Americans. 

Childhood experience. There were no significant intergroup differences in 
the predictors of childhood classes and lessons. By contrast, once parental 
education was controlled, age was a significantly positive predictor of 
parental guidance for whites but a significantly negative predictor for Afri- 
can-Americans. What this means is that whereas white parents of equivalent 
educational levels have been providing fewer kinds of parental guidance over 
the lifetimes of our respondents, comparable African-American parents have 
been providing more kinds of guidance over that same time span. This trend, 
along with increases in educational attainment among African-Americans, 
might be expected to moderate or eliminate African-American/white differ- 
ences in family guidance. 

Taste/ interest proxies. Older white respondents watched significantly more 
kinds of televised arts programs and reported liking significantly more kinds 
of art music, other things equal, than younger whites. By contrast, older 
African-American and Hispanic respondents were no more likely than other- 
wise comparable younger ones to have high scores on these scales. Significant 



87 



Race, Ethnicity, and Participation in the Arts 

Table 5-7 
Significant Differences in lUlodels Predicting Scores on Artistic 
Socialization, Taste, and Participation Scales for African- 
American, Hispanic, and White Subsamples 



SCALE 



PREDICTOR 



Parental 
guidance 



AGE 



Significantly positive for whites, negative for African- 
Americans, controlling for gender and parental 
education 



Childhood 
lessons 



None 



TV arts viewing AGE 



EDUCATION 



Significantly positive for whites, only slightly positive for 
African-Americans with sociodemographic controls 

Significantly positive for both whites and African- 
Americans, but effect for whites significantly stronger, 
with sociodemographic controls 



Art music 



AGE 



EDUCATION 



Significantly positive effect for whites, insignificant 
weak effects for African-Americans and Hispanics, both 
with sociodemographic controls only and with socio- 
demographic and socialization controls. 

With sociodemographic controls, strongly significant for 
whites, significant but less so for African-Americans; 
with socialization controls, still strongly significant for 
whites, insignificant for African-Americans 



Performance 


EDUCATION 


attendance 




Inc. jazz 




Performance 


EDUCATION 


attendance 




exc. jazz 





SMSA 



More strongly significant for whites than for African- 
Americans with sociodemographic control only; insignifi- 
cant for Hispanics with sociodemographic controls and 
negative for Hispanics with additional controls 

More strongly significant for whites than for African- 
Americans with sociodemographic and socialization 
controls; insignificant for Hispanics with sociodemo- 
graphic controls and negative with other controls 

Significantly positive for whites, negative for African- 
Americans with sociodemographic controls; negative 
and significant for African-Americans, insignificant for 
whites with sociodemographic and socialization controls 



88 



Evidence on Racial and Etiinic Differences 



Table 5-7 (Continued) 
Significant Differences in fUlodels Predicting Scores on Artistic 
Socialization, Taste, and Participation Scales for African- 
Americans, Hispanic, and White Subsamples 



SCALE 



PREDICTOR 



Exhibition 
visiting 



EDUCATION 



OCCUPATION 
(white collar) 



Strongly significant for whites, all models; for African- 
Americans, more weakly significant with sociodemo- 
graphic controls, insignificant with other controls 

More significantly positive for African-Americans than 
for whites with sociodemographic controls; significantly 
positive for Hispanics, insignificant for whites, with so- 
ciodemographic, socialization, and taste controls 





PARENTAL 




TV ARTS 


Performance 
activity 


INCOME 


Nonperformance 
activity 


AGE 




EDUCATION 




OCCUPATION 
(white collar) 




PARENTAL 
SOCIALIZATION 




TV ARTS 
VIEWING 



GENDER (female) Significantly positive for whites, insignificant for Hispan- 
ics, all models 



More significantly positive for African-Americans than 
for whites, all models 

More significantly positive for Hispanics than for whites 

Significantly negative for whites, all models; signifi- 
cantly positive for Hispanics with sociodemographic 
and with sociodemographic and socialization controls, 
and positive but insignificant with all controls 

Significantly negative for whites, positive for African- 
Americans, with all controls 

More significantly positive for whites than for African- 
Americans with sociodemographic controls 

Significantly positive for African-Americans, all models; 
significantly but less positive for whites, models with so- 
ciodemographic and with sociodemographic and sociali- 
zation controls, insignificant in model with all controls 

Significantly positive for Hispanics but not for African- 
Americans, model with socialization controls 

Significantly positive for Hispanics, less significantly 
positive for whites, not significant for African-Americans 



Note: For full models, see Appendix Tables 5-1 1 through 5-17. 



89 



Race, Ethnicity, and Participation in the Arts 

differences in the effects of age for whites as compared to African- Americans 
(for art music and TV arts viewing) and Hispanics (for art music) suggest the 
possibility of a convergence in musical taste and interest in the arts. Although 
these differences may simply represent an absence of aging effects in the 
minority subpopulations, they may instead reflect cohort change in the 
African-American and Hispanic communities. One other intergroup differ- 
ence was evident. Educational attainment was more strongly and positively 
predictive of TV arts viewing and liking for art music and related genres for 
white than for African-American respondents. 

Participation Scales. The most notable intergroup difference was that edu- 
cational attainment was more strongly related for whites than for African- 
Americans to performing-arts attendance (both including and excluding 
jazz), exhibition attendance and related activities, and nonperformance crea- 
tive activities. In most cases, the effect of education was significant for 
African-Americans as well as for whites, but smaller in magnitude. A similar 
pattern appeared in the difference between white and Hispanic respondents 
in effects of education on attending the performing arts (both including and 
excluding jazz), but not on the other participation scales. 

In other words, to use the language of economics, retums to investments 
in education in the form of increased participation in a range of artistic 
activities are larger for whites than for African-Americans or Hispanics. One 
possible explanation for such a finding is that African- American respondents 
may have received different kinds of education than white respondents. If, 
for example, African-Americans were more likely to go to high schools 
where the arts were not stressed, to take vocational rather than college 
preparatory courses, to attend community colleges rather than liberal arts 
colleges, or to major in technical or business subjects rather than in the 
humanities, any of these factors might account for the differences in the 
effects of education. 

By contrast, the effects of having a white-collar occupation on nonper- 
formance consumption and production activities were larger for African- 
Americans than for whites, as they were on visiting exhibitions and related 
activities for Hispanics. In other words, there is some evidence that, at least 
with respect to nonperformance items, occupation plays a more important 
role in determining the participation of African-Americans and Hispanics 
than of whites, whereas education is more important in determining partici- 
pation levels of whites than of African-Americans and Hispanics. 

Other intergroup differences in the effects of sociodemographic factors 
were each restricted to just one form of participation in the arts. Living in an 



90 



Evidence on Racial and Ethnic Differences 



SMSA had a positive effect on performing-arts attendance (excluding jazz) 
for white respondents, but a negative influence on attendance for African- 
Americans. Other things being equal, v^hite women were more likely to visit 
museums and exhibits than white men, but no such gender differences 
appeared for Hispanics. Family income was positively related to onstage and 
backstage performance activities for Hispanic respondents, but negatively 
related to such activities for whites. Consistent with findings described in 
Chapter 3, the gap in participation between women and men was greater 
among whites than among African-Americans for all the scales, but unlike 
those analyses, the differences never reached statistical significance. 

In general, the effects of childhood socialization on participation scores 
were weaker, although still significant, for whites than for members of other 
groups. The only difference between groups that was significant, however, 
was for visually oriented consumption activities, where parental guidance 
exerted a significantly stronger impact on participation by African-Ameri- 
cans respondents than on that of whites. 

In Chapter 2, we noted that differences between African-Americans and 
whites with respect to taking classes or lessons in the arts were relatively 
small, compared to differences in participation in the core activities, and 
speculated about the efficacy of the schools in increasing equality of oppor- 
tunity for participation in the arts. Except for the performance attendance 
scale that included jazz, the effects of lessons or classes in the arts were 
smaller for African-Americans than for whites or Hispanics. This finding is 
consistent with the weaker effects of educational attainment on participation 
for African- Americans than for whites, and may indicate either that AMcan- 
Americans took different kinds of classes or lessons than members of other 
groups or that, for some other reason, classes or lessons were less effective 
in stimulating adult activity among African-Americans than among other 
respondents. On the other hand, these differences, although pervasive, never 
reached statistical significance, so, at most, they suggest hypotheses for 
further research. 

In Chapter 2, we also noted the smaller differences in pattems of 
watching the arts on television than in pattems of live attendance between 
white Americans, on the one hand, and African- Americans and Hispanic- 
Americans, on the other, and speculated as to whether television might be a 
force for increasing minority participation in the arts. For Hispanic respon- 
dents, this hypothesis seems to be a credible one: watching televised arts 
programs is significantly related to each of the participation scales, even after 
controlling for sociodemographic factors, childhood experience measures, 
taste for art music, and amount of television viewing of all kinds. For each 



91 



Race, Ethnicity, and Participation in the Arts 

scale, the impact of viewing the arts on television is greater for Hispanics 
than for any other group, and for the nonperformance scales, both visually 
oriented consumption and nonperformance production activities, the rela- 
tionship is significantly stronger for Hispanic respondents than for African- 
Americans. By contrast, for African-Americans, viewing the arts on 
television has a weaker effect on each of the participation scales than for 
whites or for Hispanics, and is a significant predictor only of the perform- 
ance attendance scales. The effects of TV arts viewing on participation for 
whites is intermediate between that for Hispanics and African- Americans for 
each kind of participation. 

What can we make of these differences? One possibility is that televised 
arts programs boost arts participation among Hispanic-Americans more than 
among African- Americans or whites. A plausible alternative explanation is 
that participating in the arts as consumers or producers makes Hispanics 
want to watch arts programs on television more than it does African- Ameri- 
cans or whites. Or, viewing arts programs may simply be a better proxy 
measure of interest in the arts for Hispanics than for members of other 
groups. These possibilities can at best serve as hypotheses for further 
research, especially given the fact that only two of the intergroup differences 
are statistically significant. 

Taken together, however, the findings suggest an intriguing and poten- 
tially important hypothesis: the links between childhood classes and lessons 
(but not parental guidance), formal education, televised arts viewing, and 
artistic participation may be weaker among African-Americans than for the 
Hispanic or white subpopulations. Whether this conclusion would survive 
replication, given the statistical insignificance of many of the results, is 
uncertain. If the hypotheses are confirmed, it remains to be seen whether 
the differences result from differences in the kinds of education African- 
Americans and other Americans receive, the kinds of classes they take, and 
the kinds of arts programs they watch; or from aspects of the African-Ameri- 
can experience that blunt the impact of education on artistic interests and 
behavior. 



Do Intergroup Differences Vary by Gender, 
Educational Attainment, or Age? 

Table 5-8 displays means by race for subsamples based on differences 
among respondents in educational attainment, gender, and age. The educa- 
tional attainment categories are less than high school, high school graduation 



92 



Evidence on Racial and Etiinic Differences 



Table 5-8 

Means and Standard Deviations for Regression Variables 

by Race by Education, Gender, Age — Including 

Respondents without Data or Parental Education 











Art 






Attend 










N 


Lessons 


Home 


music 


Attend 


Visit 


no jazz 


Perform 


Dovis 


TV arts 


EDUCATION 






















11 & Less 






















White 


447 


0.528 


0.550 


0.894 


0.148 


1.065 


0.129 


0.079 


0.276 


0.660 






1.054 


0.674 


1.195 


0.482 


1.203 


0.447 


0.346 


0.692 


1.333 


African- 


106 


0.318 


0.516 


0.316 


0.115 


0.467 


0.058 


0.023 


0.143 


0.668 


American 




0.752 


0.659 


0.927 


0.319 


0.812 


0.233 


0.149 


0.452 


1.386 


Hispanic 


54 


0.290 


0.626 


0.886 


0.161 


0.966 


0.088 


0.152 


0.494 


0.806 






0.810 


0.919 


1.272 


0.368 


1.331 


0.284 


0.496 


0.927 


1.681 


12 Years 






















White 


801 


1.187 


1.027 


1.368 


0.387 


2.275 


0.323 


0.082 


0.694 


1.260 






1.312 


0.870 


1.287 


0.785 


1.630 


0.696 


0.341 


1.025 


1.689 


African- 


80 


1.272 


0.990 


0.872 


0.304 


1.400 


0.084 


0.181 


0.469 


1.200 


American 




1.557 


0.824 


1.168 


0.549 


1.586 


0.337 


0.502 


0.874 


1.498 


Hispanic 


38 


1.093 


0.804 


1.239 


0.332 


1.959 


0.264 


0.244 


0.781 


0.955 






1.297 


0.650 


1.062 


0.738 


1.653 


0.553 


0.642 


1.317 


1.455 


13-15 Years 






















White 


342 


1.758 


1.558 


1.833 


0.837 


2.779 


0.700 


0.166 


1.044 


1.676 






1.580 


0.963 


1.490 


1.169 


1.674 


1.023 


0.536 


1.297 


1.847 


African- 


28 


1.082 


1.285 


1.088 


0.673 


1.993 


0.432 


0.111 


0.910 


1.299 


American 




1.206 


0.855 


1.384 


0.906 


1.688 


0.735 


0.400 


1.102 


1.528 


Hispanic 


20 


0.714 


1.209 


1.344 


0.605 


2.347 


0.460 


0.101 


1.069 


1.687 






0.760 


0.589 


1.202 


1.209 


1.316 


0.896 


0.301 


1.412 


1.732 


16 & Over 






















White 


318 


1.781 


1.738 


2.339 


1.308 


3.457 


1.120 


0.194 


1.286 


2.471 






1.510 


1.025 


1.427 


1.345 


1.596 


1.196 


0.618 


1.305 


2.103 


African- 


16 


1.649 


1.438 


1.625 


1.474 


3.001 


1.046 


0.067 


1.242 


2.471 


American 




1.566 


0.940 


1.371 


1.534 


1.575 


1.137 


0.249 


1.119 


2.286 


Hispanic 


5 


1.292 


0.987 


1.022 


0.432 


2.602 


0.216 


0.000 


1.008 


1.271 






1.179 


0.458 


1.602 


0.823 


2.220 


0.411 


0.000 


1.256 


1.993 



93 



Race, Ethnicity, and Participation in thie Arts 

Table 5-8 (Continued) 

Means and Standard Deviations for Regression Variables 

by Race by Education, Gender, Age — Including 

Respondents without Data or Parental Education 











Art 






Attend 










N 


Lessons 


Home 


music 


Attend 


Visit 


no jazz 


Perform 


Dovis 


TV arts 


GENDER 






















Male 






















White 


860 


1.177 


1.033 


1.357 


0.475 


1.839 


0.389 


0.094 


0.688 


1.319 






1.344 


0.896 


1.349 


0.906 


1.594 


0.787 


0.399 


1.075 


1.770 


African- 


92 


0.921 


0.783 


0.652 


0.361 


0.890 


0.168 


0.090 


0.450 


1.055 


American 




1.450 


0.823 


1.217 


0.754 


1.249 


0.511 


0.341 


0.855 


1.621 


Hispanic 


56 


0.631 


0.817 


0.994 


0.241 


1.473 


0.161 


0.134 


0.742 


1.085 






1.020 


0.722 


1.069 


0.692 


1.351 


0.496 


0.469 


1.223 


1.593 


Female 






















White 


48 


1.296 


1.225 


1.645 


0.657 


2.690 


0.567 


0.135 


0.828 


1.480 






1.494 


1.029 


1.459 


1.096 


1.756 


0.974 


0.477 


1.162 


1.864 



African- 138 0.818 0.922 0.776 0.369 1.457 0.220 0.097 0.447 1.104 
American 1.199 0.847 1.171 0.780 1.684 0.603 0.375 0.864 1.599 



Hispanic 


61 


0.703 


0.783 


1.175 


0.368 


1.723 


0.268 


0.197 


0.675 


0.968 






1.120 


0.870 


1.368 


0.778 


1.824 


0.590 


0.556 


1.156 


1.738 


AGE 






















18-30 






















White 


605 


1.915 


1.288 


1.023 


0.528 


2.511 


0.401 


0.136 


1.142 


1.134 






1.653 


0.921 


1.113 


0.982 


1.712 


0.811 


0.467 


1.349 


1.535 


African- 


80 


1.454 


1.018 


0.680 


0.544 


1.606 


0.197 


0.140 


0.600 


1.310 


American 




1.515 


0.799 


1.099 


0.809 


1.609 


0.555 


0.465 


0.969 


1.661 


Hispanic 


44 


1.058 


0.917 


1.229 


0.282 


1.874 


0.196 


0.248 


1.167 


1.177 






1.268 


0.589 


1.145 


0.765 


1.621 


0.549 


0.629 


1.469 


1.630 


3W1 






















White 


647 


1.192 


1.129 


1.751 


0.693 


2.517 


0.595 


0.148 


0.789 


1.547 






1.303 


0.965 


1.468 


1.082 


1.696 


0.962 


0.519 


1.082 


1.877 


African- 


74 


0.737 


0.950 


0.964 


0.312 


1.244 


0.223 


0.078 


0.506 


1.044 


American 




1.238 


0.909 


1.340 


0.769 


1.611 


0.589 


0.267 


0.913 


1.517 


Hispanic 


47 


0.481 


0.794 


1.019 


0.386 


1.701 


0.287 


0.062 


0.494 


0.989 






0.811 


0.806 


1.251 


0.817 


1.709 


0.629 


0.318 


0.955 


1.645 



94 



Evidence on Racial and Ethnic Differences 



Table 5-8 (Continued) 

Means and Standard Deviations for Regression Variables 

by Race by Education, Gender, Age — Including 

Respondents without Data or Parental Education 











Art 






Attend 










N 


Lessons 


Home 


music 


Attend 


Visit 


no jazz 


Perform 


Dovis 


TV arts 


Over 51 






















White 


656 


0.660 


0.996 


1.717 


0.488 


1.852 


0.446 


0.065 


0.382 


1.511 






0.979 


1.007 


1.498 


0.961 


1.710 


0.888 


0.313 


0.746 


1.978 


African- 


78 


0.208 


0.535 


0.484 


0.184 


0.604 


0.165 


0.050 


0.172 


0.815 


American 




0.543 


0.706 


1.069 


0.648 


1.062 


0.544 


0.276 


0.492 


1.599 


Hispanic 


26 


0.309 


0.603 


0.945 


0.197 


0.914 


0.115 


0.209 


0.283 


0.828 






0.872 


1.037 


1.309 


0.475 


1.135 


0.319 


0.548 


0.618 


1.752 



but no further education, some college, and at least college graduation. Age 
categories were derived by dividing the population into three groups of 
similar size: 18 to 30, 31 to 51, and older than 51 years of age. 

The educational means must be interpreted with caution, because only 
16 African- American respondents and only 5 Hispanic respondents in the 
November/December sample had 16 or more years of formal education, and 
only 28 African-Americans and 20 Hispanics had attended college for 1 to 
3 years. Differences in means between African- American and white respon- 
dents were smaller (expressed as ratios) among college graduates than 
among other groups with respect to taste for art music, performing-arts 
attendance, museum and exhibition visiting, and nonperformance creative 
activities. By contrast, the gaps between Hispanics and whites in participa- 
tion (again, expressed in ratios) tended to be greater among the more highly 
educated. For example, Hispanics without high-school degrees had higher 
means than their non-Hispanic white counterparts on performing-arts atten- 
dance (including jazz), watching arts television, and participating in per- 
formance and nonperformance production activities. 

Comparisons of intergroup differences by age are also complicated by 
small subsample sizes. Nonetheless, the results are striking (see Table 5-9). 
Comparing mean scores of respondents 52 years of age or over, 31 to 51 
years old, and 18 to 30 years of age, we see that the ratio of white to 
African-American participation rises monotonically with age for lessons and 
classes, art music, televised arts viewing, performing-arts attendance (in- 



95 



Race, Ethnicity, and Participation in the Arts 

Table 5-9 

Ratio of White to African-American and of White to Hispanic 

Weighted Means for Socialization, Taste, and Participation 

Scales, by Age of Respondent (Nov./Dec. 1982 Subsample) 

Ratios, white means: African-American means 





Parental 








Attend 


Age 


guidance 


Lessons 


Art music 


TV arts 


(w/jazz) 


18-30 


1.27 


1.32 


1.50 


0.87 


0.97 


31-51 


1.19 


1.62 > 


1.82 


1.48 


2.23 


52+ 


1.86 
Attend 


3.17 


3.55 


1.85 
Other 


2.65 




(no jazz) 


Exhibits 


Perform 


creative 




18-30 


2.04 


1.56 


0.97 


1.90 




31-51 


2.67 


2.02 


1.90 


1.56 




52+ 


2.70 


3.07 


1.30 


2.22 




Ratios, white means 


: Hispanic means 










Parental 






- 


Attend 


Age 


guidance 


Lessons 


Art music 


TV arts 


(w/jazz) 


18-30 


1.40 


1.81 


0.83 


0.96 


1.87 


31-51 


1.48 


2.48 


1.72 


1.56 


1.80 


52+ 


1.65 
Attend 


2.14 


1.82 


1.82 
Other 


2.48 




(no jazz) 


Exhibits 


Perform 


creative 




18-30 


2.05 


1.34 


0.55 


0.98 




31-51 


2.07 


1.48 


2.39 


1.60 




52+ 


3.88 


2.03 


0.31 


1.35 




Numt)er of respondents 










Age 


White 


African-American 


Hispanic 






18-30 


605 


80 


44 






31-51 


647 


74 


47 






52+ 


656 


76 


26 







96 



Evidence on Racial and Etiinic Differences 

eluding and excluding jazz), and visually oriented activities. Indeed, a 
convergence of African-American and white participation is visible for all 
but performance and nonperformance activities. Among the youngest co- 
hort, African-American means were higher than those of whites for viewing 
art programs on television, attending performances (including jazz), and 
participating onstage and backstage in performances. 

Reduction among age groups of the white/Hispanic ratios are less 
marked than those for whites and African- Americans (perhaps due to higher 
levels of Hispanic immigration), but a monotonic trend appears with respect 
to parental guidance activities, art music, televised arts viewing, performing- 
arts attendance (excluding jazz), and visiting exhibits, museums and related 
activities. Among the youngest cohort, Hispanic means are higher than white 
means for taste for art music, television arts viewing, and both performance 
and other creative activities. 

Do these declining differences reflect changes in the net effects of race 
and ethnicity, or changes in the sociodemographic profiles of African- 
Americans and Hispanic-Americans over the past decades? There is good 
reason to believe the latter is the case, especially changes in levels of formal 
education attained by Hispanics and African- Americans. Among the over- 
51 subsample, the average white respondent had 1 1.25 years of education, 
the average African-American respondent, 7.43, and the average Hispanic, 
6.52. Among the subsample aged 18 to 30, the white average was 12.82, 
while the African-American average had risen to 12.33, and the Hispanic 
average had increased to 11.87. Given the powerful role of education in 
stimulating participation in the arts, we would expect that such relative 
advances for African-Americans and Hispanic-Americans should make 
these groups more similar to whites in patterns of taste and artistic partici- 
pation.^ 



Summary 

White respondents had higher mean scores on all the childhood experi- 
ence, taste, and participation scales than African-Americans and, with the 
exception of performance activities, Hispanic respondents. Intergroup dif- 
ferences were modest because scores for all groups were low, and differ- 
ences were greater for arts consumption than for arts production. 

Because their parents had less formal education than white parents, 
African-American and Hispanic respondents reported receiving fewer kinds 
of parental guidance experiences and taking fewer kinds of arts lessons or 



97 



Race, Ethnicity, and Participation in the Arts 

classes as children and adolescents than white respondents. African- Ameri- 
cans and Hispanics reported about the same number of classes and lessons 
and significantly more parental guidance experiences than whites of compa- 
rable age and family background. 

Hispanics hked art music as much and watched as many kinds of 
televised arts programs as whites with comparable sociodemographic char- 
acteristics, and the television arts viewing habits of African-Americans were 
similar to those of sociodemographically comparable whites. By contrast, 
sociodemographic differences account for only half of the significant ten- 
dency for African- Americans to report enjoying fewer kinds of art music and 
related genres than whites, and differences in socialization explained little of 
the remaining gap. Thus, small but significant differences in musical taste 
are directly related to race. 

The effects of being African-American or Hispanic on participation 
varied depending on whether the activities entailed the consumption or 
production of art and whether the activities involved the performing arts or 
the visual and literary arts. Both Hispanics and African-Americans, on 
average, participate in fewer activities than whites on all three arts consump- 
tion scales. By contrast, there is no significant difference between Hispanic 
and white respondents on either production scale, or between African- 
Americans and whites with respect to onstage or backstage performance 
activities. The gap between African-Americans and whites is wider for the 
visual and literary arts than for the performing arts. 

Despite the gross differences, Hispanic-Americans participate in about 
as many arts-consumption activities as sociodemographically comparable 
non-Hispanic whites. Hispanic -Americans report being involved in more 
production activities (of both kinds) than sociodemographically similar 
white Americans with similar amounts of childhood experience in the arts. 

Significant differences between African-Americans and whites in per- 
forming-arts attendance are also fully accounted for by sociodemographic 
differences between the two groups. By contrast, sociodemographic factors 
explain only about two-fifths of the African-American/white difference in 
visually oriented consumption and production activities. Controlling for 
childhood experience eliminates the significant gap between African- 
Americans and whites with respect to visual-art and literary production, but 
has little effect on African-American/white differences in exhibit atten- 
dance, literature reading, and related activities. The latter difference remains 
significant even after controls for artistic taste and interest are added. 

Taken together, these findings indicate that intergroup differences vary 



98 



Evidence on Racial and Ethnic Differences 



across different kinds of participation, that such differences are largely the 
result of sociodemographic variation between whites, African- Americans, 
and Hispanics, and that such effects of race or ethnicity as remain once 
sociodemographic factors are controlled are small relative to the impact of 
such variables as educational attainment and parental guidance. 

For the most part, childhood experience, taste, and participation mea- 
sures were predicted by the same variables for African-Americans and 
Hispanics as for whites. Two exceptions were notable, however. The first of 
these had to do with the effects of age on parental socialization, musical taste, 
and television arts watching. With parental education controlled, it appears 
that white parents offer fewer guidance experiences than they used to, while 
African-American parents offer more, suggesting that a convergence is 
occurring. Similarly, controlling for other sociodemographic factors, tastes 
for art music and viewing arts programs on TV increased with age for whites, 
but not for African-Americans or Hispanics. (Differences were significant 
except for white/Hispanic TV arts program viewing.) Although these results 
could mean that white Americans' tastes change more with aging than those 
of African- Americans or Hispanic- Americans, it seems more likely to indi- 
cate a convergence of all groups with respect to tastes for art music and 
convergence between African-American and white Americans in artistic 
interests as expressed through watching arts programs on television. These 
findings are consistent with inspection of means by race and age: among 
younger respondents, intergroup differences in parental guidance, taste for 
art music, and television viewing of the arts are smaller than for older 
respondents. 

Second, education had a stronger effect on television viewing of the arts 
and on all of the participation scales except for performance production 
activities for whites than for African-Americans, although in most cases it 
was a significant predictor for both groups. Although the differences were 
not significant, the effects on the participation scales of taking lessons or 
classes in the arts were also weaker for African-Americans than for other 
groups. The same was true of television viewing of the arts, and the differ- 
ences between African-Americans and Hispanics were significant with re- 
spect to nonperformance consumption and production activities. In other 
words, there is some evidence that formal education, both general and 
specific to the arts, is more weakly related to adult interest and participation 
in the arts for African-Americans than for other groups. 

For most participation activities, gaps between white and minority sub- 
populations were greater for older than for younger respondents. The declin- 



99 



Race, Ethnicity, and Participation in the Arts 

ing intergroup differences appear to be the result of changes in the sociode- 
mographic profiles of African- Americans, Hispanics, and white Americans, 
especially rapid increases in the educational attainment of the two former 
groups, rather than of changes in the effects of race on the participation of 
otherwise similar men and women. 



100 



Chapter 6 
CONCLUSIONS 



In Chapter 1 , we called attention to three distinct ways of thinking about 
"underrepresentation" of groups as participants in artistic activities. The 
first focuses on gross differences in rates of participation. In this view, 
any statistical underrepresentation is a matter of public concern. The second 
emphasizes differences in net rates of participation between people who are 
similar in terms of socioeconomic and demographic characteristics other than 
race or ethnicity. In this view, differing rates of participation are of concern 
only if they stem directly from racial or ethnic identity. The third perspective 
asks whether differences in participation, gross or net, result from differences 
in taste or demand between groups or from differences in the degree to which 
groups face different obstacles to participation. In this view, varying partici- 
pation is a concern only if it results from inequality of opportunity to 
participate rather than from differences in taste. 

Which of these perspectives one favors will depend on one's more 
general attitudes towards inequality. It will also depend on one's beliefs 
about artistic participation. If one believes that participation in the arts is 
absolutely essential to an acceptable quahty of Hfe, one is more Hkely to 
believe that gross differences in participation are important. If one believes 
that participation in the arts is essential, but not so important as education, 
income, or a good job, one might be more likely to focus upon net differences 
in participation. If one is not certain whether participation in the arts is 
important for people, one is more likely to focus on equality of opportunity 
and to see no virtue in stimulating demand. 

In this section, we summarize the results of our analyses of the SPPA 
data on participation in selected artistic activities by African- Americans, 
Hispanics, and white Americans.^ We organize our conclusions along the 
lines of the questions raised by the three perspectives noted above: gross 
differences in participation, net differences in participation, and evidence 
bearing on the relative roles of variation in tastes and varying exposure to 
barriers in accounting for the differences observed. 



101 



Race, Ethnicity, and Participation in tlie Arts 

Because patterns of differences among groups vary for different kinds of 
artistic activities and because the SPPA did not ask people about many kinds 
of artistic activities, we can draw no general conclusions about differences 
in artistic participation per se. Thus, we shall call attention to the kinds of 
activities to which specific conclusions do and do not apply. 

The Surveys of Public Participation in the Arts represent the best re- 
source available for investigating the question with which this report is 
concerned. But no survey, especially one designed to address a great many 
different issues, can tell us everything we wish to know. In the final section, 
we set out an agenda of questions that remain, along with some suggestions 
about how such questions might be answered. 



Do Rates of Participation Vary? 

The answer to this question is unambiguous. Rates of participation in 
most of the activities about which the SPPAs asked vary among white, 
African-American, and Hispanic respondents. White Americans participate 
at higher rates than African-Americans or Hispanic-Americans in most of 
these activities that involved attendance at museums, visual art exhibitions, 
and live performances. African-Americans participate at higher rates than 
others, however, as members of jazz audiences. 

Differences in rates of participation between whites, on the one hand, 
and African-Americans and Hispanics on the other, were modest for two 
kinds of active performing: playing a musical instrument onstage and sing- 
ing, dancing, or acting in public. With respect to the former, however, 
differences between whites and African-Americans were greater if only 
public performances of classical music or jazz were considered. Whites were 
also more likely than African-Americans or Hispanics to participate in 
producing visual arts objects such as drawings, paintings, or crafts. For most 
of these activities, rates of participation were somewhat higher for Hispanic 
than for African-American respondents, although the differences between 
white and Hispanic rates exceeded those between Hispanic and Afiican- 
American rates. 

Except for reading imaginative literature, fewer than half of the people 
surveyed participated in any of the activities about which the "core" and 
"other activity" questions of the SPPA asked. With respect to all of the 
activities but reading, visiting art exhibits, visiting science and history 
museums, visiting historical monuments, and doing needlecrafts, fewer than 
20 percent were active. Fewer than 5 percent of respondents attended opera 



102 



Conclusions 



or musical performances, or performed publicly on musical instruments or 
by singing, acting, or dancing. 

Because relatively few people participated, especially in core activities, 
absolute differences in participation rates between groups were often small. 
But absolute differences between participation rates of whites and those of 
African-Americans were .10 or more in both 1982 and 1985 for visiting art 
exhibitions, reading works of imaginative literature, visiting science or 
history museums, visiting historical monuments, attending arts and crafts 
fairs, and engaging in such needlecrafts as sewing or knitting. White rates 
exceeded Hispanic rates by this margin in both years for these same activi- 
ties, except for visiting science or history museums and visiting art exhibits. 

By contrast to the relatively small absolute margins of difference, ratios 
of whites' to others' participation rates were in many cases greater than two 
to one. African-Americans were less than half as likely as whites in both 
1982 and 1985 to work in pottery or other craft media, or to attend classical 
music concerts, opera performances, musicals, plays, arts and crafts fairs, or 
ballet performances. Hispanic respondents were less than half as likely to 
attend plays in both years. 

Thus, there were persistent and substantial gaps in the extent to which 
non-Hispanic white Americans, on the one hand, and African-Americans 
and Hispanic-Americans, on the other, reported participating in the arts 
about which the SPPA asked. African-Americans and Hispanics were less 
likely than whites to be consumers of both the performing arts and the visual 
arts. Both groups were less likely to participate in visual-arts activities. 
Differences between groups were less for performing onstage, particularly 
in popular genres. Differences were not restricted to Euro-American "high 
culture" art forms, however. They also appeared for making crafts, reading 
literature, and visiting historical or scientific museums or exhibits. 

Does Participation Vary Net of 
Sociodemographic Factors? 

Do African-American, Hispanic, and white Americans who are similar 
with respect to such characteristics as gender, age, educational attainment, 
marital status, occupation, family income, and residence in a Standard 
Metropolitan Statistical Area (SMSA) participate at different levels? Here 
the answer is more complicated. 

If we take each of the core activities, one at a time, and control for 
socioeconomic and demographic effects, we find different patterns for 



103 



Race, Ethnicity, and Participation in the Arts 

African- American and Hispanic respondents. For most of the core activities 
in which whites participated significantly more than African-Americans (all 
but jazz attendance and performing in public), between approximately 25 
and 40 percent of the gap resulted from differences in sociodemographic 
position between the races. The remaining margins were statistically signifi- 
cant, but small compared to differences associated with educational attain- 
ment and other background factors. These differences indicate that some 
factor or factors make the probability that African-Americans participate in 
these activities significantly lower than the probability of participation by 
white Americans who are similar with respect to the socioeconomic and 
demographic factors for which we controlled. Nonetheless, policies that 
made African- Americans more similar to whites with respect to educational 
attainment, occupational status, and family income would diminish differ- 
ences in rates of participation for every core activity but jazz attendance. 

Sociodemographic differences between white and Hispanic respondents 
accounted for most of the gross differences between whites and Hispanics in 
attendance at classical music concerts, ballet performances, and art exhibits. 
With such factors controlled, white participation was significantly greater 
than Hispanic participation only for attendance at musical stage perform- 
ances, plays, and (in 1985 only) opera; and for reading imaginative literature 
and (in 1982 only) acting, singing, or dancing onstage. In 1982, Hispanic 
respondents were significantly more likely than comparable non-Hispanic 
whites to attend ballet performances. Because the core activities for which 
significant differences persisted tended to be those that in the United States 
are usually presented in the English language (musicals, plays, literature), 
we speculated that the high proportion of Hispanic-Americans for whom 
Spanish is the native language may have played a role. If this speculation is 
correct, then differences between whites and Hispanics in core participation 
are largely attributable to socioeconomic and linguistic differences. Thus, 
policies that increased the educational attainment, occupational levels, and 
incomes of Hispanic-Americans would eliminate much or all of the signifi- 
cant differences between Hispanics and non-Hispanic whites in participation 
in most of the core activities. Moreover, differences in attendance at plays 
and musicals and differences in reading literature might be moderated by 
increasing the availability of such works in Spanish. 

We also looked at net differences between groups in scores on five scales 
developed with the use of factor analysis. These scales represented the 
number of activities in which respondents participated, rather than the 
probability of participating in a specific activity. Drawing on a smaller 
sample of respondents who were asked a wider range of questions, these 



104 



Conclusions 



analyses looked at scores on five scales: participation as consumers at live 
performances (with and without jazz included); participation as consumers 
of visual materials (art and history museums and exhibits and imaginative 
literature); participation as producers onstage or backstage of performing 
arts events; participation as producers of visual arts and crafts. 

Nearly all of the difference between African-Americans and whites on 
the performing-arts attendance scale (all of it when jazz was included on the 
scale) resulted from sociodemographic differences between members of the 
two racial groups. Once such factors were taken into account, no significant 
differences remained between comparable African-Americans and whites in 
the number of kinds of performing-arts activities they reported attending. 
There was no significant difference between white and African-American 
scores on the performing-arts production scale. 

African-American respondents participated in significantly fewer visual- 
arts and literature consumption and production activities than whites. More- 
over, only about 40 percent of these differences were attributable to the 
socioeconomic and demographic factors for which we controlled. 

In other words, these analyses indicate that one cannot generalize about 
net differences between African-Americans and whites in artistic participa- 
tion. African-Americans are more likely than whites to attend jazz concerts, 
and the margin only increases when sociodemographic differences between 
the races are taken into account. African-Americans are no less likely than 
whites to participate in performing-arts activities as performers or by work- 
ing backstage. African-Americans on average attend fewer kinds of the 
performing-arts activities about which the SPPA asked (excluding jazz) than 
whites, but about the same number as whites who are comparable with 
respect to socioeconomic and demographic characteristics. By contrast, 
African-Americans participate in significantly fewer kinds of visual-arts 
activities than comparable whites, both as consumers and as producers. 

Differences between white and Hispanic respondents can be described 
more succinctly. There were no significant differences between non-His- 
panic whites and Hispanics on either performing-arts or visual-arts produc- 
tion scales. Hispanic respondents scored significantly lower than whites on 
each of the consumption scales, but both of these differences resulted from 
differences between the groups in socioeconomic standing and demographic 
characteristics. In other words, there are no significant differences in any of 
these scales between sociodemographically comparable white and Hispanic 
respondents. 



105 



Race, Ethnicity, and Participation in tlie Arts 

Does Demand for Artistic Participation Vary? 

This question is the hardest to address with the resources provided by the 
SPPAs, and we have reached no definitive conclusions. The best we can do 
is to hold the data up like so many prisms and report the results from a variety 
of angles. 

The SPPAs asked a subsample of respondents whether they liked a wide 
range of musical genres. Within each group, responses were very stable 
between 1982 and 1985. White and Hispanic tastes were quite similar. 
African-American and white respondents' tastes differed more, although, 
like Hispanics and whites, African-Americans tended to prefer commer- 
cially popular genres to most other kinds of music. Larger proportions of 
whites and Hispanics liked country western, rock and easy listening music 
than any other kind of music, whereas African-American respondents were 
most likely to choose hymns/gospel music, soul/blues/rhythm and blues, and 
jazz. Those genres favored by whites and Hispanics ranked fourth, fifth, and 
sixth among Afiican- American respondents, well ahead of the seven other 
genres about which the survey asked. Moreover, substantial minorities of 
whites and Hispanics enjoyed gospel, rhythm and blues, and jazz. Such 
genres as bluegrass, barbershop, and opera were distinctly less popular 
among all three groups. Taken together, the results demonstrate strong 
similarity of tastes between whites and Hispanics, and patterns of musical 
taste for whites and African-Americans that, although different, involve 
different degrees of participation in the same commercially popular musical 
forms rather than sharply opposed preferences. 

Looking more closely at the four kinds of music related to the SPPA core 
participation activities (classical music, opera, show tunes, and jazz), we see 
that differences between African-Americans and whites in taste for classical 
music mirrored differences in African-American and white rates of atten- 
dance at classical music concerts. By contrast, the proportion of Hispanic 
respondents who said they enjoyed classical music was close to that of 
whites in 1982 and greater in 1985. Taken together with the finding that 
Hispanics were about as hkely to attend classical music concerts as sociode- 
mographically comparable white respondents, this pattern suggests that 
Hispanics would attend classical music performances at the same rates as 
whites if they had the resources to do so. 

Similarly, differences between African-Americans and whites with re- 
spect to opera, show tunes, and jazz are comparable to differences in 
attendance at operas, musicals, and jazz performances. So were differences 
between Hispanics and whites for opera and show tunes in 1982, but not in 



106 



Conclusions 

1985, when differences in attendance far exceeded differences in taste. 
Hispanic respondents were more likely than whites to report liking jazz in 
both years, but less likely to report attending jazz concerts. Taken together, 
these results again suggest that disparities between Hispanics and non-His- 
panic whites in attendance at these activities reflect socioeconomic barriers 
rather than differences in taste, whereas differences between African- Ameri- 
cans and whites would appear to be largely accounted for by differences in 
taste. 

This conclusion, however, would conflict with results of analyses pre- 
dicting probabihties of participation in the core attendance activities (other 
than jazz), which showed that between 25 and 40 percent of differences 
between African-Americans and whites were accounted for by differences 
in socioeconomic and demographic factors. In other words, it seems likely 
that some portion of differences in taste are themselves the result of socio- 
economic inequality. Consistent with this interpretation, once sociodemo- 
graphic factors are taken into account, differences in musical taste or in 
childhood socialization explain little of the intergroup variation that remains. 

The SPPAs also asked respondents if they wanted to participate in the 
seven core attendance activities more than they had in the previous year. 
Respondents from all groups who had participated in a given activity in the 
previous year were much more likely than those who had not to wish that 
they had attended more. And respondents who had not participated were 
more likely to wish that they had if they were members of groups that 
participated at relatively high rates. For most activities, the proportion of 
people who did not participate but said that they wanted to exceeded the 
proportion that actually participated. 

What this implies is that if all reported barriers to attendance were 
removed — that is, if everyone who reported wanting to participate, but did 
not, joined the ranks of attenders — the absolute differences in probabilities 
of attendance at core activities between members of different groups would 
increase. The margin between African- American attendance at jazz concerts 
and attendance by whites and Hispanics would become greater, as would the 
margin between white attendance at classical concerts, operas, musicals, 
plays, ballet performances, and art exhibitions, and African-American and 
Hispanic attendance. For many activities, however, the large increase in the 
proportions attending in each group would reduce the ratios of rates of 
participation between groups. 

We caution against taking this finding too seriously for several reasons. 
First, we are not sure what respondents meant when they said they wanted 
to attend more than they did. Second, we suspect that respondents factored 



107 



Race, Ethnicity, and Participation in tlie Arts 

in the cost of attendance in deciding whether they wished to do something 
they had not done, so that respondents facing socioeconomic barriers would 
have been less likely to report "wanting" to attend an event than more 
well-to-do respondents whose taste for the activity in question was similar 
to theirs. Finally, we suspect that many barriers to participation work by 
reducing demand for participation in such activities, rather than by keeping 
people from satisfying demand. 

Indeed, other analyses, including the results on Hispanic musical tastes 
mentioned above, cast doubt upon the degree to which whites do value the 
SPPA core arts more highly than do African- Americans or Hispanics. 
Differences in the extent to which whites, on the one hand, and African- 
Americans and Hispanics, on the other, watch the core attendance activities 
(other than jazz) on television are not so great as differences in the extent of 
live participation. This may suggest that when cost is not a factor (because 
most Americans have access to televised arts programs), intergroup rates of 
participation are more comparable than when participation is more costly 
and time-consuming. 

Moreover, parents of Hispanic and African-Americans appeared to value 
certain kinds of artistic guidance even more highly than comparable white 
parents. When age, gender, and educational attainment of parents are con- 
trolled, Hispanic and African-American respondents reported significantly 
if modestly higher scores on a parental guidance scale comprising being 
encouraged to read, being taken to museums, being exposed to classical 
music, and being taken to live performances. 

Taken individually, the results of the analyses described in this section 
point in somewhat different directions. Taken together they suggest that the 
issue of motivation is extremely complex. On the one hand, participation in 
the artistic activities for which intergroup differences appear is not some- 
thing that everyone clearly desires. For example, eliminating all barriers to 
attendance at jazz concerts or ballet or opera (by providing free vouchers, 
transportation, and babysitting), would seem unlikely to eliminate differ- 
ences between African-Americans and whites in rates of attendance. On the 
other hand, it would be simplistic, and at odds with many of our other 
findings, to suggest that African-Americans and Hispanics attend certain 
activities less than whites simply because they Hke them less. Rather, 
differences in participation rates appear to result, in part, from differences in 
socioeconomic opportunity, in part from differences in taste, and in part 
from the interaction of these two factors. 



108 



Conclusions 

Summary Conclusions 

1 . Rates of participation in the activities about which the SPPA asked differ 
by race and ethnicity. White rates are greatest for almost all these 
activities (with the notable exception of those associated with jazz, for 
which African-American rates are highest). In general, differences are 
greater for attendance and reading than for viewing the arts on television, 
parental guidance into the arts through home activities and (for African- 
Americans) taking formal classes and lessons, participating in most art 
producing activities, and (for Hispanics) musical tastes. For most activi- 
ties, absolute differences are relatively small (with minorities of any 
group participating), although ratios of white to other rates are often as 
high as two to one. 

2. African-Americans participate somewhat less than sociodemographi- 
cally comparable white Americans in most of the core activities, but most 
of these net differences are small. Net differences between African- 
Americans and whites are more marked for the visual arts than for the 
performing arts. African-Americans are significantly more likely than 
comparable whites to attend jazz concerts. 

3. Hispanic-Americans participate somewhat less than comparable whites 
in some core activities, especially those usually presented in the English 
language. In general, however, Hispanic-Americans participate at rates 
similar to those of socioeconomically comparable white Americans. 

4. Differences in participation associated with race are very small com- 
pared to those associated with educational attainment and are usually 
exceeded by those associated with income, occupational status, and 
gender. The principal barriers to participation for African-Americans 
and, especially, Hispanics, are socioeconomic. These barriers reduce 
minority participation by influencing demand for the arts and by making 
it difficult for less well-off Americans to satisfy their demand. 

5 . Measurable differences in taste or in parental guidance into the arts, other 
than those associated with differences in socioeconomic standing, play 
a small role in explaining most of the observed differences in participa- 
tion between African-American, Hispanic, and white Americans. Much 
of the observed differences in taste, demand, or childhood experience 
appears to result from socioeconomic differences between these groups. 

6. Intergroup differences in participation in most of the activities about 
which the SPPA asked are smaller for younger than for older respon- 
dents. Most of this apparent decline in the participation gap is probably 



109 



Race, Ethnicity, and Participation in the Arts 



the result of increases in socioeconomic resources, especially years of 
schooling, of African-American and Hispanic respondents. 



Further Research 

We have emphasized throughout this report that one cannot generalize 
meaningfully about "artistic participation." Patterns of differences between 
non-Hispanic whites, on the one hand, and African- Americans and Hispan- 
ics, on the other, vary among activities. 

The SPPA questions focused upon categories of participation that a 
pretest indicated were widely understood by all or most people interviewed. 
The requirements of a national survey tended to exclude such forms as 
mariachi music or clog dancing that are unfamiliar to most Americans, 
including many forms with roots in specific ethnic or racial communities. 
The questions also tended to focus on consumption activities associated with 
nonprofit cultural institutions rather than on the most widely consumed 
forms of popular culture. We suspect that there are many activities (like jazz) 
for which white participation is lower than African-American. Conse- 
quently, we would not generalize the findings of this report beyond the 
specific kinds of activities that the SPPA considered. Because the SPPA 
items cover a broad range of activities, including ones with which public 
policy has been particularly concerned, we do not regard this as a serious 
problem given the purposes of this report. But it does mean that it would be 
a mistake to consider this a full treatment of all aspects of the artistic 
participation of African-Americans, Hispanics, and white Americans. 

Even within the scope of artistic participation as defined by the SPPA, 
this report could not address a number of questions that are of substantial 
interest. The limitations we describe below are natural ones for a national 
survey that was not designed specifically to address racial and ethnic differ- 
ences in participation. Nonetheless, without questioning the importance of 
what the SPPAs have already accomplished, it may be useful to sketch a few 
tasks that remain. 

Each of the groups we examined is heterogeneous. The SPPA data did 
not permit close analysis of participation by ethnic subgroups, in part 
because the number of Hispanic respondents (other than Mexican-Ameri- 
cans) was relatively small, in part because of the way in which ethnic 
background was coded on the SPPA. In particular, it is important from the 
standpoint of public policy to distinguish among the ethnic groups that 
constitute the Hispanic and Asian categories, between native-bom and 



110 



Conclusions 



immigrant {e.g., West Indian) African-Americans, and between Native 
Americans and other respondents. A design that stratified the sample on 
ethnicity and oversampled these groups relative to their percentage of the 
population, and a coding scheme that distinguished more clearly among 
groups would be helpful in this respect. 

To mask the identity of respondents, the Census Bureau did not include 
regional data on the SPPA files. Broad regional classifications would permit 
investigation of regional differences (which may interact with race or ethnic- 
ity) without breaching confidentiality. 

Especially for those groups for which recent immigration rates have been 
relatively high (Hispanics and Asians), it is important to be able to assess the 
impact of native language on participation in activities relying on the spoken 
or written word. A question on language should be included in subsequent 
surveys. 

Even if we had found no differences in the rates of participation of 
African-Americans, Hispanics, and white Americans in the activities that the 
SPPA described, we could not have concluded that these groups participated 
in the same way. Do African-Americans, Hispanics, and whites who attend 
theater, for example, see the same kind of plays at the same kind of venues? 
If attenders vary by race or ethnic origin in the kinds of activities they prefer, 
we might be able to assess the extent to which different rates of participation 
result from the undersupply of the kinds of activity preferred by members of 
the groups that participate less. Such questions could perhaps be addressed 
best in local surveys that would ask respondents about attendance at specific 
events or specific institutions. We suspect that such questions would reveal 
greater intergroup diversity than questions phrased in more general terms. 

We found some suggestive evidence of differences in the effects of 
gender, education, occupation, and artistic socialization on participation by 
members of different groups, but could explore them only superficially due 
to the relatively small number of African-American, Hispanic, and espe- 
cially Asian respondents. Particularly with respect to activities in which 
relatively few people participate, cell sizes {e.g., for college-educated His- 
panic opera attenders) were very small. A research design that permitted 
oversampling of minority respondents (relative to their share of the popula- 
tion) would alleviate this problem to some degree. 

Although one can make rough inferences about change over time on the 
basis of cohort analysis, as we have attempted to do in this report, confident 
conclusions require data collected over a wide range of time. The 1982 and 
1985 SPPAs represent an excellent first step in this process. There can be no 
substitute for the routine collection of comparable data at regular intervals. 



111 



Race, Ethnicity, and Participation in the Arts 

How widespread is participation in the activities about which the SPPA 
could not ask because they were not sufficiently familiar to the average 
American? Some of these activities, such as reggae music or Balkan folk 
dance, might be characterized by racially or ethnically homogeneous audi- 
ences, even though the proportion of persons in the relevant ethnic groups 
who participate would be low. Such art forms may add a great deal to our 
national cultures and to specific artistic subcultures even if they are partici- 
pated in by too few persons to catch in a national sample survey. Regional 
or SMS A-based surveys might be able to explore the distribution of partici- 
pation in such activities more effectively. 

The one SPPA question that provided information about taste for large- 
scale commercial popular-culture genres, the music preference question, 
revealed patterns of racial and ethnic cleavage and convergence that were 
not apparent in responses to questions about other activities. Because much, 
probably most, of the arts that Americans consume are provided by the 
national popular-culture industries, a comprehensive analysis of differences 
in artistic participation would require attention to participation in popular 
culture broadly defined. 

The foregoing is a wish list and, as such, is unrestricted by the costs, 
multiple priorities, and tradeoffs that constrain actual research decisions. 
Some of the suggestions offered above will be impractical, while others may 
not. In conclusion, we offer the following recommendations: 

• That information on region of residence be included in the SPPA data 
file. 

• That information on native language be collected and included in the 
SPPA data file. 

• That the Arts Endowment explore the possibility of a design for the 
SPPA that oversamples African-American, Hispanic, Asian, and 
Native American respondents relative to their share of the population. 

• That ethnicity codes comparable to those provided in the Census of 
Population, including multiple ethnic origins, be collected for the 
SPPA. 

• That the Arts Endowment or other research sponsors explore the 
possibility of supporting comparable local surveys of participation in 
the arts in several regions that include questions about attendance at 
specific events or specific institutions. 



112 



NOTES 



Executive Summary 

1 . For the sake of simplicity, we drop the modifier "non-Hispanic" when 
referring to whites and African-Americans throughout this report. The 
reader should recognize that this modifier is implicit. 

2. Throughout this report, "childhood sociaHzation" or "socialization" refer 
to both forms of early experience: "parental guidance" and "lessons and 
classes." 

3. SMS A residence refers to whether respondents lived in an urban area, 
an area within an SMSA but not a city (ordinarily, a suburb), or an area 
outside a center of concentrated population. SMSA stands for Standard 
Metropolitan Statistical Area. 



Chapter One 

1. See, e.g., Richard Bach, The Place of the Arts in American Life (New 
York: The Carnegie Corporation, 1924); and Frederick P. Keppel and 
Robert L. Duffus, The Arts in American Life (New York: McGraw Hill, 
1933); Melvin E. Haggerty, Art as a Way of Life (Minneapolis: Univer- 
sity of Minnesota Press, 1935). 

2. A detailed technical description of the procedures for the 1982 SPPA 
(which were similar to those for the 1985 survey) is available in John P. 
Robinson, Carol A. Keegan, Terry Hanford, and Timothy A. Triplett, 
Public Participation in the Arts: Final Report on the 1982 Survey, 
October 1985 report to the Research Division, National Endowment for 
the Arts. Background information on the 1985 survey is available in 
Timothy A. Triplett and Jeffrey M. Holland, Public Participation in the 
Arts: The 1982 and 1985 Users Manual (draft, October 1985), report 
to the Research Division, National Endowment for the Arts. 



113 



Race, Ethnicity, and Participation in tlie Arts 

3. The text of the survey is available from the National Endowment for the 
Arts, Research Division. Because only tiny percentages engaged in any 
given activity more than once in the month preceding the survey, only 
data on participation during the previous year are analyzed in this report. 

4. The survey was administered each month for 12 months in 1982, and 6 
months in 1985, with all but the core questions rotated from month to 
month. (All questions were asked in the final two months of 1982.) 
Consequently, analyses of responses to all but the core questions are 
based on only a portion of the total number of respondents. Because 
responses were weighted to be representative of the non-institutionalized 
population over 18 for each month, as well as for each of the two years, 
findings are equally generalizable. 

5. William L. Yancey, Eugene P. Ericksen, and Richard Juliani, "Emergent 
Ethnicity: A Review and Reformulation," American Sociological Re- 
view 4\ (1976): 391-^03; Susan Olzak, "Contemporary Ethnic Mobili- 
zation " Annual Review of Sociology 9 (1983): 355-74. 

6. Ninety-five percent of respondents to the 1982 SPPA who reported their 
race as Black reported their ethnicity as Afro-American or Negro. In 
1 985, the figure was 92 percent. Each year, most other Black respondents 
reported their ethnicity as "other," a category that would have included 
such Caribbean ethnicities as Jamaican or Haitian. In each year more 
than 99 percent of respondents who reported their ethnicity as Afro- 
American or Negro reported their race as Black. We use "African- Ameri- 
can," which has come into common usage since the surveys were 
conducted. 

7. The "Other" category in 1982 consists of Asian-Americans, American 
Indians, and persons who failed to choose one of several races from a 
set presented by the interviewer. In 1985, it excluded Asian- Americans 
and included only a very small number of respondents. We do not report 
results for the "Other" category. Because of its heterogeneity and be- 
cause we do not have data on its composition, such results could not be 
interpreted. Unfortunately, then, the data do not permit us to describe the 
artistic participation of Native Americans. 

8 . Our aggregation of ethnic categories into a broader Hispanic group yields 
a category consistent with the federal government's definition of His- 
panic as "A person of Mexican, Puerto Rican, Cuban, Central or South 
American or other Spanish culture or origin, regardless of race" (0MB 
Directive Number 15, as revised May 12, 1977). 

9. By contrast, in the 1980 Census (the report cited in the note at the bottom 
of Table 1-2), 41 percent of Hispanic respondents reported their race as 



114 



Notes 



something other than Black or White. The difference is a result of 
different question phrasing: SPPA respondents were asked to designate 
their race and given a brief set of options that did not include "other." 
Census respondents, by contrast, were asked to choose from among a 
longer list that included both racial and ethnic categories. We are grateful 
to Carmen DeNavas, Helen Montagliani, and Robert Tinari of the Bureau 
of the Census for explaining the manner in which the Bureau asked about 
race and ethnicity in its interviews. 
10. As Table 1-1 indicates, the percentages of Hispanic- Americans in the 
SPPA samples are close to, but not identical with, the proportion of these 
groups in the American population. Furthermore, a review of census data 
reveals that 1985 SPPA Hispanic-American respondents were typical of 
their specific ethnic groups with respect to educational attainment, a 
variable of particular interest in the analysis of arts participation. Census 
data are from U.S. Bureau of the Census, Current Population Reports, 
Persons of Spanish Origin in the United States: March 1985 (Advance 
Report), December 1985 Series P-20, No. 403, p. 4. For insightful 
criticism of the "Hispanic" category, see David E. Hayes Bautista and 
Jorge Chapa, "Latino terminology: Conceptual bases for standardized 
itrrmnology,'' American Journal of Public Health 11 (1987): 61-68; for 
a pragmatic defense, see Fernando M. Trevino, "Standardized terminol- 
ogy for Hispanic populations," pp. 69-72 in the same issue. 



Chapter Two 

1 . A second reason for focusing on the annual rather than the monthly rates 
is the evidence reported by John Robinson and his colleagues that 
respondents' recollections "telescoped" their annual attendance into the 
previous month, thus making the monthly estimates less reliable than the 
annual ones. See John P. Robinson, Carol A. Keegan, Terry Hanford, 
and Timothy A. Triplett, Public Participation in the Arts: Final Report 
on the 1982 Survey, Report of the Research Division of the National 
Endowment for the Arts, October, 1985, pp. 227-29. 

2. Paul DiMaggio, Michael Useem, and Paula Brown, Audience Studies of 
the Performing Arts andMuseums: A Critical Review, Research Division 
Report #9 (Washington, D.C.: National Endowment for the Arts, 1978), 
pp. 37-38. 

3. For the sake of simplicity, we drop the modifier "non-Hispanic" when 
referring to whites and African-Americans throughout this report. The 



115 



Race, Ethnicity, and Participation in the Arts 

reader should recognize that this modifier is impHcit in the remainder of 
the report. 

4. For a careful analysis of responses to the SPPA questions about jazz, see 
Harold Horowitz, The American Jazz Music Audience (Washington, 
D.C.: National Jazz Service Organization, 1986). 

5. We know that people who watch an art form on television are more likely 
than those who do not to attend it in person; but, without data over time 
on the same people, we cannot ascertain whether this is the case because 
television viewing leads to attendance, because attenders are more likely 
to watch arts programming on television, or because attending live events 
and watching the arts on television are caused by some third set of factors. 

6. Pierre Bourdieu, Distinction (Cambridge: Harvard University Press, 
1984); Paul DiMaggio, "Cultural capital and school success: The impact 
of status culture participation on the grades of U.S. high school students," 
American Sociological Review 47 ( 1 982) : 1 89-20 1 ; and Paul DiMaggio 
and John Mohr, "Cultural capital, educational attainment, and marital 
selection," American Journal of Sociology 90 (1985): 1231-61. 

7. John P. Robinson et al., Public Participation in the Arts, p. 368. 

8. The 1 985 figures showed an increase for whites, African-Americans, and 
Hispanics in the extent to which parents listened to classical music while 
the respondents were growing up. Because the question referred to 
previous parental behavior, which by definition could not have changed 
between 1982 and 1985, as opposed to respondent behavior (which could 
have), we do not regard these increases as meaningful ones. None of the 
differences are statistically significant. Moreover, because the sample 
was not designed to be representative of Hispanic-Americans, it is 
possible that some portion of the difference for that group, which is the 
largest, is an artifact of sample composition changes. Because the 
number of respondents in 1982 was substantially greater than that in 
1985, we place more confidence in the results for the earlier year. 

9. Fewer than 10 percent of respondents in any group in either year reported 
that their parents "often" Ustened to classical music, though whites were 
somewhat more likely than members of other groups to give this re- 
sponse. 

10. Fewer than 5 percent of any group in either year reported that their 
parents "often" took them to art museums or galleries. 

11. White respondents were more likely than others to report that their 
parents "often" took them to such events, but fewer than 6 percent of any 
group in either year reported this frequency of attendance. 

12. The single exception: in 1982, 7.31 percent of the Hispanic respondents 



116 



Notes 



as compared to 6.92 percent of the African-Americans reported having 
taken acting lessons. 

13. It may be that Hispanic-Americans have gone to schools where fewer 
arts courses have been offered or required; that they are less likely than 
African- Americans or whites to take optional arts courses; or that they 
have in some way been excluded from courses that were available to 
African- Americans or whites. Note, however, that a far higher propor- 
tion of Hispanic-Americans than of whites or African-Americans are 
immigrants who received their schooling outside of the United States. 
Unfortunately, data on where respondents were bom are not available in 
the SPPA. 

14. To pursue this issue further, we compared the percentages of African- 
American, Hispanic, and white respondents participating in the core 
consumption items among respondents who did and did not take lessons 
relevant to each item before the age of 18. (For example, we compared 
attendance at classical music concerts by respondents who took music 
appreciation courses to the attendance by those who did not take such 
courses. See Appendix tables 2-4 and 2-5.) As expected, persons in each 
group who had taken relevant classes or lessons participated in most 
activities at higher levels than others. In 1982, the difference between 
respondents with and without lessons was greatest for most activities for 
African-Americans and Hispanics, and the odds ratio of participation 
between whites and other groups was in most cases lower among persons 
who had taken lessons or classes, suggesting the possibility that formal 
instruction tends to depress intergroup differences. (Alternatively, Afri- 
can-Americans and Hispanics who reported taking lessons or classes in 
their youth in 1982 may simply have had more of other characteristics 
that are associated with attendance than did whites who reported having 
taken lessons.) These differences were not so apparent, however, among 
respondents to the 1985 SPPA. Because the number of African- Ameri- 
can and Hispanic respondents is greater in the 1982 SPPA, we have more 
confidence in those data. But given the discrepancy in results between 
the two years, it would be incautious to regard the 1982 patterns as any 
more than bases for hypotheses for further research. 

Chapter Three 

1 . We ask the same question about Asian-Americans, but there are so few 
Asian-Americans in the sample that we answer it with less confidence 
than for the other groups. 



117 



Race, Ethnicity, and Participation in tlie Arts 

2. In Chapter 5, we shall return to other forms of participation and explore 
similar questions using different methods and a wider range of variables. 

3. The following description of our approach will be somewhat tedious for 
the reader unfamiliar with statistical analyses of the type reported here; 
but reading it is necessary if one is to interpret the tables in this chapter. 

4. For a fuller description, see John H. Aldrich and Forrest D. Nelson, 
Linear Probability, Logit, and Pr obit Models (Beverly Hills, California: 
Sage Publications, 1984). We used the LOGIST procedure provided by 
SAS (a statistical package) and developed by Frank E. Harrell, Jr. 

5. Education and income, which were categorized in the survey, were 
recoded to their natural metric (using midpoints of categories where 
appropriate). Because of changes in the federal occupational classifica- 
tion system between the 1982 and 1985 surveys, the occupational 
classifications were somewhat different, although occupations were 
aggregated to maximize comparability between the two years. For 
residence, the omitted category was "outside SMSA." For occupation, 
the omitted category in 1982 was "craft, operative, service, farm, trans- 
port, laborers, private household, and armed forces"; in 1985, it was 
"craft, operative, service, farm, armed forces." For marital status, the 
omitted category was "married." 

6. Because logistic regression analyses cannot generate standardized re- 
gression coefficients, comparison of effects is less straightforward than 
for ordinary least squares regression analysis of the kind used in Chapter 
5. We compare the magnitude of effects by comparing the R statistic for 
specific independent variables. The R statistic measures the net contri- 
bution of each predictor to the model's total explanatory power. 

7. William Julius Wilson, The Declining Significance of Race: Blacks and 
Changing American Institutions, 2nd ed. (Chicago: University of Chi- 
cago Press, 1980); Stanley Lieberson, A Piece of the Pie: Blacks and 
White Immigrants Since 1880 (Berkeley: University of California Press, 
1980). 

8. The results of these analyses are presented in Appendix Table 3-1. Note 
that the coefficient for each independent variable represents the effect of 
that variable on participation among members of the group in question. 

9. By "more likely," we mean "more likely after controlling for other 
sociodemographic differences between white men and white women." 
Unless otherwise specified, all comparative statements in this chapter 
refer to net differences after the inclusion of sociodemographic controls. 



118 



Notes 



Chapter Four 

1 . Questions on unsated demand for and barriers preventing the other three 
core activities (two kinds of pubHc performance and reading imaginative 
literature) were not included in the SPPA. 

2. See Roger A. McCain, "Reflections on the Cultivation of Ta.siQ,'' Journal 
of Cultural Economics 3, 1 (1979): pp. 30-52; and "Game Theory and 
CuiiivaXionofTasiQ,'' Journal of Cultural Economics 10, 1 (1986): 1-16. 

3. John Robinson, Public Participation in the Arts: A Project Summary 
(College Park, Maryland: University of Maryland Survey Research 
Center, 1985), pp. 2-3# 

4. Note that there is nothing circular about this argument. In the case of 
most other goods, demand is greater among those with less. If I do not 
have a washing machine, I am Hkely to want one. Once I have one, I will 
not need another until the one I have breaks down. Similarly, my demand 
for breakfast is higher before rather than after I have eaten. The arts may 
be different. 

5. See, e.g., David Featherman and Robert Hauser, "Changes in Socio- 
economic Stratification of the Races, 1962-1973,'' American Journal of 
Sociology 82 (1976): 621-51. 

6. The 1982 SPPA data set contained several precoded "Other" responses, 
none of which was chosen by even 10 percent of the would-be attenders. 

7. Results for any group are not reported when the base number of respon- 
dents — those "wanting more" of something — is smaller than 10. 
Results for a given reason are not reported when fewer than 10 percent 
of any group in either year marked that reason as applicable. 

8. In 1982, none of the percentages for Hispanic-Americans is based on 
more than 48 respondents and in 1985, none is based on more than 21. 
Ns for African- Americans ranged from 23 (for opera) to 113 (for jazz 
performance) in 1982; and from 10 (opera) to 59 (jazz) in 1985. 

9. Can we assume that underlying valuations of arts attendance are the same 
for all three groups? The answer to this question is not obvious. The most 
cautious assumption is that the underlying distributions of value that 
African-American, Hispanic, and white respondents place on the activi- 
ties in which they report wanting to take part are basically similar. On 
the one hand, we have seen in chapter 3 that after controlling for measures 
of educational and economic resources (which can be interpreted as 
measures of economic barriers to participation), African-Americans are 
more likely to attend jazz performances and less likely to attend the other 
activities than white Americans, whereas Hispanic-Americans attend 



119 



Race, Ethnicity, and Participation in the Arts 

most activities at levels not significantly different from those of white 
Americans. A rough inference from these results would be that the 
average African- American values jazz more highly and the other activi- 
ties less highly than the average white American, and that white and 
Hispanic-Americans value them more or less to the same degree. But the 
figures in this section are based not on average African-American, 
Hispanic, or white respondents but on those who did not attend but said 
that they wanted to do so. Such persons seem likely to value the arts more 
than their peers who neither participated nor wish to participate; and, if 
the speculations about social-desirability bias set out earlier in this 
chapter are correct, such bias may be less important for African- Ameri- 
cans and Hispanics than for whites. Consistent with this hypothesis, 
white respondents tended to give such reasons as procrastination, a lack 
of motivation, or a preference for watching television (each of which we 
regard as evidence of a relatively low valuation of the activity in 
question) more frequently than African-Americans or Hispanics. How- 
ever, the differences are small and inconsistent. 



Chapter Five 

1 . Non-Hispanic respondents whose race was coded as "Other" (including 
Asian-Americans, Native Americans, and those not classifiable) were 
removed from the sample. (There were too few of these respondents for 
most of our purposes and, in any case, the heterogeneity of the category 
would have made any results uninterpretable.) A few respondents for 
whom data on key variables were missing were likewise eliminated. 
Respondents classified as "Other" were not included in these analyses. 

2. Although we undertook all of the analyses reported below on both the 
full November/December subsample and the partial subsample (of re- 
spondents reporting data on parental education), in most cases we report 
only the results from the full subsample, because of the nonresponse bias 
problem. For most taste and participation outcomes, parental education 
exerts a small positive influence by virtue of its causal relationship to the 
two socialization measures, which are positively related to participation. 
In other words, because it seems that more educated parents lead their 
children to participate more in the arts as adults because they help them 
have more artistic experiences in their youth (rather than through some 
other means not measured by the socialization scores), we can use the 



120 



Notes 



more reliable full sample without fear that including measures of parental 
education would alter our results. 

3. In Chapter 2, we raised the question of whether the lesser gross differ- 
ences between African-American and white respondents in television 
viewing than in live attendance were the result of the fact that African- 
Americans also watched more television, in general, than whites. To 
explore this possibility, we controlled for hours of television viewing of 
all kinds. Although people who watch a great deal of television in general 
also watch significantly more arts television than people who do not, the 
effect is very small and does not explain the relatively high levels of arts 
viewing among African-American respondents. See Appendix Table 5-6 
for the full model. 

4. The full models are displayed in Appendix Tables 5-6 through 5-10. 

5. This finding was unexpected for the performance attendance scale that 
excluded jazz attendance, which African-American respondents re- 
ported at higher rates that whites, because the logistic regression analyses 
reported in Chapter 3 revealed that African-American respondents were 
less likely to have attended most of the activities included in the perform- 
ance attendance scale even after controlling for sociodemographic fac- 
tors. But although they were statistically significant, these differences 
were small. The apparent difference stems from the difference in sizes 
between the full sample and the November/December subsample. Be- 
cause the latter is smaller than the former, effects are less likely to be 
statistically significant. To confirm this, we reran logistic models using 
only November/December data. Although the coefficients for race were 
comparable in magnitude to those for the full sample, once sociodemog- 
raphic controls were added, the effects of race on attendance at perform- 
ing-arts events (other than jazz) were not statistically significant. We also 
considered and ruled out three altemative explanations for the apparent 
disparity in results. First, we asked if they resulted from systematic 
differences between the November/December subsample and the sample 
for 1982 as a whole. But, as Table 5-1 indicates, African- American/white 
differences in the likelihood of attendance at core performing-arts activi- 
ties were about as large for the November/December subsample as for 
the 1982 sample as a whole. Moreover, regression analyses to predict 
the performing-arts attendance scales using the full sample (Appendix 
Table 5-5) yielded results that were substantively the same as those from 
the November/December subsample (although the large size of the full 
sample made the tiny effect of race statistically significant). Second, we 



121 



Race, Ethnicity, and Participation in the Arts 

considered the possibility that racial effects might have been altered 
because a somewhat shorter list of control variables was employed in the 
analyses in Chapter 5 than in the analyses in Chapter 3 (due to the 
merging of several occupational, marital, and residential categories). If 
anything, however, this would have magnified the effects of race by 
eliminating variation in control variables with which both race and 
participation are correlated. Third, we considered the possibility that the 
logarithmic form used in the logistic regression analyses in Chapter 3 
better represented the relationship between race and participation than 
the linear models reported above. To test this possibility, we ran the 
models using the logarithmic form of the attendance scale as our depend- 
ent variable, and discovered that this transformation made no substantive 
difference to the results. Having eliminated these three alternative expla- 
nations, we feel confident in attributing the difference to the smaller size 
of the November/December subsample. Because a sample of 2,255 (the 
size of the November/December subsample) is sufficiently large that no 
substantively important effect could be deemed insignificant, we are 
satisfied with the reliability of these findings. 
6. For more detailed analyses of intercohort differences in the relationship 
between race and participation, see Paul DiMaggio and Francie Os- 
trower, "Participation in the Arts by Black and White Americans," Social 
Forces 6S(\990): 7 53-7S. 



Chapter Six 

1. We do not include Asian-Americans in this summary because the 
SPPA's information on this group was so limited. 



122 



Appendix 



Appendix Table 2-1 
Z Scores for Musical Tastes by Racial/Ethnic Group and Year 

Whites African-Americans Hispanics 



1982 1985 1982 1985 1982 1985 

Classical -0.14 -0.44 -0.44 -0.68 0.18 0.13 

Opera -1.55 -1.68 -0.96 -0.92 -1.36 -1.33 

Show tunes -0.43 -0.47 -0.62 -0.70 -0.59 -0.43 

Jazz -0.51 -0.20 0.95 1.27 0.26 0.84 

Soul/blues -0.62 -0.30 1.87 1.90 0.42 0.37 

Big band 0.31 0.20 -0.30 -0.32 0.05 -0.55 

C&W 2.38 1.95 0.00 -0.05 2.00 1.63 

Bluegrass -0.23 -0.41 -0.99 -1.10 -1.05 -0.94 

Rock 0.39 0.82 0.26 0.16 1.09 1.49 



Easy listening 


1.55 


1.74 


0.02 


0.63 


1.31 


1.17 


Folk 


-0.25 


-0.42 


-0.80 


-0.64 


-0.40 


-0.66 


Barbershop 


-1.09 


-1.20 


-1.01 


-1.10 


-1.39 


-1.51 


Hymns/gospel 


0.21 


0.43 


2.04 


1.58 


-0.52 


-0.20 



123 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 2-2 

Percentages Participating in Core Activities by Respondents Who 

Did and Did Not Watch Such Events on Television, 1982 SPPA 



Core Item: 




Jazz 






Classical 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


5.64 


29.36 


5.21 


6.56 


31.68 


4.83 


(N) 


(2,733) 


(550) 




(2,421) 


(865) 




A 


10.70 


28.39 


2.65 


3.66 


12.92 


3.53 


(N) 


(269) 


(97) > 




(308) 


(58) 




H 


4.41 


4.66 


1.06 


1.76 


20.49 


11.64 


(N) 


(170) 


(33) 




(160) 


(43) 




W/Aodds 


0.53 


1.03 




1.79 


2.45 




W/H odds 


1.28 


6.30 




3.73 


1.55 




Core item: 




Opera 






Musical ttieater 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


1.39 


11.22 


8.07 


14.94 


42.46 


2.84 


(N) 


(2,863) 


(423) 




(2,584) 


(694) 




A 


0.32 


2.83 


8.84 


5.17 


27.22 


5.26 


(N) 


(333) 


^33) 




(303) 


(63) 




H 


0.48 


9.57 


19.93 


7.68 


35.58 


4.63 


(N) 


(183) 


(20) 




(167) 


(36) 




W/Aodds 


4.34 


3.96 




2.89 


1.56 




W/H odds 


2.90 


1.17 




1.95 


1.19 




Core Item: 




Plays 






Ballet 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


7.27 


25.98 


3.57 


2.49 


16.03 


6.44 


(N) 


(2.375) 


(909) 




(2,712) 


(566) 




A 


1.83 


18.14 


9.91 


0.51 


2.87 


5.63 


(N) 


(302) 


(64) 




(327) 


(38) 




H 


4.37 


8.37 


1.92 


0.49 


17.14 


34.98 


(N) 


(173) 


(30) 




(172) 


(31) 




W/AOdds 


3.97 


1.43 




4.88 


5.59 




W/H Odds 


1.66 


3.10 




5.08 


0.94 





124 



Appendix 

Appendix Table 2-2 (Continued) 

Percentages Participating in Core Activities by Respondents Who 

Did and Did Not Watch Such Events on Television, 1982 SPPA 



Core item: 




Art 






No 


Yes 


Y/N 


W 


16.13 


47.93 


2.97 


(N) 


(2,492) 


(781) 




A 


9.14 


26.64 


2.91 


(N) 


(299) 


(67) 




H 


11.16 


44.87 


4.02 


(N) 


(169) 


(33) 




W/AOdds 


1.76 


1.80 




W/H Odds 


1.45 


1.07 





Ns unweighted, percentages weighted. "Yes" refers to respondents who watched relevant television 
programs, "No" to those who did not. Y/N=probability of participation for persons who watched 
programs to those who did not. W/A Odds=probability of participation for whites/probability of 
participation for African-Americans.W/H Odds=probability of participation for whites/probability for 
Hispanics. 



125 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 2-3 

Percentages Participating in Core Activities by Respondents Who 

Did and Did Not Watch Such Events on Television, 1985 SPPA 



Core Item: Jazz Classical 

No Yes Y/N No Yes Y/N 



w 


6.05 


25.59 


4.23 


6.00 


34.55 


5.76 


(N) 


(1,443) 


(257) 




(1.263) 


(446) 




A 


2.81 


34.27 


12.19 


4.52 


34.25 


7.58 


(N) 


(116) 


(69) 




(146) 


(41) 




H 


2.98 


24.86 


8.34 


2.32 


28.98 


12.49 


(N) 


(105) 


(18) 




(101) 


(23) 




W/Aodds 


2.15 


.75 




1.33 


1.01 




W/Hodds 


2.03 


1.03 




2.59 


1.19 




Core Item: 




Opera 






Musical theater 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


1.33 


12.02 


9.04 


13.44 


42.07 


3.13 


(N) 


(1,474) 


(234) 




(1,394) 


(313) 




A 


0.00 


11.68 


NA 


4.89 


30.96 


6.33 


(N) 


(169) 


(18) 




(151) 


(35) 




H 


0.82 


6.41 , 


7.82 


4.86 


28.07 


5.78 


(N) 


(108) 


(15) 




(104) 


(20) 




W/Aodds 


NA 


1.03 




2.75 


1.36 




W/H odds 


1.62 


1.88 




2.77 


1.50 




Core Item: 




Plays 






Ballet 






No 


Yes 


Y/N 


No 


Yes 


Y/N 



W 6.65 28.16 4.23 2.97 16.99 5.72 

(N) (1,299) (406) (1,431) (276) 

A 4.00 14.30 3.58 0.00 14.30 NA 

(N) (150) (36) (155) (32) 

H 3.97 23.41 5.90 1.97 23.34 11.85 

(N) (105) (18) (102) (21) 

W/AOdds 1.66 1.97 NA 1.19 

W/H Odds 1.68 1.20 1.51 0.72 



126 



Appendix 

Appendix Table 2-3 (Continued) 

Percentages Participating in Core Activities by Respondents Who 

Did and Did Not Watcli Such Events on Television, 1985 SPPA 



Core Item: 




Art 






No 


Yes 


Y/N 


W 


16.49 


44.09 


2.67 


(N) 


(1.242) 


(461) 




A 


11.04 


18.96 


1.72 


(N) 


(143) 


(44) 




H 


8.11 


56.30 


6.94 


(N) 


(101) 


(23) 




W/AOdds 


1.49 


2.33 




W/H Odds 


2.03 


0.78 





Ns unweighted, percentages weighted. "Yes" refers to respondents who watched relevant television 
programs, "No" to those who did not. Y/N=probability of participation for persons who watched 
programs to those who did not. W/A Odds=probability of participation for whites/probability of 
participation for African-Americans. W/H Odds=probability of participation for whites/probability for 
Hispanics. 



127 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 2-4 

Percentage Participating in Core Activities, Respondents with 

and without Specific Lessons or Classes Before Age 18, 

1982 SPP A 



Core Item: 




Jazz attendance 






Jazz attendance 




Lesson: 




Music 




I 


iAusIc appreciation 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


4.70 


13.22 


2.81 


6.37 


18.56 


2.91 


(N) 


(2,283) 


(2,301) - 




(3,593) 


(991) 




A 


8.76 


33.96 


3.87 


14.87 


34.47 


2.32 


(N) 


(318) 


(195) 




(414) 


(99) 




H 


9.18 


11.23 


1.22 


7.69 


29.36 


3.82 


(N) 


(240) 


(65) 




(276) 


(29) 




W/Aodds 


0.54 


0.39 




0.43 


0.54 




W/H odds 


0.51 


1.18 




0.83 


0.63 




Core Item: 


Classical attendance 


Classical attendance 


Lesson: 




Music 




I 


lAusic appreciation 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


7.58 


21.45 


2.83 


10.53 


29.34 


2.79 


(N) 


(2,287) 


(2,301) 




(3,597) 


(991) 




A 


3.98 


11.62 


2.92 


4.17 


18.09 


4.34 


(N) 


(318) 


(195) 




(414) 


(99) 




H 


4.42 


12.66 


2.86 


3.79 


31.32 


8.26 


(N) 


(240) 


(65) 




(276) 


(29) 




W/A odds 


1.90 


1.85 




2.53 


1.62 




W/H odds 


1.71 


1.69 




2.78 


0.94 




Core Item: 




Opera 






Opera 




Lesson: 




Music 




I 


tAusic appreciation 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


2.00 


3.81 


1.91 


2.17 


5.64 


2.60 


(N) 


(2,286) 


(2,301) 




(3,597) 


(990) 




A 


0.21 


1.79 


8.52 


0.44 


2.42 


5.50 


(N) 


(318) 


(195) 




(414) 


(99) 




H 


1.42 


1.44 


1.01 


0.54 


10.39 


19.24 


(N) 


(240) 


(65) 




(276) 


(29) 




W/A Odds 


9.52 


2.13 




4.93 


2.33 




W/H Odds 


1.41 


2.65 




4.02 


0.54 





128 



Appendix 

Appendix Table 2-4 (Continued) 

Percentage Participating in Core Activities, Respondents with 

and without Specific Lessons cr Classes Before Age 18, 

1982 SPP A 



Core Item: 




Musical 






Musical 




Lesson: 




Music 




Music appreciatior 


1 




No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


13.65 


27.50 


2.01 


15.91 


37.89 


2.38 


(N) 


(2,287) 


(2,302) 




(3,598) 


(991) 




A 


5.05 


18.31 


3.63 


6.75 


24.38 


3.61 


(N) 


(318) 


(195) 




(414) 


(99) 




H 


6.25 


14.54 


2.33 


5.23 


37.20 


7.11 


N 


(240) 


(65) 




(276) 


(29) 




W/AOdds 


2.70 


1.50 




2.36 


1.55 




W/H Odds 


2.18 


1.89 


- 


3.04 


1.02 




Core item: 




Musical 






Plays 




Lesson: 




Acting 






Acting 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


18.01 


44.53 


2.47 


11.34 


29.55 


2.61 


(N) 


(4,139) 


(450) 




(4,138) 


(449) 




A 


9.52 


22.72 


2.39 


4.24 


14.69 


3.46 


(N) 


(481) 


(32) 




(481) 


(32) 




H 


7.36 


17.55 


2.38 


3.22 


12.51 


3.89 


(N) 


(282) 


(23) 




(282) 


(23) 




W/AOdds 


1.89 


1.96 




2.67 


2.01 




W/H Odds 


2.45 


2.54 




3.52 


2.36 




Core Item: 




Art extiibits 






Art extiibits 




Lesson: 




Art 






Crafts 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


15.39 


45.43 


2.95 


16.28 


36.42 


2.24 


(N) 


(3,441) 


(1.144) 




(3.065) 


(1,519) 




A 


9.37 


33.22 


3.55 


6.70 


35.92 


5.36 


(N) 


(426) 


(88) 




(396) 


(118) 




H 


12.54 


35.10 


2.80 


12.11 


33.56 


2.77 


(N) 


(256) 


(49) 




(246) 


(59) 




W/AOdds 


1.64 


1.37 




2.43 


1.01 




W/H Odds 


1.25 


1.29 




1.34 


1.09 





129 



Race, Ethnicity, and Participation in tlie Arts 

Appendix Table 2-4 (Continued) 

Percentage Participating in Core Activities, Respondents with 

and without Specific Lessons or Classes Before Age 18, 

1982 SPPA 



Core Item: 




Art exhibits 






Ballet 




Lesson: 




Art appreciation 






Ballet 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


16.11 


49.50 


3.07 


2.65 


22.89 


8.64 


(N) 


(3,647) 


(938) 




(4,215) 


(374) 




A 


9.53 


36.39 


3.82 


1.23 


15.14 


12.31 


(N) 


(437) 


(77) 




(493) 


(21) 




H 


12.16 


50.52 


4.15 


2.19 


32.44 


14.81 


(N) 


(272) 


(33) 




(294) 


(11) 




W/AOdds 


1.69 


1.36 




2.15 


1.51 




W/H Odds 


1.32 


0.98 




1.21 


0.71 




Core Item: 




Reading literature 










Lesson: 


No 


Creative writing 
Yes 


Y/N 








W 


55.82 


84.38 


1.51 








(N) 


(3,694) 


(879) 










A 


37.18 


82.59 


2.22 








(N) 


(445) 


(68) 










H 


32.69 


60.82 


1.86 








(N) 


(266) 


(37) 










W/AOdds 


1.50 


1.02 










W/H Odds 


1.71 


1.39 











Ns unweighted, percentages weighted. Y/N=probability of participation for persons who have taken 
lessons/probability for those who have not. W/A Odds=probability of participation for whites/probability 
of participation for African-Americans. W/H Odds=probability of participation for whites/probability for 
Hispanics. 



130 



Appendix 

Appendix Table 2^ 

Percentage Participating in Core Activities, Respondents with 

and without Specific Lessons or Classes Before Age 18, 

1985 SPPA 



Core Item: 




Jazz attendance 






Jazz attendance 


Lesson: 




Music 




\ 


i/lusic appreciation 




No 


Yes 


Y/N 


No 


Yes Y/N 


W 


5.45 


15.52 


2.85 


7.38 


22.21 3.00 


(N) 


(928) 


(993) 




(1,482) 


(439) 


A 


8.12 


16.99 


2.09 


7.16 


32.15 4.49 


(N) 


(122) 


(77) 




(166) 


(33) 


H 


0.57 


14.56 


25.54 


2.55 


22.72 8.91 


(N) 


(105) 


(37) 




(129) 


(13) 


W/Aodds 


0.67 


0.93 




1.03 


0.69 


W/Hodds 


9.56 


1.07 




2.89 


0.98 


Core Item: 


Classical attendance 


Classical attendance 


Lesson: 




Music 




I 


i/lusic appreciation 




No 


Yes 


Y/N 


No 


Yes Y/N 


W 


7.34 


24.19 


3.30 


10.88 


34.18 3.14 


(N) 


(929) 


(992) 




(1,482) 


(439) 


A 


2.39 


9.29 


3.89 


3.90 


10.09 2.59 


(N) 


(122) 


(77) 




(166) 


(33) 


H 


1.68 


31.60 


18.81 


7.78 


22.91 2.94 


(N) 


(106) 


(37) 




(130) 


(13) 


W/Aodds 


3.07 


2.60 




2.79 


3.39 


W/H odds 


4.37 


0.77 




1.40 


1.49 


Core Item: 




Opera 






Opera 


Lesson: 




Music 




I 


i^usic appreciation 




No 


Yes 


Y/N 


No 


Yes Y/N 


W 


1.71 


4.51 


2.64 


2.23 


6.43 2.88 


(N) 


(928) 


(992) 




(1,481) 


(439) 


A 


0.00 


3.56 


NA 


0.00 


7.76 NA 


(N) 


(122) 


(77) 




(166) 


(33) 


H 


0.00 


2.24 


NA 


0.58 


0.00 0.00 


(N) 


(105) 


(37) 




(129) 


(13) 


W/AOdds 


NA 


1.27 




NA 


0.85 


W/H Odds 


NA 


2.01 




3.84 


NA 



131 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 2-5 (Continued) 

Percentage Participating in Core Activities, Respondents with 

and without Specific Lessons or Classes Before Age 18, 

1985 SPP A 



Core Item: 




Musical 






Musical 




Lesson: 




Music 




Music appreciation 


1 




No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


10.80 


26.25 


2.43 


14.39 


34.17 


2.37 


(N) 


(929) 


(991) 




(1,481) 


(439) 




A 


7.59 


11.69 


1.54 


6.36 


22.56 


3.55 


(N) 


(122) 


(77) 




(166) 


(33) 




H 


3.46 


18.92 


5.47 


5.48 


29.60 


5.40 


(N) 


(106) 


(37) 




(130) 


(13) 




W/AOdds 


1.42 


2.25 




2.26 


1.51 




W/H Odds 


3.12 


1.39 




2.63 


1.15 




Core Item: 




Musical 






Plays 




Lesson: 




Acting 






Acting 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


15.84 


42.24 


2.67 


11.57 


33.56 


2.90 


(N) 


(1,714) 


(206) 




(1,715) 


(207) 




A 


8.65 


14.04 


1.62 


4.41 


15.79 


3.58 


(N) 


(182) 


(17) 




(181) 


(17) 




H 


6.58 


18.93 


2.88 


6.58 


0.00 


0.00 


(N) 


(133) 


(10) 




(133) 


(10) 




W/AOdds 


1.83 


3.01 




2.62 


2.13 




W/H Odds 


2.41 


2.23 




1.76 


NA 




Core Item: 




Art exhibits 






Art exhibits 




Lesson: 




Art 






Crafts 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


16.26 


46.46 


2.86 


10.57 


37.53 


3.55 


(N) 


(1,383) 


(538) 




(1,183) 


(738) 




A 


4.25 


30.71 


7.23 


6.40 


14.93 


2.33 


(N) 


(161) 


(38) 




(140) 


(59) 




H 


17.91 


36.89 


2.06 


19.05 


25.46 


1.34 


(N) 


(123) 


(20) 




(116) 


(27) 




W/AOdds 


3.83 


1.51 




1.65 


2.51 




W/H Odds 


0.91 


1.26 




0.55 


1.47 





132 



Appendix 

Appendix Table 2-5 (Continued) 

Percentage Participating in Core Activities, Respondents with 

and without Specific Lessons or Classes Before Age 18, 

1985 SPPA 



Core Item: 




Art exhibits 






Ballet 




Lesson: 




Art appreciation 






Ballet 






No 


Yes 


Y/N 


No 


Yes 


Y/N 


W 


16.37 


54.98 


3.36 


3.56 


18.40 


5.17 


(N) 


(1,492) 


(429) 




(1,731) 


(191) 




A 


4.93 


28.00 


5.68 


1.14 


14.38 


12.61 


(N) 


(166) 


(33) 




(193) 


(6) 




H 


18.63 


39.15 


2.10 


2.95 


0.00 


0.00 


(N) 


(133) 


(10) 




(137) 


(6) 




W/AOdds 


3.32 


1.96 




3.12 


1.28 




W/H Odds 


0.88 


1.40 




1.21 


NA 




Core Item: 




Reading literature 










Lesson: 


No 


Creative writing 
Yes 


Y/N 








W 


53.45 


90.37 


1.69 








(N) 


(1.508) 


(410) 










A 


38.55 


62.99 


1.62 








(N) 


(173) 


(25) 










H 


36.04 


89.18 


2.47 








(N) 


(138) 


(5) 










W/AOdds 


1.39 


1.43 










W/H Odds 


1.48 


1.01 











Ns unweighted, percentages weighted. Y/N=probability of participation for persons who have taken 
lessons/probability for those who have not. W/A Odds=probability of participation for whites/probability 
of participation for African-Americans. W/H Odds=probability of participation for whites/probability for 
Hispanics. 



133 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 3-1 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1982 Disaggregated Subsamples: Whites (W), 

African-Americans (A), and Hispanics (H) 



AGE 



EDUC 



WOMEN 

b 

se 

sig 

SMSA 
b 
se 
sig 

b 
se 
sig 

b 
se 
sig 

b 
se 
sig 

b 
se 
sig 

MARIT 
b 
se 
sig 

INT 



INC 



OCC 





Attend Jazz 






Attend 






Attend 






concerts 




classical concerts 




opera 




w 


A 


H 


W 


A 


H 


W 


A 


H 


.101 


-.451 


-.026 


.650 


.226 


.066 


.444 


.521 


1.573 


.065 


.158 


.252 


.057 


.228 


.260 


.107 


.518 


.558 


NS 


a 


NS 


d 


NS 


NS 


d 


NS 


a 


.373 


.658 


-.318 


.107 


.602 


.453 


.666 


.357 


-.387 


.077 


.207 


.341 


.062 


.310 


.453 


.139 


.686 


.702 


d 


a 


NS 


NS 


NS 


NS 


d 


NS 


NS 


-.028 


-.041 


-.028 


.016 


.018 


.003 


.031 


-.012 


.013 


.003 


.007 


.012 


.002 


.008 


.010 


.003 


.022 


.019 


d 


d 


a 


d 


a 


NS 


d 


NS 


NS 


.216 


.235 


.083 


.335 


.274 


.191 


.250 


.385 


.427 


.015 


.038 


.046 


.012 


.048 


.049 


.022 


.120 


.101 


d 


d 


NS 


d 


d 


c 


d 


a 


d 


.073 


.002 


.071 


.114 


.037 


.142 


.200 


.325 


.143 


.021 


.065 


.095 


.018 


.088 


.095 


.032 


.158 


.163 


b 


NS 


NS 


d 


NS 


NS 


d 


a 


NS 


.184 


.422 


.838 


.252 


.546 


.463 


.392 


.519 


.063 


.072 


.178 


.282 


.061 


.258 


.287 


.121 


.628 


.512 


a 


a 


a 


d 


a 


NS 


a 


NS 


NS 


.810 


.236 


.010 


.467 


.736 


.602 


.773 


.031 


.753 


.073 


.172 


.290 


.067 


.245 


.293 


.124 


.550 


.536 


d 


NS 


NS 


d 


a 


a 


d 


NS 


NS 



4.98 -3.66 -2.52 -7.89 -7.93 -5.83 -9.85 -10.58 -10.72 



134 



Appendix 

Appendix Table 3-1 (Continued) 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1982 Disaggregated Subsamples: Whites (W), 

African-Americans (A), and Hispanics (H) 



WOMEN 

b 

se 

sig 



SMSA 



AGE 



EDUC 



b 
se 
sig 

b 
se 
sig 

b 
se 
sig 

b 
se 
sig 

b 
se 
sig 



MARIT 
b 
se 
sig 



INC 



occ 





Attend 






Attend 






Attend 






musical 






play 






ballet 




w 


A 


H 


W 


A 


H 


W 


A 


H 


.583 


-.139 


.510 


.581 


-.050 


-.175 


1.101 


.960 


1.202 


.049 


.186 


.235 


.058 


.240 


.316 


.097 


.477 


.394 


d 


NS 


a 


d 


NS 


NS 


d 


a 


a 


.482 


.659 


.737 


.202 


1.266 


.226 


.617 


1.155 


8.073 


.055 


.258 


.427 


.064 


.423 


.539 


.114 


.705 


* 


d 


a 


NS 


a 


a 


NS 


d 


NS 


* 


.009 


.001 


-.001 


.013 


.017 


.009 


.010 


-.032 


.019 


.002 


.007 


.009 


.002 


.009 


.012 


.003 


.020 


.014 


d 


NS 


NS 


d 


NS 


NS 


b 


NS 


NS 


.241 


.219 


.189 


.288 


.211 


.275 


.283 


.462 


.380 


.010 


.041 


.044 


.012 


.051 


.064 


.019 


.106 


.077 


d 


d 


d 


d 


d 


d 


d 


d 


d 


.189 


.262 


.197 


.187 


.256 


.146 


.112 


.032 


.211 


.016 


.068 


.086 


.018 


.086 


.111 


.028 


.167 


.131 


d 


c 


a 


d 


a 


NS 


d 


NS 


NS 


.400 


.314 


.644 


.459 


.310 


.738 


.471 


-.088 


.769 


.053 


.214 


.255 


.063 


.277 


.351 


.102 


.493 


.399 


d 


NS 


a 


d 


NS 


a 


d 


NS 


NS 


.271 


.184 


.360 


.504 


.582 


.145 


.525 


.668 


.645 


.059 


.206 


.263 


.063 


.264 


.366 


.104 


.458 


.415 


d 


NS 


NS 


d 


a 


NS 


d 


NS 


NS 



INT 



-6.25 -^.02 -5.91 -7.54 -7.92 -7.22 -9.12 -10.94 -18.09 



135 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 3-1 (Continued) 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1982 Disaggregated Subsamples: Whites (W), 

African-Americans (A), and Hispanics (H) 





Visit art museum 


Perform: Play 




Perform: Act, 






or gallery 




musical instrument 


sing, or dance 




W 


A 


H 


W 


A 


H 


W 


A 


C 


WOMEN 




















b 


.436 


-.157 


.101 


.011 


.046 


.237 


.300 


.069 


.295 


se 


.047 


.177 


.202 


.091 


.293 


.424 


.086 


.248 


.420 


sig 


d 


NS 


NS 


NS 


NS 


NS 


b 


NS 


NS 


SMSA 




















b 


.245 


1.364 


-.071 


-.006 


-.522 


8.080 


-.061 


-.190 


7.959 


se 


.051 


.297 


.298 


.099 


.304 


* 


.092 


.274 


* 


sig 


d 


d 


NS 


NS 


NS 


* 


NS 


NS 


* 


AGE 




















b 


.001 


-.000 


-.012 


-.011 


-.001 


-.018 


-.020 


-.016 


-.020 


se 


.002 


.007 


.009 


.003 


.011 


.018 


.003 


.010 


.018 


sig 


NS 


NS 


NS 


b 


NS 


NS 


d 


NS 


NS 


EDUC 












- 








b 


.320 


.279' 


.279 


.106 


.091 


.001 


.112 


.084 


-.072 


se 


.011 


.040 


.042 


.019 


.057 


.067 


.019 


.050 


.065 


sig 


d 


d 


d 


d 


NS 


NS 


d 


NS 


NS 


INC 




















b 


.115 


.152 


.226 


-.089 


.072 


-.145 


-.062 


.042 


.202 


se 


.015 


.068 


.076 


.033 


.117 


.199 


.030 


.100 


.166 


sig 


d 


a 


a 


a 


NS 


NS 


a 


NS 


NS 


OCC 




















b 


.255 


.710 


.444 


-.055 


.161 


.657 


.209 


.403 


.906 


se 


.050 


.197 


.219 


.103 


.343 


.488 


.095 


.282 


.488 


sig 


d 


b 


a 


NS 


NS 


NS 


a 


NS 


NS 


MARIT 




















b 


.415 


.194 


.019 


.418 


.515 


.018 


.245 


.275 


.444 


se 


.056 


.193 


.234 


.105 


.325 


.478 


.099 


.273 


.473 


sig 


d 


NS 


NS 


d 


NS 


NS 


a 


NS 


NS 


INT 


-6.29 


-7.07 


-5.02 


-3.96 


^.40 


-10.98 


-3.80 


-3.58 


-10.94 



136 



Appendix 

Appendix Table 3-1 (Continued) 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1982 Disaggregated Subsamples: Whites (W), 

African-Americans (A), and Hispanics (H) 



Read novels, short stories, 
poems, or plays 



W A H 



WOMEN 

b 

se 

sig 

SMSA 
b 
se 
sig 

AGE 
b 
se 
sig 

EDUC 
b 
se 
sig 

INC 

b 
se 
sig 

OCC 
b 
se 
sig 

MARIT 
b 
se 
sig 

INT -4.03 -3.09 -3.15 

Notes: b is the unstandardized logistic regression coefficient, se is the standard error, sig refers to the level of 
statistical significance, where a=probability less than .05, b=probability less than .01 , c=probability less than .001 , 
c=probability less than .00005, and NS= not significant. Variables are defined in the text. The coefficients and 
standard errors for INC are multiplied by 1 0,000 for purposes of display. *=The program does not compute reliable 
standard errors and significance tests for coefficients of this magnitude. 



137 



.889 


.552 


.496 


.041 


.123 


.156 


d 


d 


a 


.085 


.310 


-.120 


.043 


.137 


.219 


a 


a 


NS 


.002 


-.014 


-.004 


.001 


.004 


.006 


NS 


a 


NS 


.288 


.204 


.??? 


.009 


.026 


.028 


d 


d 


d 


.079 


.126 


.034 


.015 


.053 


.065 


d 


a 


NS 


.225 


.711 


.411 


.047 


.146 


.183 


d 


d 


a 


.245 


.261 


.029 


.052 


.134 


.183 


d 


NS 


NS 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 3-2 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1985 Disaggregated Subsamples: 

Whites (W), African-Americans (A), and Hispanics (H) 







Attend jazz 






Attend 






Attend 








concerts 




classical concerts 




opera 






W 


A 


H 


W 


A 


H 


W 


A 


H 


WOMEN 




















b 


.143 


-.491 


-.161 


.542 


.336 


.276 


.473 


.156 


-1.415 


se 


.072 


.182 


.316 


.064 


.256 


.315 


.125 


.501 


1.262 


sig 


a 


a 


NS 


C 


NS 


NS 


c 


NS 


NS 


SMSA 




















b 


.359 


.358 


-.388 


.165 


.090 


.269 


.099 


8.227 


7.111 


se 


.088 


.232 


.403 


.072 


.331 


.474 


.147 


* 


* 


sig 


c 


NS 


NS 


a 


NS 


NS 


NS 


* 


* 


AGE 




















b 


-.024 


-.034 


-.010 


.016 


.024 


.012 


.019 


.029 


-.032 


se 


.003 


.008 


.014 


.002 


.009 


.012 


.004 


.016 


.054 


sig 


c 


c 


NS 


c 


a 


NS 


C 


NS 


NS 


EDUC 




















b 


.265 


.235 


.175 


.342 


.280 


.369 


.320 


.361 


.214 


se 


.017 


.043 


.059 


.014 


.051 


.065 


.027 


.103 


.215 


sig 


c 


c 


a 


c 


c 


c 


c 


b 


NS 


INC 




















b 


.081 


.230 


-.014 


.103 


.268 


.100 


.133 


.031 


.470 


se 


.021 


.061 


.112 


.018 


.079 


.102 


.034 


.169 


.267 


sig 


c 


b 


NS 


c 


b 


NS 


c 


NS 


NS 


OCC 




















b 


.193 


-.340 


.429 


.299 


.068 


-.592 


.549 


.345 


7.805 


se 


.082 


.214 


.349 


.071 


.280 


.354 


.150 


.561 


* 


sig 


a 


NS 


NS 


c 


NS 


NS 


b 


NS 


* 


MARIT 




















b 


.548 


.132 


.040 


.477 


.816 


.337 


.431 


.819 


-.155 


se 


.083 


.200 


.348 


.076 


.277 


.350 


.147 


.536 


1.135 


sig 


c 


NS 


NS 


c 


a 


NS 


a 


NS 


NS 


INT 


-5.72 


-3.97 


^.12 


-8.08 


-8.31 


-7.97 


-9.92 


-18.67 


-21.56 



138 



Appendix 

Appendix Table 3-2 (Continued) 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1985 Disaggregated Subsamples: 

Whites (W), African-Americans (A), and Hispanics (H) 







Attend 






Attend 






Attend 








musical 






play 






ballet 






W 


A 


H 


W 


A 


H 


W 


A 


H 


WOMEN 




















b 


.513 


.560 


.186 


.436 


.285 


.189 


.980 


.198 


-.161 


se 


.057 


.233 


.284 


.065 


.269 


.325 


.105 


.416 


.429 


sig 


d 


a 


NS 


d 


NS 


NS 


d 


NS 


NS 


SMSA 




















b 


.593 


.836 


-.021 


.402 


.594 


.111 


.233 


8.544 


-.084 


se 


.067 


.353 


.417 


.077 


.393 


.477 


.120 


* 


.600 


sig 


d 


a 


NS 


d 


NS 


NS 


NS 


* 


NS 


AGE 




















b 


.007 


.014 


.023 


.009 


.009 


.021 


.004 


.006 


-.002 


se 


.002 


.008 


.011 


.002 


.010 


.012 


.003 


.014 


.019 


sig 


d 


NS 


a 


d 


NS 


NS 


NS 


NS 


NS 


EDUC 




















b 


.217 


.255 


.394 


.294 


.372 


.251 


.318 


.259 


.205 


se 


.012 


.046 


.060 


.014 


.059 


.061 


.022 


.082 


.082 


sig 


d 


d 


d 


d 


d 


d 


d 


a 


a 


INC 




















b 


.162 


.188 


.089 


.117 


.226 


.126 


.125 


.016 


.170 


se 


.016 


.072 


.093 


.018 


.083 


.105 


.027 


.143 


.133 


sig 


d 


a 


NS 


d 


a 


NS 


d 


NS 


NS 


occ 




















b 


.377 


.492 


.448 


.414 


-.180 


-.080 


.287 


.744 


.311 


se 


.063 


.242 


.308 


.073 


.296 


.363 


.115 


.453 


.479 


sig 


d 


a 


NS 


d 


NS 


NS 


a 


NS 


NS 


MARIT 




















b 


.244 


.458 


-.071 


.415 


.712 


.304 


.445 


-.726 


.106 


se 


.069 


.242 


.323 


.077 


.290 


.363 


.115 


.482 


.477 


sig 


a 


NS 


NS 


d 


a 


NS 


c 


NS 


NS 


INT 


-6.11 


-7.88 


-8.35 


-7.45 


-9.25 


-6.97 


-8.94 


-15.93 


-6.12 



139 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 3-2 (Continued) 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1985 Disaggregated Subsamples: 

Wliites (W), African-Americans (A), and IHispanics (H) 





Visit art museum 


Perform: Play 


i 


Perfonn: Act, 






or gallery 




musical instrument 


Sing, or dance 




W 


A 


H 


W 


A 


H 


W 


A 


H 


WOMEN 
b 


.383 


.276 


.141 


-.147 


.051 


-1.185 


.240 


.129 


-.258 


se 


.052 


.203 


.203 


.124 


.433 


.604 


.103 


.326 


.471 


sig 


d 


NS 


NS 


NS 


NS 


a 


a 


NS 


NS 


SMSA 




















b 


.495 


.362 


-.083 


-.483 


.154 


-.125 


196.278 


.020 


8.004 


se 


.060 


.269 


.286 


.131 


.533 


.687 


.109 


.396 


* 


sig 


d 


NS 


NS 


b 


NS 


NS 


a 


NS 


* 


AGE 

b 


-.001 


-.007 


.007 


-.023 


-.042 


6.026 


-.016 


.003 


-.066 


se 


^ .002 


.008 


.008 


.005 


.020 


.026 


.004 


.012 


.029 


sig 


NS 


NS 


NS 


d 


a 


NS 


d 


NS 


a 


EDUC 












- 








b 


.312 


.272 - 


.155 


.179 


.058 


.153 


.143 


.101 


-.037 


se 


.012 


.044 


.037 


.028 


.095 


.097 


.022 


.064 


.084 


sig 


d 


d 


d 


d 


NS 


NS 


d 


NS 


NS 


INC 




















b 


.097 


.208 


.114 


.090 


.017 


.019 


-.056 


.084 


-.031 


se 


.015 


.065 


.072 


.039 


.154 


.180 


.032 


.112 


.173 


sig 


d 


a 


NS 


a 


NS 


NS 


NS 


NS 


NS 


OCC 




















b 


.257 


.385 


.688 


.412 


.743 


.058 


.348 


.413 


.820 


se 


.058 


.217 


.224 


.143 


.466 


.585 


.117 


.360 


.537 


sig 


d 


NS 


a 


a 


NS 


NS 


a 


NS 


NS 


MARIT 




















b 


.297 


.252 


.125 


.155 


-.104 


.149 


.101 


.631 


.074 


se 


.063 


.218 


.228 


.143 


.471 


.571 


.120 


.356 


.532 


sig 


d 


NS 


NS 


NS 


NS 


NS 


NS 


NS 


NS 


INT 


-6.23 


-6.34 


-3.98 


-4.64 


-3.58 


^.26 


-4.35 


-5.33 


-9.10 



140 



Appendix 

Appendix Table 3-2 (Continued) 

Logistic Regression Analyses Predicting Participation in Core 

Activities for 1985 Disaggregated Subsamples: 

Whites (W), African-Americans (A), and Hispanics (H) 



Read novels, short stories, 
poems, or plays 



W A H 



WOMEN 

b 

se 

sig 

SMSA 
b 
se 
sig 

AGE 
b 
se 
sig 

EDUC 
b 
se 
sig 

INC 
b 
se 
sig 

OCC 
b 
se 
sig 

MARIT 
b 
se 
sig 

INT -3.60 -3.10 -3.72 

Notes: b is the unstandardized logistic regression coefficient, se is tiie standard error, sig refers to tfie level of 
statistical significance, where a=probability less than .05, b=probability less than .01 , c=probability less than .001 , 
c=probability less than .00005, and NS=not significant. Variables are defined in the text. The coefficients and 
standard errors for INC are multiplied by 10,000 for purposes of display. *=Program does not compute reliable 
standard errors or significance statistics when regression coefficients are this high. 



.912 


.362 


.595 


.046 


.127 


.171 


d 


a 


b 


.122 


.797 


.119 


.048 


.155 


.239 


a 


d 


NS 


.002 


-.007 


.013 


.001 


.004 


.007 


NS 


NS 


a 


.240 


.171 


.187 


.010 


.026 


.029 


d 


d 


d 


.091 


.069 


.102 


.015 


.050 


.066 


c 


NS 


NS 


.290 


.648 


.776 


.052 


.152 


.193 


d 


d 


c 


.101 


.031 


.234 


.057 


.143 


.193 


NS 


NS 


NS 



141 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 4-1 

Weighted Percentages of Respondents Who Wished to Attend 

Jazz Music Performances More, Citing Selected Reasons for Not 

Doing So: Whites (W), African-Americans (A), and Hispanics (H) 





Attended during 
previous 12 months 


Did not attend during 
previousi 2 months 




W 


A 


H 


W 


A 


H 


1982 














Tickets sold out 


3.96 


4.77 


14.87 


1.18 


0.80 


0.00 


Cost 


31.35 


59.05 


34.14 


25.62 


45.05 


39.94 


Not available 


29.07 


16.66 


31.52 


22.43 


12.67 


14.06 


Child care 


3.80 


8.90 


12.87 


7.94 


8.26 


11.49 


Too far to go 


13.98 


2.02 


20.04 


15.59 


7.00 


12.95 


Transportation 


7.28 


10.91 


28.47 


5.66 


13.75 


6.35 


Fear crime 


0.66 


0.00 


0.00 


2.89 


5.17 


0.00 


Lack motivation 


8.33 


6.11 


14.08 


13.85 


9.53 


11.12 


Too little time 


42.97 


37.83 


29.52 


41.39 


24.20 


37.22 


N (unweighted) 


220 


55 - 


15 


532 


113 


39 


1985 














Tickets sold out 


0.87 


3.90 


NA 


1.40 


1.23 


0.00 


Cost 


21.68 


51.31 


NA 


28.63 


39.26 


54.62 


Not available 


24.25 


30.81 


NA 


23.43 


12.73 


15.56 


Child care 


8.69 


8.76 


NA 


10.97 


2.57 


21.15 


Too far to go 


15.13 


9.30 


NA 


14.22 


5.75 


0.00 


Transportation 


5.74 


22.93 


NA 


5.20 


7.56 


2.68 


Fear crime 


2.71 


11.42 


NA 


1.12 


3.52 


4.04 


Lack motivation 


11.05 


0.00 


NA 


14.88 


3.62 


31.32 


Too little time 


47.23 


19.83 


NA 


45.16 


41.48 


30.54 


N (unweighted) 


102 


20 


4 


241 


59 


21 



In 1 985, too few Hispanic attenders reported wanting to go more to report statistics. Fewer than 1 
percent of any group reported discomfort, no one to go with, handicap, poor quality, publicity, work- 
related reasons, performance times, or transience. 



142 



Appendix 

Appendix Table 4-2 

Weighted Percentages of Respondents Wishing to Attend 

Classical Music Performances More, Citing Selected Reasons for 

Not Doing So: Whites (W), African-Americans (A), 

and Hispanics (H) 





Attended during 
previous 12 months 


Did not attend during 
previous 12 months 




W 


A 


H 


W 


A 


H 


1982 














Cost 


32.83 


34.71 


55.96 


28.30 


43.96 


48.43 


Not available 


21.44 


0.00 


8.92 


23.27 


14.32 


8.07 


No one to go with 


7.63 


9.61 


19.11 


7.18 


6.40 


2.31 


Child care 


5.28 


21.98 


0.00 


7.51 


9.96 


8.66 


Handicap 


2.32 


0.00 


8.78 


10.04 


7.81 


2.88 


Too far to go 


17.24 


0.00 


15.76 


15.02 


12.31 


19.69 


Transportation 


7.19 


8.00 


19.11 


8.15 


20.50 


15.43 


Lack motivation 


11.79 


17.61 


9.00 


14.71 


6.73 


3.18 


Too little time 


41.40 


35.30 


16.13 


39.10 


34.53 


32.96 


N (unweighted) 


303 


14 


10 


552 


48 


36 


1985 














Cost 


22.93 


NA 


NA 


30.08 


23.53 


NA 


Not available 


18.34 


NA 


NA 


23.79 


3.76 


NA 


No one to go with 


9.14 


NA 


NA 


5.83 


8.32 


NA 


Child care 


7.65 


NA 


NA 


10.17 


10.74 


NA 


Handicap 


3.27 


NA 


NA 


6.43 


3.03 


NA 


Too far to go 


11.50 


NA 


NA 


25.46 


10.28 


NA 


Transportation 


5.76 


NA 


NA 


8.55 


16.86 


NA 


Lack motivation 


16.42 


NA 


NA 


12.06 


9.50 


NA 


Too little time 


51.06 


NA 


NA 


34.70 


47.98 


NA 


N (unweighted) 


130 


7 


4 


207 


23 


9 



In 1985, too few African-American and Hispanic attenders and Hispanic nonattenders reported 
wanting to go more to report statistics. Fewer than 1 percent of any group reported tickets sold out, 
discomfort, crime, poor quality, publicity, work-related reasons, performance times, or transience as 
reasons for not attending. 



143 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 4-3 

Weighted Percentages of Respondents Wishing to Attend Opera 

Performances More, Citing Selected Reasons for Not Doing So: 

Whites (W), African-Americans (A), and Hispanics (H) 





Attended during 
previous 12 months 


H 


Did not attend di 
previous 12 moi 

W A 


jring 
iths 




W 


A 


H 


1982 














Cost 


38.47 


NA 


NA 


34.55 


38.52 


68.01 


Not available 


32.95 


NA 


NA 


25.98 


8.63 


6.82 


No one to go with 


11.68 


NA 


NA 


9.22 


3.30 


11.98 


Handicap 


4.32 


NA 


NA 


10.56 


2.09 


7.24 


Too far to go 


14.15 


NA 


NA 


17.40 


12.27 


36.14 


Transportation 


9.84 


NA 


NA 


8.39 


11.52 


7.24 


Lack motivation 


7.02 


NA 


NA 


10.34 


10.28 


0.00 


Too little time 


20.21 


NA 


NA 


30.27 


30.48 


15.05 


N (unweighted) 


50 


1 


1 


311 


23 


14 



1985 



Cost 


43.54 


NA 


NA 


36.59 


8.99 


NA 


Not available 


17.55 


NA 


NA 


14.31 


10.87 


NA 


No one to go with 


3.56 


NA 


NA 


6.38 


5.68 


NA 


Handicap 


0.00 


NA 


NA 


5.42 


0.00 


NA 


Too far to go 


16.14 


NA 


NA 


25.97 


13.59 


NA 


Transportation 


4.51 


NA 


NA 


13.00 


30.16 


NA 


Lack motivation 


4.61 


NA 


NA 


16.49 


0.00 


NA 


Too little time 


56.36 


NA 


NA 


33.48 


61.20 


NA 


N (unweighted) 


21 


2 





141 


10 


7 



In 1 982 and 1 985, too few African-American and Hispanic attenders, and in 1 985 too few Hispanic nonattenders, 
reported wanting to go more to report statistics. Fewer than 10 percent of any group reported tickets sold out, 
discomfort, child care, crime, poor quality, publicity, work-related reasons, performance times, or transience as 
reasons for not attending. 



144 



Appendix 

Appendix Table 4-4 

Weighted Percentages of Respondents Wishing to Attend l\/lusical 

Theater Performances IMore, Citing Selected Reasons for Not 

Doing So: Whites (W), African-Americans (A), and Hispanics (H) 





Attended during 
previous 12 months 


Did not attend during 
previous 12 montiis 




W 


A 


H 


W 


A 


H 


1982 














Cost 


36.56 


55.08 


62.86 


30.93 


47.08 


37.41 


Not available 


22.75 


13.90 


19.15 


21.02 


11.85 


13.12 


No one to go with 


6.92 


9.33 


7.14 


8.79 


3.09 


8.81 


Child care 


6.09 


10.98 


3.18 


7.58 


11.68 


16.15 


Too far to go 


15.65 


5.26 


15.98 


15.95 


4.95 


29.42 


Transportation 


7.42 


4.46 


8.92 


7.56 


8.18 


17.40 


Fear crime 


2.75 


7.49 


3.95 


3.57 


3.41 


5.17 


Lack motivation 


10.79 


24.01 


3.99 


12.07 


12.18 


9.44 


Too little time 


36.58 


44.78 


33.08 


36.52 


29.30 


32.90 


N (unweighted) 


620 


24 


23 


969 


81 


44 


1985 














Cost 


28.83 


53.27 


NA 


32.10 


43.48 


53.39 


Not available 


17.86 


13.47 


NA 


19.07 


15.20 


0.00 


No one to go with 


5.81 


4.81 


NA 


5.55 


12.04 


14.46 


Child care 


4.99 


0.00 


NA 


10.33 


7.07 


17.07 


Too far to go 


16.31 


0.00 


NA 


18.84 


15.40 


5.99 


Transportation 


7.26 


4.29 


NA 


8.83 


14.24 


0.00 


Fear crime 


2.35 


0.00 


NA 


2.95 


12.24 


0.00 


Lack motivation 


13.19 


10.17 


NA 


15.14 


2.95 


14.38 


Too little time 


47.05 


33.31 


NA 


34.07 


25.51 


37.28 


N (unweighted) 


247 


17 


5 


373 


40 


15 



In 1985, too few Hispanic attenders reported wanting to go more to report statistics. Fewer than 10 percent of 
any group reported tickets sold out, discomfort, handicap, poor quality, publicity, work-related reasons, perform- 
ance times, or transience as reasons for not attending. 



145 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 4-5 

Weighted Percentages of Respondents Wishing to Attend Plays 

More, Citing Selected Reasons for Not Doing So: Whites (W), 

African-Americans (A), and Hispanics (H) 





Attended during 
previous 12 months 




Did not attend during 
previous 12 months 




W 


A 


H 


W 


A 


H 


1982 














Cost 


30.83 


69.64 


NA 


31.10 


24.18 


43.63 


Not available 


22.32 


7.81 


NA 


19.59 


20.23 


10.72 


Child care 


5.90 


0.00 


NA 


8.49 


13.64 


9.50 


Handicap 


2.79 


11.45 


NA 


5.34 


3.91 


0.00 


Too far to go 


14.91 


8.69 . 


NA 


15.35 


9.93 


6.07 


Transportation 


5.79 


12.08 


NA 


5.98 


8.51 


5.93 


Fear crime 


1.52 


11.45 


NA 


3.20 


0.00 


3.45 


Poor quality 


5.18 


0.00 


NA 


3.27 


10.29 


0.00 


Lack motivation 


11.33 


13.20 


NA 


13.85 


10.12 


11.53 


Too little time 


41.17 


42.89 


NA 


38.68 


14.80 


40.82 


N (unweighted) 


364 


15 


NA 


852 


48 


27 


1985 














Cost 


28.32 


43.07 


NA 


25.14 


37.64 


60.19 


Not available 


22.01 


18.70 


NA 


21.39 


9.54 


0.00 


Child care 


8.87 


8.06 


NA 


7.40 


7.96 


5.76 


Handicap 


2.58 


0.00 


NA 


3.82 


4.73 


0.00 


Too far to go 


14.44 


10.63 


NA 


17.25 


4.14 


0.00 


Transportation 


5.78 


10.63 


NA 


5.08 


13.51 


5.82 


Fear crime 


3.30 


8.06 


NA 


1.61 


4.14 


0.00 


Poor quality 


6.37 


4.66 


NA 


1.15 


0.00 


0.00 


Lack motivation 


12.23 


0.00 


NA 


12.81 


2.41 


25.30 


Too little time 


43.38 


21.54 


NA 


39.10 


38.72 


50.91 


N (unweighted) 


149 


12 


1 


358 


24 


13 



in 1982 and 1985, too few Hispanic attenders reported wanting to go more to report statistics. Under 10 percent 
of any group reported tickets sold out, discomfort, no one to go with, publicity, work-related reasons, performance 
times, or transience. 



146 



Appendix 

Appendix Table 4-6 

Weighted Percentages of Respondents Wishing to Attend Ballet 

Performances More, Citing Selected Reasons for Not Doing So: 

Whites (W), African-Americans (A), and Hispanics (H) 

Attended during Did not attend during 

previous 1 2 months previous 1 2 montlis 



W A H W A H 

1982 

Cost 43.21 NA NA 28.98 42.93 45.93 

Not available 27.91 NA NA 27.08 13.90 16.33 

No one to go with 8.15 NA NA 11.96 6.97 12.85 

Child care 6.07 NA NA 7.22 11.62 16.72 

Handicap 3.00 NA NA 8.27 3.09 10.01 

Too far to go 10.91 NA NA 15.88 13.61 20.38 

Transportation 8.37 NA NA 7.74 10.69 17.29 

Fear crime 1.76 NA NA 2.84 0.00 14.32 

Lack motivation 3.12 NA NA 13.35 2.61 21.39 

Too little time 27.35 NA NA 32.23 32.72 26.47 

N (unweighted) 100 2 4 468 31 22 

1985 

Cost 23.60 NA NA 33.31 36.98 43.72 

Not available 26.81 NA NA 17.72 0.00 7.50 

No one to go with 8.71 NA NA 10.90 9.07 14.03 

Child care 8.26 NA NA 11.23 6.19 12.60 

Handicap 1.87 NA NA 5.89 0.00 0.00 

Too far to go 10.61 NA NA 22.11 5.65 15.63 

Transportation 1.94 NA NA 6.93 0.00 5.43 

Fear crime 0.00 NA NA 3.18 11.84 0.00 

Lack motivation 8.26 NA NA 14.20 0.00 9.05 

Too little time 42.47 NA NA 35.29 51.09 28.31 

N (unweighted) 45 5 2 204 16 18 

In 1982 and 1985, too few African-American and Hispanic attenders reported wanting to go more to report 
statistics. Under 1 percent of any group reported tickets sold out, discomfort, poor quality, publicity, work-related 
reasons, performance times, or transience. 



147 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 4-7 

Weighted Percentages of Respondents Wishing to Attend Art 

Museums and Galleries More, Citing Selected Reasons for Not 

Doing So: Whites (W), African-Americans (A), and Hispanics (H) 





Attended during 
previous 12 months 


Did not attend during 
previous 12 months 




W 


A 


H 


W 


A 


H 


1982 














Cost 


6.95 


7.40 


28.98 


10.73 


22.66 


0.00 


Not available 


25.09 


5.79 


19.00 


24.81 


15.31 


7.11 


No one to go with 


5.23 


2.99 


10.22 


5.97 


6.62 


9.67 


Child care 


3.49 


15.86 ■ 


4.69 


5.44 


6.46 


15.43 


Too far to go 


16.96 


5.05 


28.05 


20.25 


10.46 


5.85 


Transportation 


5.23 


22.31 


6.97 


6.46 


11.68 


13.21 


Lack motivation 


14.58 


13.34 


2.81 


12.84 


17.76 


13.19 


Too little time 


51.01 


51.80 


67.86 


40.12 


30.85 


47.33 


N (unweighted) 


606 


26 


40 


812 


89 


48 



1985 



Cost 


9.40 


11.07 


11.00 


13.78 


17.37 


29.65 


Not available 


20.04 


0.00 


16.61 


23.98 


15.09 


0.00 


No one to go with 


2.39 


20.42 


9.98 


6.74 


5.37 


0.00 


Child care 


5.79 


0.00 


5.92 


5.39 


3.60 


8.54 


Too far to go 


25.20 


15.51 


7.81 


21.28 


6.80 


8.96 


Transportation 


6.30 


14.82 


3.33 


10.70 


18.94 


14.82 


Lack motivation 


14.32 


0.00 


21.71 


16.75 


7.94 


34.49 


Too little time 


48.29 


33.72 


45.77 


38.90 


52.72 


74.05 


N (unweighted) 


274 


52 


21 


323 


20 


18 



Under 1 percent of any group reported tickets sold out, discomfort, handicap, crime, poor quality, publicity, work- 
related reasons, performance times, or transience. 



148 



Appendix 



Appendix Table 5-1 

Results of Factor Analysis of Core and Other Activity 

Participation Measures: Rotated Factor Loadings 



Variables 



Factors 



.407 .343 .060 

.551 .029 .108 

.479 .001 .039 

■573 -.019 .047 

.607 .040 .064 

.602 .151 .054 

.341 .172 -.018 

-.002 -.045 716 

.002 .002 777 

.148 .210 .044 

.164 .023 .011 

.153 .144 .000 

.234 415 .092 

.084 .173 .034 

.038 i03 .150 

-.216 435 .065 

-.150 .136 .102 

.105 .187 ,565 

.155 .087 ,55Z 

.220 ,618 .167 

.070 452 -.044 

-.025 .622 -.077 



Based on data from November and December, 1982. Underlined variables are included in additive 
scales. 



Attend jazz 


.018 


Attend classical 


.307 


Attend opera 


-.031 


Attend musical 


.360 


Attend play 


.272 


Attend ballet 


.003 


Visit art exhibit 


.599 


Perform on 




instrument 


.049 


Perform: act, 




sing, dance 


.057 


Read novels, etc. 


.481 


Visit science or 




history museum 


.651 


Visit historic 




monument 


.686 


Read/listen to 




poetry 


.284 


Visit art/craft fair 


.670 


Art lessons 


.098 


Make pottery 


.297 


Do needlecrafts 


.397 


Work on play set 


.104 


Work on music set 


-.044 


Creative writing 


-.014 


Do photography 


.192 


Paint or draw 


.172 



149 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-2 

Results of Factor Analysis of Socialization Measures: 

Rotated Factor Loadings 



Factors 



Variables 



1 


2 


.097 


.748 


.174 


.747 


.128 


.777 


.208 


.604 


.494 


.320 


.693 


.082 


.542 


.072 


.322 


.268 


.667 


.113 


.559 


.083 


.680 


.220 


.618 


.256 



Parents listened to classical music 
Parents took child to art museums 
Parents took child to plays/concerts 
Parents encouraged child to read 
Instrumental/singing class/lessons 
Art class/lessons 
Acting class/lessons 
Ballet class/lessons 
Writing class/lessons 
Craft class/lessons 
Art appreciation class/lessons 
Music appreciation class/lessons 



Based on data from November and December, 1982. Underlined variables are included in additive 
scales. Only lessons taken before age of 1 8 are included. 



150 



Appendix 



Appendix Table 5-3 
Results of Factor Analysis of Music Preference Measures: 

Rotated Factor Loadings 



Factors 



Variables 



1 


2 


3 


.713 


.030 


-.004 


.665 


-.075 


-.092 


,695 


.115 


.165 


,353 


-.006 


.648 


,204 


.099 


.659 


,549 


.299 


.153 


,214 


.701 


.024 


,063 


.730 


.207 


.202 


.058 


.712 


,437 


.226 


.202 


,362 


.623 


.144 


.420 


.526 


-.107 


.214 


.430 


-.315 



Classical/chamber 

Opera 

Operetta/show tunes 

Jazz 

Soul/blues 

Big band 

Country western 

Bluegrass 

Rock 

Mood/easy listening 

Folk 

Barbershop 

Hymns/gospel 



Based on data from November and December, 1982. Underlined variables are included in additive 
scales. 



151 



Race, Ethnicity, and Participation in the Arts 

Table 5-4 

Means and Standard Deviations for Regression Variables for the 

Full Sample, and by Race and Spanish Origin, by Education, by 

Gender and by Age — Including Respondents without Data on 

Parental Education 



Art Attend 

N Lessons None music Attend Visit no jazz Perform Dovis TV arts 



FULLSAMPLE 2,255 1.166 1.005 1.396 0.533 2.177 0.436 0.116 0.723 1.347 

1.408 0.957 1.407 0.980 1.745 0.854 0.438 1.105 1.795 



RACE/SP ORIGIN 

White 1,908 1.240 1.134 ,1.509 0.571 2.208 0.483 0.116 0.762 1.404 

1.426 0.973 1.416 1.015 1.734 0.895 0.443 1.124 1.822 

African-American 230 0.864 0.860 0.720 0.365 1.203 0.197 0.094 0.449 1.092 

1.318 0.839 1.193 0.768 1.531 0.564 0.360 0.860 1.609 

Hispanic 117 0.667 0.900 1.084 0.305 1.597 0.214 0.166 0.708 1.027 

1.072 0.799 1.230 0.738 1.609 0.547 0.515 1.191 1.665 



EDUCATION 

11 & Less 607 0.467 0.550 0.785 0.143 0.945 0.112 0.075 0.273 0.675 

0.989 0.697 1.179 0.446 1.176 0.403 0.337 0.685 1.379 

12 Years 919 1.192 1.014 1.312 0.376 2.174 0.296 0.099 0.675 1.241 

1.338 0.858 1.276 0.763 1.648 0.667 0.380 1.027 1.662 

13-15 Yrs 390 1.646 1.517 1.744 0.811 2.689 0.665 0.158 1.034 1.644 

1.547 0.945 1.486 1.153 1.676 0.999 0.516 1.288 1.820 

16&0ver 339 1.766 1.708 2.275 1.305 3.416 1.101 0.183 1.279 2.450 

1.510 1.020 1.445 1.356 1.614 1.190 0.399 1.294 2.118 



GENDER 

Male 1,008 1.118 0.993 1.259 0.449 1.714 0.352 0.096 0.665 1.277 

1.348 0.884 1.341 0.882 1.576 0.752 0.398 1.065 1.747 

Female 1247 1.209 1.167 1.517 0.608 2.494 0.511 0.134 0.775 1.409 

1.458 1.010 1.453 1.054 1.805 0.929 0.471 1.137 1.835 



152 



Appendix 

Table 5-4 (Continued) 

Means and Standard Deviations for Regression Variables for the 

Full Sample, and by Race and Spanish Origin, by Education, by 

Gender and by Age — Including Respondents without Data on 

Parental Education 











Art 






Attend 










N 


Lessons 


None 


music 


Attend 


Visit 


no jazz 


Perform 


Dovis 


TV arts 


AGE 






















1M0 


729 


1.800 


1.229 


0.989 


0.515 


2.350 


0.361 


0.143 


1.071 


1.160 



1.632 0.897 1.121 0.950 1.725 0.771 0.479 1.325 1.559 



31-51 



768 1.098 1.089 1.618 0.632 2.325 0.535 0.135 0.739 1.457 
1.289 0.955 1.471 1.046 1.742 0.920 0.487 1.063 1.837 



Over 51 



758 0.603 0.937 1.569 0.448 1.696 0.407 0.065 0.358 1.419 
0.953 0.994 1.505 0.928 1.688 0.852 0.322 0.723 1.951 



153 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-5 

Regression Analyses Predicting Number of Performing-Arts 

Events Attended, 1982 Full Sample 



I.V. 




Jazz 
included 




Jazz not 
included 




African-American 


-.074 
d 




-.008 


-.103 
d 


-.033 
d 


Hispanic 


-.056 
d 




.013 


-.061 
d 


.012 


Female 






.088 
d 




.100 
d 


Age 






.095 
d 




.132 
d 


Education 






.299 
d 




.298 
d 


Occupation 






.109 
d 


- 


.105 
d 


Income X 10,000 






.132 
d 




.141 
d 


Single/divorced 






.110 
d 




.089 
d 


SMSA residence 






.063 
d 




.057 
d 


d.f. 

R squared 


15,012 
.008 




15,012 
.193 


15,012 
.013 


15,012 
.191 



Standardized regression coefficients, a: p less than or equal to .05 b: p less than or equal to .01 c: p 
less than or equal to .001 d: p less than or equal to .0001 



154 



Appendix 



Appendix Table 5-6 

Regression Analyses Predicting Scores 

on Fine-Arts Music Scale 



I.V. 


1 


2 


3 


African-American 


-.179 


-.093 


-.080 




d 


d 


d 


Hispanic 


-.067 
b 


.001 


.019 


Female 




.107 


.073 






d 


d 


Age 




.273 
d 


.312 
d 


Education 




.338 


.234 






d 


d 


Occupation 




.073 
c 


.057 
b 


ncomex 10,000 




.095 


.074 






d 





Single/divorced 




.035 


.007 


Lives in SMSA 




.076 


.055 






d 


b 


Childhood lessons 






.163 
d 


Parental guidance 


- 




.200 
d 


d.f. 


2,254 


2,254 


2,254 


R Squared 


.034 


.218 


.286 



Additive scale of number of kinds of fine-arts music respondents reported enjoying. Based on 
November/December 1982 subsample. a: p less than or equal to .05 b: p less than or equal to .01 c: 
p less tiian or equal to .001 d: p less than or equal to .0001 



155 



Race, Ethnicity, and Participation in tlie Arts 

Appendix Table 5-7 
Regression Analyses Predicting Scores on Television Arts Scale 



I.V. 


1 


2 


3 


African-American 


-.057 
b 


.017 


.028 


Hispanic 


-.047 
a 


.009 


.030 


Female 


> 


.054 
b 


.009 


Age 




.192 
d 


.233 
d 


Education 




.306 
d 


.181 
d 


Occupation 




.042 


.029 


Income X 10,000 




.100 
d 


.078 
c 


Single/divorced 


' 


.081 
c 


.047 
a 


Lives in SMSA 




.072 


.047 






c 


a 


Childhood lessons 






.173 
d 


Parental guidance 






.266 
d 


Hours watch TV 






.044 
a 


d.f. 

R Squared 


2,254 
.004 


2,254 
.140 


2,254 
.243 



Additive scale of number of kinds of artistic programs respondents reported watching on television 
during previous year. Models based on data from November/December 1982 subsample. a: p less 
than or equal to .05 b: p less than or equal to .01 c: p less than or equal to .001 d: p less than or equal 
to .0001 



156 



Appendix 

Appendix Table 5-8 
Regression Analyses Predicting Number of Performing Events 

Attended (Including Jazz) 



I.V. 


1 


2 


3 


4 


African-American 


-.067 
b 


.015 


.026 


.031 


Hispanic 


-.061 
b 


-.002 


.012 


.001 


Female 




.107 
d 


.080 
d 


.076 
d 


Age 




.119 
d 


.150 
d 


.065 
b 


Education 




.280 
d 


.196 
d 


.123 
d 


Occupation 




.125 
d 


.112 
d 


.090 
d 


Income X 10,000 




.144 
d 


.128 
d 


.098 
d 


Single/divorced 




.088 
d 


.065 
b 


.052 
b 


Metropolitan 




.051 


.034 


.017 


Parental guidance 




Wl 


.165 
d 


.076 
c 


Childhood lessons 






.129 
d 


.074 
c 


Hours watch TV 








-.077 
d 


Likes art music 


- 






.085 
c 

.257 
d 


Watch TV arts 








d.f. 

R Squared 


2,254 
.007 


2,254 
.182 


2,254 
.226 


2,254 
.299 



Standardized beta coefficients, a: p less than or equal to .05 b: p less than or equal to .01 c: p less 
than or equal to .001 d: p less than or equal to .0001 Models based on data from November/December 
1982subsample. 



157 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-9 
Regression Analyses Predicting Number of Performing Events 

Attended (Excluding Jazz) 



i.v. 



African-American 


-.107 
d 


-.022 


-.013 


-.006 


Hispanic 


-.070 
c 


-.008 


.004 


-.007 


Female 


w 


.117 
d 


.094 
d 


.089 
d 


Age 




.152 
d 


.178 
d 


.093 
d 


Education 




.275 
d 


.202 
d 


.129 
d 


Occupation 




.123 
d 


.112 
d 


.089 
d 


Income X 10,000 




.150 
d 


.136 
d 


.106 
d 


Single/divorced 




.068 

b - 


.048 
a 


.036 


Metropolitan 


- 


.043 
a 


.029 


.012 


Parental guidance 






.144 
d 


.058 
b 


Childhood lessons 






.106 
d 


.052 
a 


Hours watch TV 








-.079 
d 


Likes art music 








.101 
d 


Watch TV arts 








.234 
d 


d.f. 

R Squared 


2,254 
.014 


2,254 
.183 


2,254 
.215 


2,254 
.283 



Standardized beta coefficients, a: p less than or equal to .05 b: p less than or equal to .01 c: p less 
than or equal to .001 d: p less than or equal to .0001 Based on data from November/December 1 982 
subsample. 



158 



Appendix 

Appendix Table &-10 
Regression Analyses Predicting Number of Visually Oriented 

Consumption Activities 



I.V. 


1 


2 


3 


4 


African-American 


-.199 
d 


-.115 
d 


-.101 
d 


-.094 
d 


Hispanic 


-.088 
d 


-.024 


-.004 


-.016 


Female 




.255 
d 


.219 
d 


.213 
d 


Age 




-.059 
b 


-.013 


-.103 
d 


Education 




.371 
d 


.260 
d 


.184 
d 


Occupation 




.080 
d 


.063 
c 


.040 
a 


Income X 10,000 




.084 
d 


.062 
b 


.031 


Single/divorced 




-.012 


-.043 
a 


-.056 
b 


Metropolitan 




.024 


.002 


-.016 


Parental guidance 






.208 
d 


.118 
d 


Childhood lessons 






.186 
d 


.129 
d 


Hours watch TV 


, 






-.077 
d 


Likes art music 








.106 
d 


Watch TV arts 








.247 
d 


d.f. 

R Squared 


2,254 
.043 


2,254 
.310 


2,254 
.390 


2,254 
.465 



Standardized beta coefficients, a: p less than or equal to .05 b: p less than or equal to .01 c: p less 
than or equal to .001 d: p less than or equal to .0001 Based on November/December 1 982 subsample. 



159 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-11 

Regression Analyses Predicting Number 

of Kinds of Performance Activities 



I.V. 


1 


2 


3 


4 


African-American 


-.016 


-.004 


.003 


.007 


Hispanic 


.025 


.038 


.049 
a 


.045 
a 


Female 


> 


.050 


.035 


.031 


Age 




-.039 


-.010 


-.049 


Education 




- .088 


.044 


.013 


Occupation 




.039 


.033 


.025 


Income X 10,000 




-.042 


-.051 
a 


-.062 
b 


Single/divorced 




.034 


.020 


.015 


Metropolitan 


' 


-.030 


-.041 


-.048 
a 


Parental guidance 






.063 
b 


.028 


Childhood lessons 






.112 
d 


.088 




Hours watch TV 








-.015 


Likes art music 








.059 
a 


Watch TV arts 








.085 
b 


d.f. 

R Squared 


2,254 
.000 


2,254 
.016 


2,254 
.032 


2,254 
.006 



Standardized beta coefficients, a: p less than or equal to .05 b: p less than or equal to .01 c: p less 
than or equal to .001 d: p less than or equal to .0001 Data from November/December 1 982 subsample. 



160 



Appendix 



Appendix Table 5-12 

Regression Analyses Predicting Number of Kinds of 

Nonperformance Creative Activities 



.V. 



African-American 


-.091 
d 


-.056 
b 


-.035 


-.032 


Hispanic 


-.011 


.016 


.043 


.034 


Female 




.074 


.038 


.039 






c 


a 


a 


Age 




-.151 


-.074 


-.133 






d 


c 


d 


Education 




.215 


.104 


.050 






d 


d 


a 


Occupation 




.098 
d 


.082 
d 


.062 
b 


Income X 10,000 




-.024 


-.047 
a 


-.070 




Single/divorced 




.096 
d 


.059 
b 


.049 
a 


Metropolitan 




-.018 


-.044 
a 


-.056 
b 


Parental guidance 






148 
d 


.081 
c 


Childhood lessons 






.300 
d 


.261 
d 


Hours watch TV 








-.088 
d 


Likes art music 








.039 


Watch TV arts 








.205 
d 


d.f. 

R Squared 


2,254 
.007 


2,254 
.149 


2,254 
.258 


2,254 
.301 



Standardized beta coefficients, a: p less than or equal to .05 b: p less than or equal to .01 c: p less 
than or equal to .001 d: p less than or equal to .0001 Based on data from November/December 1982 
subsample. 



161 



Race, Ethnicity, and Participation in the Arts 

Appendix Table &-1 3 

Regression Analyses Predicting Number of Kinds of Art Lessons 

Taken Before Age 18 and Number of Kinds of Activities with 

Parents as Child, by Race 







Kinds of lessons 




Activities with paren 
W A 


ts 


I.V.S 


W 


A 


H 


H 


Age 


-.016 


-.025 


-.015 


.009 


-.003 


.007 




.002 


.066 


.008 


.001 


.004 


.005 




-.194 


-.333 


-.206 


.159 


-.064 


.132 




d 


d 




d 






Female 


.185 


-.106 


.168 


.223 


.295 


.048 




.067 


.204 


.253 


.043 


.124 


.164 




.064 


-.039 


.073 


.114 


.166 


.030 




b 






d 


a 




Father's 


.070 


.061 


.109 


.100 


.098 


.068 


education 


.014 


.047 


.052 


.009 


.029 


.033 




.161 


.130 


.313 


.340 


.317 


.283 




d 




a 


d 


d 


a 


Mother's 


.085 


.061 


-.055 


.090 


.085 


.053 


education 


.016 


.051 


.064 


.011 


.031 


.041 




.171 


.128 


-.127 


.269 


.273 


.181 




d 






d 


b 




d.f. 


1,525 


140 


83 


1,525 


140 


83 


Adj. R squared 


.187 


.217 


.047 


.267 


.329 


.154 



a: p less than or equal to .05 b: p less than or equal to .01 c: p less than or equal to .001 d: p less 
than or equal to .0001 Based on data from November/December 1 982 subsample, respondents with 
information on parents' education only. 



162 



Appendix 

Appendix Table 5-14 

Regression Analyses Predicting Number of Kinds of Art Music 

and Related Genres Enjoyed, by Race 

Model 1 Model 2 



I.V.S 


W 


A 


H 


W 


A 


H 


Age 


.023 


.004 


.003 


.026 


.011 


.003 




.002 


.005 


.008 


.002 


.005 


.008 




.296 


.069 


.046 


.337 


.174 


.036 




d 






d 


a 




Female 


.353 


.110 


.201 


.243 


.089 


.198 




.058 


.153 


.229 


.057 


.140 


.220 




.124 


.046 


.082 


.086 


.037 


.080 




d 






d 






Education 


.175 


.070 


.077 


.122 


.041 


.038 




.012 


.028 


.038 


.013 


.025 


.038 




.346 


.225 


.243 


.241 


.131 


.122 




d 


a 


a 


d 






Occupation 


.190 


.574 


-.186 


.175 


.265 


-.193 




.068 


.196 


.278 


.066 


.183 


.266 




.065 


.205 


-.067 


.060 


.095 


-.069 




b 


b 




b 






Income 


.098 


.008 


-.052 


.080 


.023 


-.059 


X 10,000 


.021 


.075 


.103 


.021 


.069 


.100 




.105 


.008 


-.051 


.086 


.022 


-.058 




d 






d 






Single/divorced 


.128 


-.106 


.405 


.023 


-.074 


.280 




=073 


.178 


.268 


.071 


.161 


.261 




.039 


-.041 


.155 


.007 


-.029 


.107 


Lives in SMSA 


.201 


.358 


.305 


.153 


.222 


.239 




.062 


.189 


.411 


.060 


.172 


.395 




.068 


.124 


.069 


.052 


.077 


.054 




b 




- 


a 







163 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-14 (Continued) 

Regression Analyses Predicting Number of Kinds of Art Music 

and Related Genres Enjoyed, by Race 

Model 1 Model 2 



I.V.S W A H W A H 

Parental .261 .491 .350 

guidance .033 .095 .148 

.180 .345 .227 

d d a 



Childhood lessons 








.158 

.023 

.159 

d 


.176 

.061 

.195 

b 


.205 
.114 
.179 


d.f. 


1,907 


229 


116 


1,907 


229 


116 


Adj. R Squared 


.214 


.133 


.028 


.272 


.290 


.108 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December subsample. 



164 



Appendix 

Appendix Table 5-15 

Regression Analyses Predicting Number of Kinds of Television 

Arts Program Watched in Previous 12 Months, by Race 

Model 1 Model 2 



I.V.S 


W 


A 


H 


W 


A 


H 


Age 


.020 


.003 


.005 


.024 


.015 


.001 




.002 


.007 


.012 


.002 


.007 


.011 




.204 


.037 


.046 


.245 


.174 


.015 




d 






d 


a 




Female 


.255 


.099 


-.096 


.073 


.034 


-.074 




.078 


.208 


.316 


.075 


.188 


.285 




.070 
c 


.031 


-.029 


.020 


.010 


-.022 


Education 


.206 


.064 


.096 


.124 


.026 


.017 




.016 


.037 


.052 


.017 


.034 


.049 




.317 


.151 


.225 


.191 


.061 


.041 




d 






d 






Occupation 


.120 


.594 


-.291 


.114 


.169 


-.303 




.091 


.266 


.383 


.087 


.243 


.347 




.032 


.157 
a 


-.077 


.030 


.045 


-.080 


Income 


.116 


.166 


.037 


.090 


.188 


.048 


X 10.000 


.029 


.102 


.142 


.027 


.091 


.129 




.097 


.117 


.027 


.075 


.132 


.034 




d 






c 


a 




Single/divorced 


.335 


.323 


.370 


.177 


.363 


.177 




.098 


.242 


.369 


.093 


.213 


.338 




.080 
c 


.094 


.104 


.042 


.105 


.050 


Lives in SMSA 


.256 


.477 


-.230 


.183 


.244 


-.303 




.083 


.256 


.566 


.079 


.228 


.516 




.067 


.123 


-.038 


.048 


.063 


-.050 




b 






a 







165 



Race, Ethnicity, and Participation in tlie Arts 

Appendix Table 5-1 5 (Continued) 

Regression Analyses Predicting Number of Kinds of Television 

Arts Program Watched in Previous 12 Months, by Race 

Model 1 Model 2 



I.V.S W A H W A H 

Parental .451 .693 .875 

guidance .044 .126 .192 

.241 .361 .419 

d d d 

Childhood lessons .215 .294 .223 

.030 .082 .147 

.168 .240 .143 

d 

Hours TV .030 .037 -.001 

.018 .028 .069 

.035 .077 -.001 



d.f. 1,907 229 116 


1,907 


229 


116 


Adj. R Squared .151 .121 -.004 


.236 


.319 


.186 


First row is unstandardized regression coefficient. 








Second row is standard error. 








Third row is standardized regression coefficient. 








Fourth row indicates significance: 








a less than or equal to .05 








b less than or equal to .01 








c less than or equal to .001 








d less than or equal to .0001 








Based on data from November/December 1982 subsample. 









166 



Appendix 

Appendix Table 5-16 
Regression Analyses Predicting Number of Kinds of Performing 
Events Attended in Previous 12 Months, Including Jazz, by Race 

Model 1 Model 2 Model 3 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Age 


.006 


.002 


.001 


.008 


.004 


.000 


.003 


.002 


.000 




.001 


.003 


.005 


.001 


.003 


.005 


.001 


.003 


.005 




.118 


.042 


.027 


.151 


.097 


.011 


.063 


.045 


.001 




d 






d 






b 






Female 


.257 


.018 


.127 


.192 


.007 


.132 


.175 


.004 


.145 




.042 


.093 


.137 


.042 


.091 


.134 


.040 


.090 


.131 




.126 


.012 


.086 


.095 


.005 


.090 


.086 


.002 


.098 




d 






d 






d 






Education 


.112 


.045 


.003 


.081 


.034 


-.016 


.052 


.029 


-.019 




.009 


.017 


.022 


.009 


.016 


.023 


.009 


.016 


.023 




.309 


.222 


.014 


.223 


.168 


-.082 


.144 


.145 


-.099 




d 


b 




d 


a 




d 






Ocxjupation 


.198 


.529 


.269 


.189 


.417 


.266 


.141 


.386 


.296 




.050 


.119 


.166 


.048 


.119 


.162 


.046 


.117 


.159 




.094 


.294 


.161 


.090 


.232 


.159 


.067 


.214 


.177 




d 


d 




d 


c 




b 


b 




Income 


.086 


.113 


.121 


.076 


.117 


.124 


.055 


.096 


.113 


X 10,000 


.016 


.046 


.061 


.015 


.044 


.061 


.015 


.044 


.059 




.128 


.168 


.196 


.113 


.174 


.201 


.082 


.143 


.183 




d 


a 




d 


b 


a 


c 


a 




Single/ 


.242 


-.013 


.151 


.181 


-.093 


.107 


.150 


-.038 


.087 


divorced 


.053 


.108 


.160 


.052 


.112 


.159 


.050 


.103 


.155 




.104 


-.008 


.096 


.078 


-.050 


.068 


.064 


-.023 


.055 




d 






c 






b 






Lives in 


116 


-.046 


-.203 


.088 


-.093 


-.220 


.051 


-.126 


-.154 


SMSA 


.045 


.114 


.245 


.041 


.112 


.241 


.042 


.110 


.237 




.055 


-.025 


-.077 


.042 


-.050 


-.083 


.024 


-.068 


-.058 




a 






a 












Parental 








.153 


.191 


.204 


.069 


.096 


.088 


guidance 








.025 


.062 


.090 


.024 


.066 


.096 










.147 


.209 


.220 


.066 


.105 


.096 










d 


b 


a 


b 







167 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-16 (Continued) 
Regression Analyses Predicting Number of Kinds of Performing 
Events Attended in Previous 12 {Months, including Jazz, by Race 

Model 1 Model 2 Model 3 



I.V.S W AH W 


A 


H 


W 


A 


H 


Childhood lessons .092 


.053 


.050 


.053 


.016 


.023 


.017 


.040 


.069 


.016 


.040 


.068 


.129 


.092 


.073 


.075 


.027 


.033 


d 






d 






Hours TV 






-.037 


-.007 


-.018 








.010 


.013 


.032 








-.078 
c 


-.030 


-.050 


Art music 






.063 


.037 


-.020 








.018 


.046 


.066 








.088 
c 


.058 


-.033 


TV art programs 






.143 


.109 


.142 




' 




.014 


.035 


.051 


- 






.257 


.229 


.320 








d 


b 


b 


d.f. 1,907 229 116 1,907 


229 


116 


1.907 


229 


116 


R Squared .188 .234 .039 .226 


.280 


.081 


.301 


.315 


.134 


First row is unstandardized regression coefficient. 












Second row is standard error. 












Third row is standardized regression coefficient. 












Fourth row indicates significance: 












a less than or equal to .05 












b less than or equal to .01 












c less than or equal to .001 












d less than or equal to .0001 












Based on data from November/December 1982 subsample. 











168 



Appendix 

Appendix Table 5-17 
Regression Analyses Predicting Number of Kinds of Performing 
Events Attended in Previous 12 Months, Excluding Jazz, by Race 

Model 1 Model 2 Models 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Age 


.007 


.007 


.002 


.008 


.008 


.001 


.004 


.006 


.001 




.001 


.002 


.004 


.001 


.002 


.004 


.001 


.002 


.004 




.140 


.237 


.061 


.168 


.258 


.041 


.079 


.205 


.034 




d 


b 




d 


b 




b 


b 




Female 


.239 


.070 


.095 


.190 


.057 


.100 


.172 


.057 


.108 




.037 


.067 


.100 


.037 


.067 


.098 


.036 


.066 


.098 




.134 


.062 


.087 


.106 


.050 


.091 


.096 


.050 


.098 




d 






d 






d 






Education 


.097 


.038 


.010 


.073 


.032 


-.004 


.048 


.028 


-.005 




.008 


.012 


.016 


.008 


.012 


.017 


.008 


.012 


.017 




.304 


.254 


.073 


.230 


.215 


-.027 


.149 


.188 


-.039 




d 


b 




d 


b 




d 


a 




Occupation 


.174 


.411 


.196 


.168 


.358 


.194 


.125 


.331 


.210 




.044 


.086 


.122 


.043 


.087 


.119 


.042 


.086 


.118 




.094 


.311 


.158 


.090 


.270 


.156 


.067 


.250 


.169 




d 


d 




d 


d 




b 


c 




Income 


.077 


.113 


.099 


.069 


.112 


.102 


.051 


.098 


.095 


X 10,000 


.014 


.033 


.045 


.014 


.033 


.044 


.013 


.032 


.044 




.131 


.227 


.216 


.117 


.225 


.222 


.086 


.198 


.209 




d 


c 


a 


d 


c 


a 


d 


b 


a 


Single/ 


.158 


.041 


.098 


.111 


.046 


.066 


.085 


.025 


.055 


divorced 


.047 


.078 


.117 


.047 


.077 


.116 


.044 


.076 


.116 




.077 
c 


.034 


.084 


.054 
a 


.038 


.057 


.041 


.021 


.047 


Lives in 


.095 


-.155 


-.003 


.074 


-.171 


-.014 


.040 


-.194 


.023 


SMSA 


.040 


.083 


.180 


.039 


.082 


.176 


.038 


.081 


.177 




.051 


-.113 


-.002 


.039 


-.125 


-.007 


.022 


-.142 


.012 




a 








a 






a 




Parental 








.118 


.122 


.163 


.044 


.054 


.099 


guidance 








.022 


.045 


.066 


.022 


.048 


.071 










.128 


.181 


.239 


.048 


.080 


.145 










d 


b 


a 


a 







169 



Race, Ethnicity, and Participation in the Arts 

Appendix Table &-17 (Continued) 
Regression Analyses Predicting Number of Kinds of Performing 
Events Attended in Previous 12 Months, Excluding Jazz, by Race 

Model 1 Model 2 Models 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Childhood lessons 








.070 


.001 


.031 


.037 


-.025 


.016 










.015 


.029 


.051 


.015 


.030 


.051 










.112 


.002 


.062 


.058 


-.058 


.032 










d 






a 






Hours TV 














-.032 

.009 

-.076 

d 


-.007 

.010 

-.043 


-.010 

.024 

-.037 


Art music 














.067 

.016 

.105 

d 


.038 
.034 
.081 


-.012 

.049 

-.027 


TV art programs 












- 


.118 

.012 

.239 

d 


.070 

.026 

.198 

b 


.079 

.038 

.241 

a 


d.f. 


1,907 


229 


116 


1,907 


229 


116 


1,907 


229 


116 


R Squared 


.182 


.258 


.058 


.210 


.278 


.106 


.283 


.309 


.125 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



170 



Appendix 

Appendix Table 5-18 

Regression Analyses Predicting Number of Kinds of Visually 

Oriented Consumption Activities in Previous 12 Months, by Race 



I.V.S 

Age 



Female 



Education 



Occupation 



Income 
X 10,000 



Single/ 
divorced 



Lives in 
SMSA 



Parental 
guidance 



Model 1 Model 2 Model 3 



w 


A 


H 


W 


A 


H 


W 


A 


H 


-.007 


-.009 


-.002 


-.002 


-.008 


-.001 


-.011 


-.005 


-.001 


.002 


.006 


.010 


.002 


.006 


.009 


.002 


.006 


.008 


-.071 


-.109 


-.017 


-.025 


-.009 


-.015 


-.118 


-.062 


-.012 


b 












d 






1.015 


.537 


.182 


.874 


.503 


.152 


.838 


.518 


.197 


.067 


.178 


.276 


.065 


.158 


.250 


.061 


.157 


.215 


.292 


.174 


.057 


.252 


.163 


.047 


.241 


.168 


.061 


d 


b 




d 


b 




d 


b 




.256 


.089 


.141 


.188 


.051 


.081 


.136 


.039 


.078 


.014 


.032 


.045 


.015 


.029 


.043 


.014 


.028 


.037 


.413 


.222 


.343 


.303 


.127 


.197 


.220 


.097 


.190 


d 


b 


b 


d 






d 




a 


.129 


1.009 


.732 


.111 


.613 


.721 


.024 


.531 


.890 


.079 


.228 


.334 


.075 


.207 


.302 


.071 


.204 


.260 


.036 


.281 


.201 


.031 


.171 


.198 


.007 


.148 


.244 




d 


a 




b 


a 




b 


c 


.085 


.130 


.048 


.062 


.146 


.016 


.026 


.116 


.002 


.025 


.088 


.124 


.024 


.078 


.113 


.022 


.076 


.097 


.074 


.097 


.035 


.054 


.109 


.012 


.022 


.086 


.002 


c 






b 












.036 


-.427 


-.107 


-.100 


-.386 


-.359 


-.151 


-.430 


-.436 


.084 


.207 


.323 


.081 


.183 


.296 


.076 


.180 


.254 


.009 


-.130 
a 


-.031 


-.025 


-.118 
a 


-.105 


-.038 
a 


-.131 
a 


-.127 


.092 


-.012 


-.361 


.030 


-.181 


-.515 


-.037 


-.233 


-.423 


.072 


.219 


.494 


.069 


.195 


.448 


.064 


.191 


.388 


.025 


-.003 


-.062 


.008 


-.049 


-.089 


-.010 


-.063 


-.073 








.326 


.658 


.421 


.178 


.485 


.009 








.038 


.108 


.168 


.037 


.114 


.157 








.183 


.360 


.209 


.100 


.266 


.004 








d 


d 


a 


d 


d 





171 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-18 (Continued) 

Regression Analyses Predicting Number of Kinds of Visually 

Oriented Consumption Activities in Previous 12 Months, by Race 

Model 1 Model 2 Model 3 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Childhood lessons 








.214 


.203 


.485 


.146 


.143 


.384 










.026 


.070 


.129 


.025 


.071 


.120 










.176 


.174 


.323 


.120 


.123 


.256 










d 


b 


c 


d 


a 


c 


Hours TV 














-.065 

.015 

-.079 

d 


-.029 

.023 

-.064 


.057 
.052 
.074 


Art music 














.136 

.028 

.111 

d 


.113 
.080 
.088 


-.025 

.107 

-.019 


TV art programs 












' 


.234 

.021 

.246 

d 


.163 

.061 

.171 

b 


.476 

.083 

.494 

d 


d.f. 


1,907 


229 


116 


1,907 


229 


116 


1,907 


229 


116 


R Squared 


.302 


.290 


.176 


.367 


.448 


.330 


.445 


.476 


.511 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on November/December 1982 subsample. 



172 



Appendix 

Appendix Table 5-19 

Regression Analyses Predicting Number of Kinds of Performance 

Activities in Previous 12 Months, by Race 



Model 1 



Model 2 



Model 3 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Age 


-.001 


-.000 


-.004 


-.000 


-.000 


-.004 


-.001 


-.000 


-.004 




.001 


.002 


.004 


.001 


.002 


.003 


.001 


.002 


.003 




-.045 


-.018 


-.123 


-.014 


-.008 


-.155 


-.058 
a 


-.016 


-.140 


Female 


.050 


-.001 


.090 


.034 


-.002 


.091 


.028 


-.011 


.101 




.020 


.050 


.096 


.020 


.050 


.094 


.020 


.051 


.095 




.057 
a 


-.001 


.087 


.039 


-.003 


.088 


.032 


-.006 


.098 


Education 


.017 


.003 


-.016 


.010 


.002 


-.029 


.004 


.002 


-.030 




.004 


.009 


.016 


.005 


.009 


.016 


.005 


.009 


.016 




.109 


.031 


-.123 


.061 


.020 


-.223 


.024 


.024 


-.224 




d 






a 












Occupation 


.025 


.099 


.042 


.024 


.088 


.039 


.015 


.092 


.039 




.024 


.064 


.116 


.024 


.066 


.114 


.024 


.067 


.115 




.027 


.117 


.036 


.026 


.105 


.034 


.016 


.109 


.033 


Income 


-.017 


-.017 


.091 


-.020 


-.017 


.091 


-.024 


-.018 


.085 


X 10,000 


.008 


.024 


.043 


.007 


.025 


.042 


.007 


.025 


.043 




-.059 


-.053 


.212 


-.068 


-.052 


.212 


-.082 


-.055 


.197 




a 




a 


b 




a 


b 






Single/ 


.020 


.084 


.186 


.002 


.085 


.148 


-.003 


.084 


.151 


divorced 


.026 


.058 


.112 


.025 


.058 


.111 


.025 


.059 


.113 




.019 


.109 


.169 


.002 


.110 


.135 


-.002 


.109 


.137 


Lives in 


-.035 


.002 


.188 


-.043 


-.002 


.170 


-.050 


-.008 


.204 


SMSA 


.022 


.061 


.172 


.022 


.062 


.168 


.022 


.063 


.172 




-.038 


.002 


.102 


-.046 
a 


-.002 


.092 


-.054 
a 


-.009 


.110 


Parental 








.026 


.019 


.133 


.010 


.015 


.111 


guidance 








.012 


.034 


.063 


.012 


.037 


.070 










.057 


.044 


.207 


.022 


.035 


.172 










a 




a 








Childhood lessons 








.037 


.004 


.055 


.029 


-.000 


.054 










.008 


.022 


.048 


.008 


.023 


.050 










.118 


.014 


.115 


.093 


-.001 


.111 










d 






c 







173 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-19 (Continued) 

Regression Analyses Predicting Number of Kinds of Performance 

Activities in Previous 12 Months, by Race 

Model 1 Model 2 Model 3 



I.V.S W A H W A H W A H 

Hours TV -.004 .006 -.012 

.005 .008 .023 

-.021 .057 -.047 

Art music .022 .011 -.038 

.009 .026 .048 

.070 .036 -.090 
a 

TV art programs .022 -.000 .042 

.007 .020 .037 

.089 -.001 .135 
b 

d.f. 1,907 229 116 1,907 229 116 1.907 229 116 

R Squared .018 .002 .028 .034 -.004 .075 .047 -.014 .063 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1 982 subsample. 



174 



Appendix 



Appendix Table 5-20 

Regression Analyses Predicting Number of Kinds of 

Nonperformance Activities in Previous 12 Months, by Race 



Model 1 Model 2 Model 3 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Age 


-.010 


-.003 


-.011 


-.005 


.004 


-.012 


-.009 


.003 


-.013 




.001 


.004 


.008 


.001 


.004 


.007 


.001 


.004 


.007 




-.168 


-.054 


-.157 


-.091 


.092 


-.176 


-.150 


.060 


-.185 




d 






d 






d 






Female 


.223 


-.041 


-.042 


.124 


.001 


-.038 


.116 


.010 


.010 




.048 


.107 


.219 


.045 


.098 


.203 


.044 


.099 


.184 




.099 


-.024 


-.018 


.055 


.000 


-.016 


.052 


.006 


.004 




d 






b 






b 






Education 


.099 


.039 


.040 


.052 


.026 


-.008 


.028 


.022 


-.011 




.010 


.019 


.036 


.010 


.018 


.035 


.010 


.018 


.032 




.247 


.172 


.132 


.129 


.117 


-.025 


.069 


.099 


-.037 




d 


a 




d 






b 






Occupation 


.129 


.702 


.311 


.121 


.527 


.303 


.073 


.499 


.376 




.056 


.137 


.266 


.053 


.128 


.245 


.052 


.129 


223 




.056 


.348 


.115 


.052 


.261 


.112 


.031 


.247 


.139 




a 


d 




a 


d 






d 




Income 


-.023 


-.043 


.020 


-.039 


-.015 


.019 


-.057 


-.025 


-.011 


X 10,000 


.018 


.053 


.098 


.017 


.048 


.092 


.016 


.048 


.083 




-.031 


-.056 


.020 


-.053 
a 


-.019 


.019 


-.076 
c 


-.033 


-.011 


Single/ 


.313 


-.022 


.075 


.206 


.002 


-.064 


.179 


-.016 


-.094 


divorced 


.060 


.124 


.256 


.057 


.113 


.240 


.055 


.114 


.217 




.121 


-.012 


.030 


.080 


.001 


-.025 


.069 


-.008 


-.037 




d 






c 






b 






Lives in 


-.064 


.072 


-.171 


-.111 


-.044 


-.236 


-.142 


-.059 


-.070 


SMSA 


.051 


.132 


.392 


.048 


.121 


.363 


.047 


.121 


.332 




-.028 


.035 


-.040 


-.047 
a 


-.021 


-.055 


-.060 
b 


-.028 


-.016 


Parental 








.158 


.067 


.481 


.088 


.009 


.210 


guidance 








.027 


.066 


.136 


.027 


.072 


.134 










.137 


.065 


.323 


.077 


.009 


.141 










d 




c 


c 







175 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-20 (Continued) 

Regression Analyses Predicting Number of Kinds of 

Nonperformance Activities in Previous 12 Months, by Race 

Model 1 Model 2 Model 3 



I.V.S 


W 


A 


H 


W 


A 


H 


W 


A 


H 


Childhood lessons 








.229 


.268 


.202 


.200 


.249 


.146 










.018 


.043 


.104 


.018 


.044 


.096 










.291 


.410 


.182 


.254 


.381 


.131 










d 


d 




d 


d 




Hours TV 














-.046 

.011 

-.088 

d 


-.013 

.015 

-.050 


-.026 

.044 

-.046 


Art music 














.041 

.020 

.051 

a 


.026 
.051 
.035 


-.121 

.092 

-.125 


TV art programs 










- 




.120 

.015 

.194 

d 


.061 
.039 
.115 


.362 

.071 

.503 

d 


d.f. 


1,907 


229 


116 


1,907 


229 


116 


1,907 


229 


116 


R Squared 


.158 


.188 


.053 


.257 


.331 


.194 


.299 


.336 


.347 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



176 



Appendix 

Appendix Table 5-21 

Regression Analyses Predicting Number of Kinds of Art Lessons 

Taken Before Age 18, Number of Kinds of Activities with Parents 

as Child, and Number of Kinds of Television Arts Programs 

Watched, by Gender 





Lessons 


Parents 


Arts on TV 


Arts 


on TV 


I.V.S 


M 


F 


M 


F 


M 


F 


M 


F 


African- 


-.251 


-.595 


-.248 


-.334 


.163 


.036 


.167 


.159 


American 


.126 


.118 


.088 


.087 


.172 


.154 


.164 


.145 




-.058 


-.133 


-.088 


-.108 


.029 


.006 


.030 


.028 




a 


d 


b 


d 










Hispanic 


-.689 


-.707 


-.260 


-.472 


.243 


-.100 


.357 


.134 




.171 


.177 


.120 


.131 


.232 


.228 


.222 


.214 




-.117 


-.105 


-.068 


-.101 


.032 


-.012 


.047 


.016 




d 


d 


a 


c 










Age 


-.029 


-.028 


-.009 


-.008 


.021 


.017 


.026 


.021 




.002 


.002 


.002 


.002 


.003 


.003 


.003 


.003 




-.381 


-.365 


-.178 


-.141 


.218 


.176 


.265 


.213 




d 


d 


d 


d 


d 


d 


d 


d 


Education 










.149 

.020 

.291 

d 


.215 

.021 

.324 

d 


.098 

.019 

.191 

d 


.120 

.021 

.180 

d 


Occupation 










.460 

.123 

.127 

c 


-.069 

.114 

-.018 


.322 

.119 

.089 

b 


-.035 

.108 

-.009 


Income x 










.063 


.161 


.040 


.133 


10.000 










.040 
.054 


.037 

.131 

d 


.038 
.034 


.035 

.108 

d 


Single/ 










.462 


.220 


.349 


.063 


divorced 










.130 

.119 

c 


.121 
.052 


.124 

.090 

b 


.113 
.015 



177 



Race, Ethnicity, and Participation in tlie Arts 

Appendix Table 5-21 (Continued) 

Regression Analyses Predicting Number of Kinds of Art Lessons 

Taken Before Age 18, Number of Kinds of Activities with Parents 

as Child, and Number of Kinds of Television Arts Programs 

Watched, by Gender 





Lesi 


sons 


Pai 


rents 


Arts on TV 


Arts on TV 


I.Vs 


M 


F 


M 


F 


M 


F 


M 


F 


Lives in 
SMSA 




> 






.225 

.114 

.060 

a 


.313 

.107 

.080 

b 


.120 
.109 
.032 


.236 

.100 

.060 

a 


Childhood 
lessons 






> 








.239 

.042 

.184 

d 


.205 

.036 

.163 

d 


Home 
activities 














.426 

.064 

.216 

d 


.536 

.053 

.295 

d 


Hours TV/ 
day 














.041 
.025 
.047 


.031 
.018 
.043 


d.f. 

R Squared 


1,007 
.153 


1,246 
.148 


1,007 
.039 


1,246 
.034 


1,007 
.141 


1,246 
.150 


1,007 
.219 


1,246 
.265 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



178 



Appendix 

Appendix Table 5-22 

Regression Analyses Predicting Number of Kinds of Art Music 

and Related Genres Enjoyed, by Gender 



Mode! 1 Model 2 

I.V.S M F M 



African- 


-.305 


-.507 


-.297 


-.398 


American 


.125 


.118 


.122 


.112 




-.071 


-.113 


-.069 


-.089 




a 


d 


a 


c 


Hispanic 


.111 


-.094 


.184 


.059 




.169 


.175 


.164 


.166 




.019 


-.014 


.032 


.009 


Age 


.020 


.021 


.024 


.024 




.002 


.002 


.002 


.002 




.273 


.269 


.315 


.307 




d 


d 


d 


d 


Education 


.135 


.178 


.105 


.111 




.014 


.016 


.014 


.016 




.343 


.338 


.268 


.211 




d 


d 


d 


d 


Occupation 


.324 


.119 


.235 


.124 




.089 


.088 


.088 


.083 




.116 


.039 


.085 


.040 




c 




b 




Income x 


.077 


.096 


.063 


.075 


10,000 


.029 


.028 


.028 


.027 




.085 


.099 


.069 


.076 




b 


c 


a 


b 


Single/ 


.039 


.176 


-.024 


.063 


divorced 


.094 


.093 


.092 


.088 




.013 


.052 


-.008 


.019 


Lives in 


.230 


.214 


.171 


.160 


SMSA 


.083 


.082 


.081 


.078 




.080 


.069 


.060 


.052 




b 


b 


a 


a 



179 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-22 (Continued) 

Regression Analyses Predicting Number of Kinds of Art Music 

and Related Genres Enjoyed, by Gender 



I.V.S 


Model 1 
M 


F 


M 


Model 2 


F 


Childhood 
lessons 






.155 

.031 

.155 

d 




.164 

.028 

.164 

d 


Parental 
guidance 






.217 

.047 

.143 

d 




.341 

.041 

.237 

d 


d.f. 

R Squared 


1,007 
.223 


1.246 
.201 


1,007 
.270 




1.246 
.287 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



180 



Appendix 

Appendix Table 5-23 

Regression Analyses Predicting Number of Kinds of Performing 

Events Attended, Including Jazz, by Gender 





Model 1 


1 


Model 2 


f 


iAodeiS 


I.V.S 


M 


F 


M 


F 


M 


F 


African- 


.108 


-.003 


.113 


.053 


.101 


.088 


American 


.085 


.086 


.083 


.085 


.080 


.081 




.038 


-.001 


.040 


.016 


.036 


.027 


Hispanic 


-.016 


-.002 


.027 


.073 


-.022 


.030 




.115 


.128 


.112 


.126 


.108 


.119 




-.004 


-.000 


.007 


.015 


-.006 


.006 


Age 


.004 


.008 


.006 


.010 


.002 


.004 




.002 


.002 


.002 


.002 


.002 


.002 




.086 


.145 


.123 


.176 


.051 


.080 




a 


d 


c 


d 




b 


Education 


.061 


.125 


.038 


.096 


.023 


.064 




.010 


.012 


.010 


.012 


.010 


.012 




.237 


.327 


.149 


.252 


.090 


.168 




d 


d 


d 


d 


a 


d 


Occupation 


.334 


.211 


.266 


.215 


.212 


.183 




.061 


.064 


.059 


.063 


.058 


.060 




.183 


.094 


.146 


.096 


.116 


.082 




d 


c 


d 


c 


c 


b 


Income x 


.066 


.116 


.055 


.107 


.049 


.077 


10,000 


.020 


.021 


.019 


.020 


.018 


.019 




.111 


.164 


.093 


.151 


.082 


.109 




c 


d 


b 


d 


b 


d 


Single/ 


.156 


.204 


.111 


.150 


.068 


.133 


divorced 


.064 


.068 


.062 


.067 


.060 


.063 




.080 


.084 


.056 


.062 


.035 


.054 




a 


b 




a 




a 


Lives in 


.133 


.071 


.089 


.046 


.072 


-.004 


SMSA 


.057 


.060 


.055 


.059 


.053 


.056 




.070 


.032 


.047 


.020 


.038 


-.002 




a 













181 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-23 (Continued) 

Regression Analyses Predicting Number of Kinds of Performing 

Events Attended, Including Jazz, by Gender 





Model 1 


Model 2 


1 


\/lodel3 


I.V.S 


M F 


M 


F 


M 


F 


Parental 




.089 


.093 


.058 


.048 


guidance 




.021 


.021 


.021 


.020 






.137 


.128 


.088 


.066 




« 


d 


d 


b 


a 


Home 




.196 


.131 


.134 


.018 


activities 




.032 


.031 


.032 


.031 




• 


.197 


.125 
d 


.135 
d 


.017 


Hours TV/day 








-.023 

.012 

-.053 


-.031 

.010 

-.075 

b 


Art music 








.015 


.093 










.023 


.023 










.023 


.128 
d 


TV arts programs 








.131 

.017 

.259 

d 


.149 

.018 

.258 

d 


d.f. 


1,007 1,246 


1,007 


1,246 


1,007 


1,246 


R Squared 


.169 .196 


.224 


.229 


.280 


.313 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



182 



Appendix 

Appendix Table 5-24 

Regression Analyses Predicting Number of Kinds of Performing 

Events Attended, Excluding Jazz, by Gender 



I.V.S 

African- 
American 



Hispanic 



Age 



Education 



Occupation 



Income x 
10.000 



Single/ 
divorced 



Lives in 
SMSA 



Model 1 


1 


Model 2 


\ 


Model 3 


M 


F 


M 


F 


M 


F 


-.017 


-.091 


-.013 


-.049 


-.018 


-.013 


.073 


.076 


.071 


.075 


.069 


.072 


-.007 


-.032 


-.005 


-.017 


-.007 


-.005 


-.012 


-.050 


.018 


.007 


-.021 


-.031 


.098 


.113 


.096 


.112 


.093 


.106 


-.004 


-.012 


.005 


.002 


-.006 


-.007 


.006 


.008 


.007 


.010 


.004 


.005 


.001 


.001 


.001 


.001 


.001 


.001 


.137 


.168 


.168 


.194 


.098 


.098 


d 


d 


d 


d 


b 


c 


.056 


.104 


.039 


.082 


.026 


.054 


.008 


.010 


.008 


.011 


.008 


.011 


.251 


.310 


.177 


.243 


.118 


.160 


d 


d 


d 


d 


b 


d 


.247 


.214 


.198 


.217 


.152 


.187 


.052 


.057 


.051 


.056 


.050 


.054 


.158 


.109 


.127 


.110 


.097 


.095 


d 


c 


d 


d 


b 


c 


.066 


.101 


.058 


.094 


.052 


.068 


.017 


.018 


.016 


.018 


.016 


.017 


.130 


.161 


.115 


.150 


.103 


.108 


d 


d 





d 


b 


d 


.109 


.126 


.076 


.085 


.045 


.070 


.055 


.060 


.054 


.059 


.052 


.056 


.065 


.059 


.046 


.040 


.027 


.033 


a 


a 










.085 


.064 


.053 


.044 


.038 


.001 


.048 


.053 


.047 


.052 


.046 


.050 


.053 


.032 


.033 


.022 


.024 


.001 



183 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-24 (Continued) 
Regression Analyses Predicting Number of Kinds of Performing 
Events Attended, Excluding Jazz, by Gender 





Model 1 


Model 2 


I 


i^odelS 


I.V.S 


M F 


M 


F 


M 


F 


Childhood 




.063 


.069 


.037 


.030 


lessons 




.018 


.019 


.018 


.018 






.112 


.108 


.067 


.047 












a 




Parental 




.143 


.104 


.092 


.006 


guidance 




.028 


.028 


.027 


.027 






.168 


.113 


.108 


.007 






d 


c 







Hours TV/day 








-.023 

.011 

-.060 

a 


-.027 

.009 

.076 

b 


Art nfiusic 








.023 


.091 










.020 


.020 










.042 


.142 
d 


TV arts programs 








.100 

.015 

.232 

d 


.122 

.016 

.241 

d 


d.f. 


1,007 1,246 


1,007 


1,246 


1,007 


1,246 


R Squared 


.170 .191 


.208 


.215 


.258 


.297 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1 982 subsample. 



184 



Appendix 

Appendix Table 5-25 

Regression Analyses Predicting Number of Kinds of Visually 

Oriented Consumption Activities, by Gender 





Model 1 


Model 2 


Models 


I.V.S 


M 


F 


M 


F 


M 


F 


African- 


-.414 


-.813 


-.402 


-.664 


-.407 


-.596 


American 


.142 


.139 


.134 


.132 


.128 


.124 




-.082 


-.147 


-.080 


-.120 


-.081 


-.108 




b 


d 


b 


d 


b 


d 


Hispanic 


.129 


-.492 


.228 


-.294 


.131 


-.369 




.192 


.208 


.181 


.195 


.171 


.182 




.019 


-.059 
a 


.033 


-.035 


.019 


-.044 
a 


Age 


-.004 


-.007 


.001 


-.002 


-.007 


-.011 




.003 


.003 


.003 


.003 


.003 


.003 




-.042 


-.069 
a 


.006 


-.019 


-.080 
b 


-.115 
d 


Education 


.181 


.255 


.132 


.181 


.099 


.126 




.016 


.019 


.016 


.019 


.015 


.018 




.391 


.390 


.285 


.276 


.213 


.193 




d 


d 


d 


d 


d 


d 


Occupation 


.420 


.196 


.274 


.207 


.159 


.148 




.101 


.103 


.096 


.097 


.092 


.092 




.128 


.051 


.084 


.054 


.049 


.039 




d 




b 


a 






Income x 


.050 


.129 


.027 


.106 


.011 


.055 


10,000 


.033 


.033 


.031 


.031 


.029 


.029 




.047 


.107 
d 


.026 


.087 
c 


.011 


.045 


Single/ 


-.039 


-.059 


-.138 


-.200 


-.213 


-.230 


divorced 


.107 


.109 


.101 


.103 


.096 


.097 




-.011 


-.014 


-.039 


-.048 


-.061 
a 


-.055 
a 


Lives in 


.133 


.035 


.039 


-.032 


-.001 


-.116 


SMSA 


.094 


.097 


.089 


.091 


.085 


.085 




.040 


.009 


.011 


-.008 


-.000 


-.030 



185 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-25 (Continued) 

Regression Analyses Predicting Number of Kinds of Visually 

Oriented Consumption Activities, by Gender 



i.v^ 

Childhood 
lessons 



Parental 
guidance 



Hours TV/day 



Art music 



TV arts programs 



Modell 


Model 2 


1 


\/1odel3 


M F 


M 


F 


M 


F 




.207 


.249 


.143 


.174 




.035 


.033 


.033 


.031 




.177 


.201 


.122 


.140 




. d 


d 


d 


d 




.406 


.333 


.281 


.142 




.052 


.048 


.050 


.047 




.228 


.186 


.158 


.080 




d 


d 


d 

-.051 

.020 

-.065 

b 

.080 


b 

-.055 

.016 

-.078 

c 

.170 








.037 


.035 








.068 
a 

.236 

.028 

.262 

d 


.137 

d 

.241 
.027 

.245 

d 


1.007 1.246 


1,007 


1,246 


1,007 


1,246 


277 .288 


.358 


.366 


.428 


.448 



d.f. 

R Squared 



First rov; is unstandardized regression coefficient. 
Second row is standard error, 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



186 



Appendix 



Appendix Table 5-26 

Regression Analyses Predicting Number of Kinds 

of Performance Activities, by Gender 



I.V.S 

African- 
American 



Hispanic 



Age 



Education 



Occupation 



Income x 
10,000 



Single/ 
divorced 



Lives in 
SMSA 



Model 1 


Model 2 


1 


\4odel3 


M 


F 


M 


F 


M 


F 


.006 


-.023 


.007 


.001 


.006 


.008 


.042 


.042 


.042 


.042 


.042 


.043 


.004 


-.016 


.005 


.000 


.005 


.006 


.046 


.105 


.055 


.134 


.043 


.130 


.057 


.062 


.056 


.062 


.056 


.062 


.027 


.048 


.032 


.061 
a 


.025 


.059 
a 


-.000 


-.001 


.000 


-.000 


-.001 


-.001 


.001 


.001 


.001 


.001 


.001 


.001 


-.011 


-.056 


.006 


-.017 


-.036 


-.053 


.005 


.022 


.000 


.016 


-.004 


.011 


.005 


.006 


.005 


.006 


.005 


.006 


.047 


.131 
d 


.003 


.092 
b 


-.030 


.065 


.053 


.020 


.038 


.023 


.026 


.020 


.030 


.031 


.030 


.031 


.030 


.032 


.065 


.020 


.046 


.023 


.031 


.020 


-.014 


-.013 


-.016 


-.015 


-.018 


-.019 


.010 


.010 


.010 


.010 


.010 


.010 


-.052 


-.041 


-.061 


-.047 


-.068 


-.059 


.038 


.033 


.028 


.015 


.019 


.013 


.032 


.033 


.031 


.033 


.032 


.033 


.043 


.030 


.031 


.014 


.021 


.011 


.001 


-.056 


-.009 


-.064 


-.014 


-.072 


.028 


.029 


.027 


.029 


.028 


.029 


.001 


-.056 


-.010 


-.064 
a 


-.016 


-.072 
a 



187 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-26 (Continued) 

Regression Analyses Predicting Number of Kinds 

of Performance Activities, by Gender 





Model 1 


Model 2 


I 


lAodelS 


I.V.S 


M F 


M 


F 


IM 


F 


Childhood 




.018 


.047 


.011 


.040 


lessons 




.011 


.011 


.011 


.011 






.063 


.147 
d 


.036 


.124 
c 


Parental 




.045 


.012 


.031 


-.005 


guidance 




.016 


.015 


.017 


.016 






.100 


.026 


.068 


-.010 






b . 








Hours TV/day 








-.003 

.006 

-.017 


-.001 

.005 

-.005 


Art music 








.010 
.012 


.025 
.012 








- 


.035 


.075 




- 








a 


TV arts programs 








.028 

.009 

.121 

b 


.016 
.009 
.061 


d.f. 


1,007 1,246 


1,007 


1,246 


1,007 


1,246 


R Squared 


.005 .024 


.016 


.042 


.029 


.050 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



188 



Appendix 



Appendix Table 5-27 

Regression Analyses Predicting Number of Kinds of 

Nonperformance Activities^ by Gender 



I.V.S 

African- 
American 



Hispanic 



Age 



Education 



Occupation 



Income x 
10,000 



Single/ 
divorced 



Lives in 
SMSA 



Model 1 


Model 2 


[ 


i/lodel3 


M 


F 


M 


F 


M 


F 


-.093 


-.293 


-.083 


-.164 


-.084 


-.140 


.104 


.095 


.098 


.089 


.096 


.087 


-.027 


-.084 
b 


-.024 


-.047 


-.025 


-.040 


.137 


.020 


.239 


.181 


.197 


.133 


.140 


.141 


.132 


.132 


.130 


.128 


.029 


.004 


.052 


.034 


.043 


.025 


-.008 


-.010 


-.004 


-.005 


-.007 


-.009 


.002 


.002 


.002 


.002 


.002 


.002 


-.140 


-.062 


-.067 


-.079 


-.119 


-.142 


d 


d 


a 


b 


c 


d 


.056 


.103 


.023 


.057 


.009 


.031 


.012 


.013 


.012 


.013 


.012 


.013 


.178 


.250 


.075 


.139 


.028 


.076 


d 


d 


a 


d 




a 


.442 


.043 


.343 


.059 


.286 


.022 


.074 


.071 


.070 


.066 


.070 


.065 


.200 


.018 


.156 


.025 


.130 


.009 


d 




d 




d 




-.060 


.013 


-.077 


-.000 


-.084 


-.027 


.024 


.023 


.023 


.021 


.022 


.021 


-.084 


.017 


-.107 


-.001 


-.117 


-.035 


a 




c 




c 




.203 


.292 


.131 


.185 


.099 


.171 


.078 


.075 


.074 


.070 


.073 


.068 


.086 


.111 


.055 


.070 


.042 


.065 


b 


d 




b 




a 


.044 


-.130 


-.025 


-.180 


-.041 


-.218 


.069 


.066 


.065 


.062 


.064 


.060 


.019 


-.054 


-.011 


-.074 


-.018 


-.090 




a 




b 




c 



189 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-27 (Continued) 

Regression Analyses Predicting Number of Kinds of 

Nonperformance Activities, by Gender 



I.V.S 

Childhood 
lessons 



Parental 
guidance 



Hours TV/day 



Art music 



TV arts programs 



d.f. 

R Squared 



Model 1 


Model 2 


1 


\/lodel3 


M F 


M 


F 


M 


F 




.218 


.246 


.191 


.214 




.025 


.022 


.025 


.022 




.276 


.316 


.242 


.274 




d 


d 


d 


c 




.193 


.138 


.135 


.050 




.038 


.033 


.038 


.033 




. .160 


.123 


.112 


.045 




d 


d 


c 

-.038 

.028 

-.071 

a 

.028 
.028 


-.043 

.011 

-.098 

d 

.033 
.024 








.036 

.107 

.021 

.175 

d 


.043 

.139 

.019 

.224 

d 


1,007 1,246 


1,007 


1,246 


1,007 


1.246 


.158 .156 


.252 


.267 


.284 


.317 



First row is unstandardized regression coefficient. 
Second row is standard error. 
Third row is standardized regression coefficient. 
Fourth row indicates significance: 

a less than or equal to .05 

b less than or equal to .01 

c less than or equal to .001 

d less than or equal to .0001 
Based on data from November/December 1982 subsample. 



190 



Appendix 

Appendix Table 5-28 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Gender (Femaie [G]) for Selected Models 





18 to 31 years 


32 to 51 years 


over 51 years 






(N=728) 






(N=767) 






(N=757) 






b 


se 


beta 


b 


se 


beta 


b 


se 


beta 


D.V. 


: Lessons/* 


















A 


.060 


.209 


.012 


-.109 


.175 


-.023 


-.240 


.161 


-.063 


H 


-.235 


.273 


-.035 


-.251 


.221 


-.044 


-.041 


.283 


-.006 


G 


.293 


.122 


.091(a) 


.034 


.097 


.013 


.158 


.088 


.075 


D.V. 


• Homer 


















A 


.265 


.106 


.092(a) 


.222 


.119 


.065 


-.138 


.148 


-.036 


H 


.192 


.138 


.052 


.178 


.151 


.042 


.280 


.260 


.041 


G 


.256 


.062 


.144(d) 


.215 


.066 


.112(b) 


.187 


.081 


.088(a) 


D.V. 


TV arts -Model! 


















A 


.132 


.172 


.029 


-.148 


.200 


-.025 


.179 


.232 


.027 


H 


.064 


.241 


.010 


-.112 


.260 


-.014 


.184 


.372 


.017 


G 


.116 


.115 


.037 


.281 


.121 


.077(a) 


.221 


.132 


.056 


D.V. 


TV arts -Model 2 


















A 


.312 


.166 


.068 


-.164 


.186 


-.028 


.143 


.215 


.022 


H 


.349 


.229 


.054 


-.009 


.243 


-.001 


.052 


.345 


.005 


G 


-.017 


.111 - 


-.005 


.100 


.114 


.027 


.034 


.123 


.009 


D.V. 


Art music -Model 1 


















A 


-.383 


.121 - 


-.117(b) 


-.471 


.153 


-.101(b) 


-.568 


.169 


-.113(c) 


H 


.252 


.170 


.054 


-.328 


.200 


-.053 


-.167 


.271 


-.020 


G 


.243 


.081 


.108(b) 


.421 


.093 


.143(d) 


.288 


.096 


.095(b) 


D.V. 


Art music -Model 2 


















A 


-.254 


.115 ■ 


-.077(a) 


-.482 


.147 


-.103(b) 


-.585 


.157 


-.116 


H 


.461 


.161 


.099(b) 


-.266 


.192 


-.043 


-.255 


.251 


-.031 


G 


.154 


.077 


.069(a) 


.307 


.090 


.104(c) 


.154 


.090 


.051 


D.V. 


Attend performance, inc. jazz - Model 1 
















A 


.081 


.101 


.029 


-.110 


.109 


-.033 


.085 


.111 


.027 


H 


-.157 


.141 ■ 


-.040 


.016 


.143 


.004 


.114 


.178 


.022 


G 


.130 


.068 


.069 


.347 


.066 


.166(d) 


.178 


.063 


.095(b) 


D.V. 


Attend performance, inc. j 


azz - Model 2 
















A 


.161 


.098 


.058 


-.115 


.105 


-.035 


.073 


.109 


.023 


H 


-.028 


.138 


-.007 


.061 


.137 


.014 


.084 


.175 


.017 


G 


.074 


.066 


.039 


.273 


.064 


.131(d) 


.141 


.062 


.075(a) 


D.V. 


: Attend performance, inc. j 


azz - Model 3 
















A 


.171 


.095 


.061 


-.033 


.103 


-.010 


.071 


.105 


.023 


H 


-.125 


.131 - 


-.031 


.066 


.133 


.015 


.073 


.166 


.014 


G 


.085 


.063 


.045 


.243 


.063 


.116(d) 


.136 


.059 


.073(a) 



191 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-28 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Gender (Female [G]) for Selected Models 



18 to 31 years 

(N=728) 

b se beta 



32 to 51 years 

(N=767) 
b se beta 



over 51 years 

(N=757) 
b se beta 



D.V.: Attend performance, exc. jazz - Model 1 

A -.141 .082 -.063 -.141 

H -.111 .114 -.034 -.033 

G .158 .055 .102(b) .306 

D.V.: Attend performance, exc. jazz - Model 2 

A -.086 .080 -.038 -.144 

H -.022 .113 -.007 .001 

G .120 .054 .077(a) .250 

D.V.: Attend performance, exc. jazz - Model 3 

A -.065 .079 -.029 -.065 

H -.092 .109 -.029 .012 

G .133 .053 .086(a) .217 

D.V.: Visually oriented consumption - Model 1 

A -.772 .172 -.153(d) -.750 

H -.296 .241 -.041 -.220 

G .919 .115 .266(d) 1.022 

D.V.: Visually oriented consumption - Model 2 

A -.564 .160 -.112(c) -.738 

H .034 .224 .005 -.107 

G .771 .107 .223(d) .900 

D.V.: Visually oriented consumption - Model 3 

A -.572 .154 -.113(c) -.581 

H -.114 .213 -.016 -.093 

G .791 .103 .229(d) .838 

D.V.: Performance activities- Model 1 

A -.012 .054 -.009 -.025 

H .102 .075 .051 -.024 

G .018 .036 .071 .066 



.097 -.048 .067 
.126 -.009 .027 
.058 .166(d) .164 



.094 -.050 .057 
.123 .000 .003 
.058 .136(d) .133 



.092 -.022 .052 
.119 .003 -.008 
.056 .118(d) .129 



.173 -.136(d) -.347 
.226 -.030 -.004 
.105 .293(d) .783 



.164 
.214 
.100 



.058 
.076 
.035 



-.134(d) .377 

-.015 -.093 

.258(d) .662 



.157 -.105(c) -.312 
.203 -.013 -.090 
.096 .240(d) .641 



-.016 

-.012 

.068 



.020 
.197 
.008 



D.V.: Performance activities- Model 2 

A .011 .054 .008 -.025 .058 -.016 .019 

H .141 .075 .071 -.010 .076 -.005 .193 

G .053 .036 .055 .048 .035 .049 .003 



.102 
.163 
.058 



.101 
.162 
.058 



.097 
.154 
.055 



.182 
.291 
.103 



.159 
.253 
.090 



.023 
.006 
.095(b) 



.020 
.001 
.078(a) 



.018 
-.002 
.075(a) 



-.061 
-.000 
.230(d) 



.172 -.067(a) 
.276 -.010 
.098 .195(d) 



-.055 
-.010 
.188(d) 



.042 .018 
.067 .112(b) 
.024 .013 



.042 .018 
.067 .110(b) 
.024 .004 



192 



Appendix 

Appendix Table 5-28 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Gender (Female [G]) for Selected Models 

18 to 31 years 32 to 51 years over 51 years 

(N=728) (N=767) (N=757) 

b se beta b se beta b se beta 

D.V.: Performance activities - Model 3 

A -.006 .054 -.004 .007 .058 .004 .008 .043 .008 

H .114 .075 .058 -.001 .075 -.001 .187 .067 .106(b) 

G .066 .036 .045 .033 .035 .034 .005 .024 .007 

D.V.: Nonperformance activities - Model 1 

A -.533 .143 -.137(c) -.037 .120 -.011 .083 .090 .034 

H .103 .201 .019 -.002 .156 -.000 .265 .144 .067 

G .243 .096 .091(a) .222 .072 .104(b) .114 .051 .078(a) 

D.V.: Nonperformance activities - Model 2 

A -.349 .131 -.090(b) -.025 .116 -.007 .073 .085 .030 

H .411 .184 .075(a) .067 .151 .015 .224 .137 .057 

G .124 .088 .047 .160 .071 .075(a) .054 .049 .037 

D.V.: Nonperformance activities - Model 3 

A -.316 .127 -.081(a) .038 .115 .011 .079 .083 .033 

H .293 .175 .053 .067 .148 .015 .214 .131 .054 

G .162 .085 .061 .138 .070 .065(a) .052 .047 .036 



*For Starred models only, respondents without data on father's and mother's education excluded, and 
mother's and father's educational attainment used as controls. Ns for these models are 629 for the 
18-30 group, 629 for the 31-51 group, and 480 for the over 51 group. Model numbers refer to their 
counterparts in Appendix tables 5-14 through 5-20. a: p less than or equal to .05; b: p less than or 
equal to .01 ; c: p less than or equal to .001 ; d: p less than or equal to .001 . Analyses based on 
November/December 1982 subsample. 



193 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-29 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Female Gender (G), by Own Educational Attainment, for 

Selected Models 









1-11 years 






high school 










N=806 






N=918 




D.V. 


I.V. 


b 


se 


beta 


b 


se 


beta 


Childhood 


A 


-.156 


.128 


-.056 


.250 


.178 


.050 


lessons 


H 


-.031 


.181 


-.008 


-.101 


.234 


-.017 




G 


.202 


.095 


.094(a) 


.137 


.094 


.051 


Parental 


A 


.166 


.093 


.085 


.191 


.113 


.061 


guidance 


H 


.350 


.132 , 


.127(b) 


.194 


.150 


.049 




G 


.109 


.070 


.072 


.166 


.060 


.098(b) 


TV arts 


A 


.065 


.150 


.018 


.115 


.188 


.021 


programs 


H 


.115 


.204 


.024 


-.172 


.257 


-.021 


(Model 1)/* 


G 


.084 


.114 


.031 


.207 


.110 


.062 


TV arts 


A 


.091 


.142 


.026 


.139 


.178 


.025 


programs 


H 


.182 


.196 


.038 


.023 


.257 


.003 


(Model 2) 


G 


-.047 


.110 


-.017 


.083 


.103 


.025 


Art music 


A 


-.501 


.123 


-.166(d) 


-.241 


.138 


-.057 


(ModeM) 


H 


-.010 


.168 


-.002 


.103 


.202 


.016 




G 


.160 


.093 


.068 


.361 


.081 


.140(d) 


Art music 


A 


-.466 


.120 


-.154(d) 


-.206 


.132 


-.049 


(Model 2) 


H 


.061 


.164 


.015 


.215 


.194 


.034 




G 


.076 


.091 


.032 


.294 


.078 


.114(c) 


Performing-arts 


A 


-.008 


.048 


-.007 


-.032 


.085 


-.013 


attendance, 


H 


-.007 


.066 


-.004 


-.024 


.125 


-.006 


including jazz 


G 


.032 


.037 


.036 


.214 


.050 


.139(d) 


(ModeM) 
















Performing-arts 


A 


.006 


.047 


.005 


-.022 


.085 


-.009 


attendance, 


H 


.022 


.065 


.014 


.012 


.124 


.003 


including jazz 


G 


.004 


.036 


.004 


.191 


.050 


.124(d) 


(Model 2) 

















194 



Appendix 

Appendix Table 5-29 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Female Gender (G), by Own Educational Attainment, for 

Selected Models 









1-11 years 






high school 










N=606 






N=918 




D.V. 


I.V. 


b 


se 


beta 


b 


se 


beta 


Performing-arts 


A 


.018 


.047 


.016 


.005 


.081 


.002 


attendance, 


H 


.006 


.064 


.004 


-.016 


.117 


-.004 


including jazz 


G 


.008 


.036 


.009 


.155 


.047 


.101(b) 


(Model 3) 
















Performing-arts 


A 


-.052 


.044 


-.050 


-.162 


.074 


-.073(a) 


attendance, 


H 


.056 


.059 


-.040 


-.003 


.128 


-.001 


excluding jazz 


G 


.049 


.033 


.060 


.209 


.043 


.155(d) 


(Model 1) 






13-15 years 
N=389 






16 or more 
N=338 




D.V. 


I.V. 


b 


se 


beta 


b 


se 


beta 


Childhood 


A 


-.375 


.296 


-.065 


.133 


.393 


.019 


lessons 


H 


-.856 


.359 


-.121(a) 


-.473 


.645 


-.041 




G 


.159 


.156 


.051 


.300 


.167 


.100 


Parental 


A 


.193 


.169 


.055 


.090 


.266 


.019 


guidance 


H 


.058 


.205 


.013 


-.543 


.403 


-.068 




G 


.438 


.089 


.230(d) 


.412 


.105 


.199(d) 


TV arts 


A 


-.247 


.328 


-.038 


-.096 


.470 


-.011 


programs 


H 


.228 


.413 


.028 


-1.003 


.901 


-.059 


(Model 1)/* 


G 


.187 


.180 


.051 


.714 


.233 


.167(b) 


TV arts 


A 


.011 


.317 


.002 


.034 


.443 


.004 


programs 


H 


.530 


.401 


.064 


-.475 


.846 


-.028 


(Model 2) 


G 


-.046 


.177 


-.013 


.412 


.224 


.096 


Art music 


A 


-.693 


.259 


-.131(b) 


-.666 


.302 


-.110(a) 


(Model 1) 


H 


-.200 


.326 


-.030 


-1.068 


.579 


-.092 




G 


.577 


.142 


.194(d) 


.330 


.150 


.113(a) 



195 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-29 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Femaie Gender (G), by Own Educational Attainment, for 

Selected Models 









13-1 5 years 






16 or more 










N=389 






N=338 




D.V. 


I.V. 


b 


se 


beta 


b 


se 


beta 


Art music 


A 


-.441 


.247 


-.083 


-.565 


.295 


-.093 


(Model 2) 


H 


.123 


.312 


.018 


-.802 


.567 


-.069 




G 


.396 


.138 


.133(b) 


.184 


.150 


.063 


Performing-arts 


A 


-.137 


.208 


-.034 


.156 


.299 


.027 


attendance, 


H 


-.126 


.262 


-.024 


-.862 


.572 


-.079 


including jazz 


G 


.380 


.114 


.165 


.653 


.148 


.238(d) 


(Model 1) 
















Performing-arts 


A 


.035 


.202 


.008 


.279 


.289 


.049 


attendance, 


H 


.115 


.256 


.022 


-.546 


.555 


-.050 


including jazz 


G 


.290 


.113 


.126(a) 


.475 


.147 


.173(b) 


(Model 2) 
















Performing-arts 


A 


.106 


.192 


.026 


.310 


.278 


.054 


attendance. 


H 


.039 


.241 


.008 


-.427 


.529 


-.039 


including jazz 


G 


.277 


.108 


.120(a) 


.388 


.140 


.141(b) 


(Models) 
















Performing-arts 


A 


-.246 


.180 


-.069 


-.069 


.263 


-.014 


attendance. 


H 


-.122 


.227 


-.027 


-.848 


.503 


-.088 


excluding jazz 


G 


.366 


.099 


.183(c) 


.552 


.130 


.229(d) 


(Model 1) 

















196 



Appendix 

Appendix Table 5-29 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Female Gender (G), by Own Educational Attainment, for 

Selected Models 



D.V. 



I.V. 



1-11 years 



high school 



Performing-arts 


A 


-.041 


.043 


-.039 


-.157 


.073 


-.071(a) 


attendance, 


H 


-.034 


.059 


-.024 


-.022 


.107 


-.007 


excluding jazz 


G 


.027 


.033 


.033 


.192 


.043 


.143(d) 


(Model 2) 
















Performing-arts 


A 


-.029 


.043 


-.028 


-.125 


.071 


-.057 


attendance, 


H 


-.047 


.058 


-.033 


-.003 


.102 


-.001 


excluding ja?? 


G 


.030 


.033 


.037 


.162 


.041 


■121(d) 


(Model 3) 
















Visually oriented 


A 


-.456 


.120 


-.151(c) 


-.857 


.177 


-.157(d) 


consumption 


H 


-.068 


.163 


-.017 


-.301 


.260 


-.037 


activities 


G 


.559 


.090 


.235(d) 


1.041 


.104 


.313(d) 


(Model 1) 
















Visually oriented 


A 


-.429 


.113 


-.142(c) 


-.801 


.167 


-.147(d) 


consumption 


H 


-.015 


.155 


-.004 


-.129 


.245 


-.016 


activities 


G 


.462 


.086 


.195(d) 


.942 


.098 


.283(d) 


(Model 2) 
















Visually oriented 


A 


-.376 


.110 


-.125(c) 


-.742 


.158 


-.136(d) 


consumption 


H 


-.059 


.149 


-.014 


-.174 


.228 


-.021 


activities 


G 


.464 


.083 


.196(d) 


.891 


.092 


.268(d) 


(Model 3) 
















Performance 


A 


-.034 


.037 


-.040 


.082 


.043 


.065 


activities 


H 


.101 


.051 


.086(a) 


.149 


.064 


.078(a) 


(Model 1) 


G 


.009 


.028 


.013 


.014 


.025 


.018 


Performance 


A 


-.026 


.037 


-.030 


.087 


.043 


.069(a) 


activities 


H 


.119 


.051 


.101(a) 


.161 


.064 


.085(a) 


(Model 2) 


G 


-.005 


.028 


-.007 


.007 


.025 


.009 


Performance 


A 


-.045 


.038 


-.052 


.080 


.044 


.063 


activities 


H 


.113 


.051 


.096(a) 


.159 


.063 


.084(a) 


(Model 3) 


G 


.001 


.029 


.001 


.001 


.026 


.001 



197 



Race, Ethnicity, and Participation in tlie Arts 

Appendix Table 5-29 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Female Gender (G), by Own Educational Attainment, for 

Selected Models 



D.V. 



I.V. 



1-11 years 



high school 



Nonperformance 


A 


-.118 


.072 


-.067 


-.360 


.115 


-.106(b) 


activities 


H 


.150 


.098 


.063 


-.027 


.168 


-.005 


(Modell) 


G 


.119 


.055 


.086(a) 


.103 


.067 


.049 


Nonperformance 


A 


-.060 


.065 


-.034 


-.308 


.107 


-.090(b) 


activities 


H 


.268 


.089 


.112(b) 


.097 


.157 


.019 


(Model 2) 


G 


.032 


.050 


.023 


.041 


.062 


.020 


Nonperformance 


A 


-.055 


.065 , 


-.031 


-.270 


.106 


-.079(a) 


activities 


H 


.247 


.088 


.103(b) 


.081 


.153 


.016 


(Model 3) 


G 


.037 


.049 


.027 


.031 


.062 


.015 


D.V. 


I.V. 




IS-ISyears 






16 or more 




Performing-arts 


A 


-.113 


.177 


-.032 


.022 


.258 


.004 


attendance, 


H 


.077 


.223 


.017 


-.617 


.495 


-.064 


excluding jazz 


G 


.317 


.099 


.158(b) 


.421 


.131 


.175(b) 


(Model 2) 




' 












Performing-arts 


A 


-.045 


.168 


-.013 


.066 


.248 


.013 


attendance, 


H 


.015 


.211 


.003 


-.497 


.472 


-.052 


excluding jazz 


G 


.300 


.094 


.150(b) 


.343 


.125 


.143(b) 


(Model 3) 
















Visually oriented 


A 


-.931 


.289 


-.156(b) 


-.554 


.352 


-.082 


consumption 


H 


-.251 


.364 


-.033 


-.874 


.675 


-.067 


activities 


G 


1.205 


.158 


.360(d) 


1.089 


.175 


.334(d) 


(Model 1) 
















Visually oriented 


A 


-.640 


.276 


-.110(a) 


-.426 


.339 


-.063 


consumption 


H 


-.656 


.349 


.013 


-.531 


.651 


-.041 


activities 


G 


1.009 


.155 


.301(d) 


.906 


.173 


■277(d) 


(Model 2) 
















Visually oriented 


A 


-.506 


.260 


-.085 


-.383 


.321 


-.057 


consumption 


H 


.011 


.327 


.001 


-.373 


.609 


-.029 


activities 


G 


.947 


.146 


.282(d) 


.790 


.162 


.242(d) 


(Model 3) 

















198 



Appendix 

Appendix Table 5-29 (Continued) 

Effects of Race (African-American [A]), Ethnicity (Hispanic [H]), 

and Female Gender (G), by Own Educational Attainment, for 

Selected Models 



D.V. 


I.V. 

A 




13-15 years 






16 or more 




Performance 


-.069 


.096 


-.038 


-.166 


.138 


-.066 


activities 


H 


-.061 


.121 


-.026 


-.196 


.264 


-.041 


(ModeH) 


G 


.098 


.053 


.095 


.165 


.068 


.136(a) 


Performance 


A 


-.017 


.096 


-.009 


-.153 


.138 


-.061 


activities 


H 


.011 


.122 


.005 


-.158 


.265 


-.033 


(Model 2) 


G 


.069 


.054 


.067 


.146 


.061 


.120(a) 


Performance 


A 


.009 


.095 


.005 


-.108 


.138 


-.043 


activities 


H 


-.009 


.120 


-.004 


-.096 


.263 


-.020 


(Model 3) 


G 


.056 


.054 


.054 


.131 


.070 


.108 


Nonperformance 


A 


-.249 


.225 


-.054 


-.149 


.293 


-.028 


activities 


H 


-.048 


.283 


-.008 


-.348 


.562 


-.033 


(Model!) 


G 


.374 


.123 


.145(b) 


.411 


.145 


.157(b) 


Nonperformance 


A 


-.011 


.213 


-.002 


-.055 


.283 


-.010 


activities 


H 


.286 


.269 


.049 


-.091 


.544 


-.009 


(Model 2) 


G 


.253 


.119 


.098(a) 


.276 


.144 


.105 


Nonperformance 


A 


.057 


.207 


.012 


.043 


.273 


.008 


activities 


H 


.231 


.260 


.040 


.052 


.520 


.005 


(Model 3) 


G 


.242 


.116 


.094(a) 


.200 


.138 


.076 



*/ For starred analyses only, cases without information on mother's or father's education were omitted 
and controls for mother's and father's education were included. For these models, Ns are 365 for 1 -1 1 
years, 717 for high school graduate, 352 for 13-15 years, and 317 for 16 or more years. Model 
numbers refer to their counterparts in Appendix tables 5-14 through 5-20. a=p less than or equal to 
.05; b=p less than or equal to .01 ; c=p less than or equal to .001 ; d=p less than or equal to .0001 . 
Results based on analyses of November/December 1982 subsample. 



199 



Race, Ethnicity, and Participation in the Arts 

Appendix Table 5-30: Coefficients Representing Effects of 

African-American (A) and Hispanic (H) on Core Participation 

Items (1) with Race/Ethnicity only (2) with Demographic Controls 

for November/December 1982 Subsample 



H 





Jazz 


Classical 


Opera 


Musical 


1 


Play 




1 


2 


1 


2 


1 


2 


1 


2 


1 


2 


b 


.745 


1.039 


-.931 


-.513 


1.188 


-.953 


-.961 


-.501 


-.930 


-.436 


se 


.185 


.207 


.295 


.319 


.920 


.934 


.229 


.246 


.296 


.318 


sig 


b 


c 


a 


NS 


NS 


NS 


c 


a 


a 


NS 


b 


.030 


.223 


-1.790 ■ 


■1.275 


-.725 


-.374 


-.601 


-.056 


-1.164 


-.547 


se 


.331 


.351 


.638 


.654 


1.079 


1.098 


.291 


.312 


.479 


.499 


sig 


NS 


NS 


a 


NS 


NS 


NS 


a 


NS 


a 


NS 




Ballet 


Art 




Instrument 


Act, 


sing 


Read 




1 


2 


1 


2 


1 


2 


1 


2 


1 


2 


b 


-1.660 


-1.422 


-1.032 


-.730 


-.024 


.078 


.034 


.129 


-.883 


-.564 


se 


.729 


.741 


.216 


.236 


.351 


.365 


.324 


.339 


.136 


.157 


sig 


a 


NS 


c 


a 


NS 


NS 


NS 


NS 


c 


b 


b 


-.333 


.044 


-.472 


-.008 


.207 


.410 


.644 


.863 


-.713 


-.282 


se 


.574 


.597 


.257 


.282 


.456 


.474 


.363 


.384 


.192 


.219 


sig 


NS 


NS 


NS 


NS 


NS 


NS 


NS 


a 


b 


NS 



b is the logistic regression coefficient, se is the standard error, sig refers to the level of statistical 
significance, where a=probability less than .05, b=probability less than .001 , c=probability less than 
.00005, and NS=not significant. 



200 



ABOUT THE AUTHORS 



Paul DiMaggio is Professor of Sociology at Princeton University and author 
of Managers of the Arts (NEA Research Division Report #20, Seven Locks 
Press, 1986) 

Francie Ostrower received her Ph.D. in sociology from Yale University and 
is currently Assistant Professor of Sociology at Harvard University. 



201 




Race, Ethnicity, and 
Participation in the Arts 



Based on evidence from two trail-blazing national surveys, this 
volume examines rates of participation of African-Americans, 
Hispanic-Americans, and whites in activities ranging from per- 
forming-arts attendance and visits to art museums, to creating craft 
works and performing onstage. The study reveals a pattern of 
differentiation without segmentation for the activities included in 
the survey: significant differences, but many activities and experi- 
ences held in common. The study goes beyond simply demon- 
strating that patterns of participation vary among the groups to ask 
why these differences exist, exploring the impact of such factors as 
education, gender, and income. The authors conclude by high- 
lighting the need for future surveys to collect information on more 
activities with roots in many different ethnic and racial communi- 
ties, and on the rich ethnic diversity within the groups studied. By 
opening up the issues, however, this volume serves as essential 
background for any discussion of cultural diversity and the arts in 
the United States. 

A survey report by 

Paul DiMaggio Francie Ostrower 

Princeton Harvard University 



NATIONAL 
ENDOWMENT 
FOR^(S^THE 

ARTS 



Research Division Report #25 



Seven Locks Press 

Washington, D.C. 



780 



929"765037 



ISBN D-TET7bS-D3-fe