The Geography of Participation in the Arts and Culture
Report 41: Executive Summary
With the new attention on cultural policy and cultural funding at the regional and state levels in
the United States, it is becoming increasingly important to collect basic information on the arts
and culture on a geographic basis. This monograph explores the geographic variation in the
participation of the American adult population in arts and cultural activities. It is based
primarily on data from the 1997 Survey of Public Participation in the Arts, the latest is a sequence
of participation studies commissioned by the National Endowment for the Arts as a way of documenting
the cultural consumption patterns of the American adult population.
The primary goals of this monograph are twofold:
to establish a baseline of results on the geographic variation of participation in the arts and
cultural activities in the United States, and
to provide some preliminary analyses that suggest possible explanations for the observed
The 1997 Survey of Public Participation in the Arts was conducted at a scale sufficient to allow the
consideration of participation levels as well as the construction of a series of profiles of the
audiences for various art forms and cultural activities across all nine regions of the country. The
data are such that they also allow an analysis at the state level for ten of the largest states:
California, Florida, Illinois, Massachusetts, Michigan, New Jersey, New York, Ohio, Pennsylvania,
and Texas. Analyses at both the regional level and state level are reported in this monograph.
The Survey of Public Participation in the Arts focuses on attendance at eight key art forms: jazz,
classical music, opera, musical stage plays or operettas, non-musical stage plays, ballet, dance
other than ballet, and art museums or art galleries. The data also allow a consideration of
participation in three other cultural activities: reading literature, visiting historic parks or
monuments, and visiting art or crafts fairs or festivals. A wide variety of ancillary analyses are
possible as well, and several are reported in this monograph.
It is rather difficult to summarize briefly all of the findings and results of the many analyses
that we have conducted, given that they consider eleven art forms and cultural activities over nine
regions and ten states in relationship to a wide variety of other variables. In this summary we
report some selected findings, hoping that they will encourage the reader to dig deeper into the
The basic findings concerning participation rates are presented in Sections 1 and 2. These results
Generally speaking, these ten states have higher than average participation rates across all
eight key art forms.
Some art forms (art museums and musical plays) enjoy high participation rates across the board,
while others (opera and ballet, in particular) have much lower participation rates.
Certain states, most notably New York, Massachusetts, and New Jersey, stand out from the other
six as states with generally high participation rates. Pennsylvania and Texas, on the other hand,
systematically have lower than average participation rates.
Nevertheless, there is more variation across art forms than there is variation across states,
i.e., participation levels for a particular art form are quite similar across states while
participation levels for each state vary quite widely across art forms.
Participation in the three other cultural activities is quite a bit higher than participation
in the eight key art forms.
- The data suggest that there may be some substitution among types of cultural participation,
with the citizens of a particular state trading off participation in one art form with participation
in another. The possibility of substitution is particularly strong when considering the tradeoff
between the eight key art forms and the other three types of cultural activities, which are more
popular in their appeal.
At a regional level, the highest participation rates can be found in New England, the Middle
Atlantic region, and the Pacific region. New England has the highest participation rate for five of
the eight key art forms and the second highest rate for two others. The East South Central region,
on the other hand, reports the lowest participation rates for six of the eight art forms. The
pattern differs somewhat for the three other cultural activities, but the East South Central region
still reports the lowest participation rates by a considerable margin.
The SPPA also allows the measurement of participation in the arts and culture through various media.
These results are discussed in Section 3.
Nearly seven out of ten American adults report having participated in at least one of the eight
key art forms through the medium of television or video in the previous year. Among the ten states,
this rate is highest in California and lowest in Pennsylvania.
Participation via radio is at a
somewhat lower level; slightly less than six out of ten American adults report having participated
in at least one of the eight key art forms via radio in the previous twelve months. Massachusetts
and California show the highest levels of participation via radio broadcasts, and Ohio and
Pennsylvania report the lowest levels.
Participation via listening to records, compact discs,
or tape cassettes is lower still; slightly less than half of the adult population reports
participation via one of these media. New Jersey has the highest participation rate, followed by
Section 4 of the monograph uses the SPPA data to gauge a more direct form of participation:
participation through direct personal involvement in artistic creation or performance.
Five out of nine American adults report having been involved in one or another form of direct
artistic creation in the previous twelve months. Higher than average levels of participation in
creation are reported for Massachusetts and New Jersey; a lower than average level is reported in
Approximately four out of every ten American adults report participation in one
or another art form through personal performance. Of the ten states considered here, Florida has the
highest rate of participation in performance followed by Massachusetts. Ohio reports the lowest rate
of participation in personal performance and California the second lowest.
Sections 5 and 6 of the monograph ask what the relationship is between participation rates across
art forms. Is relatively high participation in one art form accompanied by relatively high
participation in another art form? Or does it tend to be accompanied by a relatively low
participation rate? Or does there seem to be no relationship? Section 5 looks at this question from
the perspective of regions and Section 6 from the perspective of the ten states.
At the regional level, all of the participation rates are positively correlated with one
another, whether they are for the eight key art forms or for the additional three cultural
activities, and many of these correlations are quite high. Thus, at the regional level participation
rates tend to parallel one another. High participation in one art form or cultural activity will be
an indicator of high participation in another.
At the state level, however, a slightly different pattern emerges. While the correlation
coefficients for the eight key art forms are, with one exception, positive, they are not as strong
as they are at the regional level. This is not too surprising because one would expect to observe
more nuance and variation at the lower geographic aggregation. When this analysis is extended to
other cultural activities, however, negative correlations appear with respect to attendance at
historic parks or monuments and attendance at art or crafts fairs or festivals, suggesting that at
the state level there is some degree of substitution between participation in the eight key art
forms and participation in these cultural activities.
Sections 7 and 8 of the monograph begin to explore possible explanations for the observed geographic
variations in participation rates. Section 7 looks at this question from the perspective of regions,
Section 8 from the perspective of the ten states. In each section, two sets of independent variables
are considered: ones that measure socio-economic characteristics of the area's population and ones
that measure the presence of cultural organizations of various types.
At the regional level:
Education, particularly as measured by the percentage of the adult population with a bachelor's
degree, is an excellent predictor of participation rates in all of the art forms as well as in the
three other cultural activities.
Median household income is positively correlated with participation in all of the art forms,
while percentage below the poverty level is negatively correlated with ten of the eleven art forms
and cultural activities. Median household income is the better predictor.
The percentage of the population that is minority has mixed value as a predictor of
participation. The strongest correlations are with attendance at historic parks or monuments and
attendance at fairs or festivals, suggesting that these cultural activities may be less attractive
to minority audiences.
The density of the population as measured by persons per square mile is not a particularly good
predictor of participation rates, but two other indicators of urbanization - "percentage
non-metropolitan" and "percentage rural" -- are both strongly negatively correlated with participation
in each of the art forms, as one might expect.
The density of arts and cultural organizations when measured per capita is strongly and
positively correlated with participation rates when the boundaries of the sectors for which the data
have been collected are comparable. When density is measured per square mile it is generally not as
good a predictor.
At the state level:
- Education, at least as measured by the percentage of high school graduates, is not a
particularly good predictor of participation rates for these ten states. Percentage of the adult
population with a bachelor's degree, on the other hand, is a much better predictor.
- Median household income is positively correlated with participation in nine of the eleven art
forms and cultural activities. Percentage below the poverty level is negatively correlated with
participation in seven of the eleven art forms and cultural activities.
- Percentage minority is once again a mixed predictor of participation rates.
- Population per square mile is a very good predictor of participation rates in a number of art
forms. Percentage non-metropolitan is a reasonably good predictor as well. Percentage rural is
generally a less useful predictor.
- At the state level, the density of arts and cultural organizations when measured per capita is
moderately and positively correlated with participation rates when the boundaries of the sectors for
which the data have been collected are comparable. When density is measured per square mile,
however, the correlation coefficients increase and a number of very strong correlations are
observed, particularly with respect to attendance at both musical and non-musical plays.
Do responses to the SPPA suggest points of leverage or particular policy instruments that might be
particularly important in increasing participation rates? Section 9 explores this question by
looking at three other sets of questions asked in the survey: questions concerning interest in
increased participation, questions concerning perceived barriers to increased participation, and
questions concerning various socialization experiences that might affect later participation in the
arts and culture.
Nearly two-thirds of the American adult population would like to attend art museums and
galleries more often. Over half the population would like to attend both musical plays and
non-musical plays more often. There is less interest in increased participation in the other art
Residents of California, New York and New Jersey report more interest in increased
participation for all of the eight key art forms than do the residents of the United States on
average. Because these three states generally turn up as high participation states in many of the
analyses reported here, this might be due to a concentration of cultural institutions in these
states raising the population's expectations or the demand of a population that, socio-economically,
is particularly inclined toward these forms of cultural consumption.
Residents of Pennsylvania and Ohio, on the other hand, show less interest than average in
- What is most important to notice, however, is that an interest in increased participation is
expressed much more often by those who have attended a particular art form in the previous twelve
months than by those who have not, and this is true irrespective of the state under consideration.
With respect to barriers to increased attendance, the most often cited reason, cited by nearly
two-thirds of those who would like to attend more often, is a broad one: "It is difficult to make
time to go out." Roughly half of those who would like to attend more often cite "Tickets are too
expensive," "There are not many performances held or art museums or galleries in my area," and "The
location is usually not convenient." These reasons are more susceptible to policy intervention.
Nearly half of American adults report having had lessons or classes in music at one time or
another in their lives. Roughly one-quarter reports having taken lessons or a class in each of the
following: the visual arts, creative writing, art appreciation or art history, and music
appreciation. Lower percentages have had acting or dance lessons.
California, Florida, Massachusetts and New Jersey have higher than average levels of
socialization for all of the eight key art forms. Texas, on the other hand, has lower than average
Finally, Section 10 of this monograph, uses the SPPA data to construct demographic profiles of the
audiences for various art forms, facilitating comparisons across art forms as well as across states.
The analyses reported in this section consider four important demographic variables: education
level, income level, race/ethnicity, and gender. A careful distinction is drawn in this section
between an audience profile of visitors (separately identifiable individuals making no adjustment
for their relative frequency of attendance) and an audience profile of visits (adjusting for the
fact that some visitors attend more frequently than others).
Visitors are more highly distributed toward upper educational levels than the overall
population, clearly indicating the importance of education in predicting whether someone will be a
visitor to any of the art forms.
Because individuals with higher educational levels also have higher frequencies of attendance,
the distribution of visits by educational level is even more highly skewed toward individuals with
higher levels of education.
Upper income individuals are over-represented among visitors to each of the art forms.
Weighting individuals by their frequency of attendance and constructing an income distribution
of visits results in a more complicated picture because frequency of attendance does not necessarily
rise with household income and the pattern differs for different art forms.
With respect to race and ethnicity, the patterns become more complex. According to the SPPA
data, members of certain minority groups are under-represented among visitors to some art forms,
while they are over-represented among visitors to others. The same is true of the profile of visits
to various art forms.
Women are over-represented among visitors to all of the eight key art forms except jazz. With
respect to visits, however, they are under-represented in the audiences for jazz, classical music,
and dance forms other than ballet.
While these are the main findings of this monograph, they only begin to scratch the surface of the
detail contained in these pages. Before encouraging the reader to wade into the main text, however,
it is necessary to add a few words of caution to aid in the interpretation of the findings.
Because the SPPA data are the result of sampling, all of the estimates of participation rates
in this monograph are subject to random sampling error. Because of that error many of the observed
differences in participation rates may be attributable, at least in part, to random sampling error
rather than to any real differences in participation rates. Only very large observed differences are
likely to be immune from this complication. This issue is discussed at some length in Section 2 of
the monograph, but is very much present in the other sections as well.
Each of the correlation analyses that looks at all of the ten identified states simultaneously
needs to be understood in a rather modest manner. Because these ten states are not a simple random
sample of the fifty states, the results of these analyses cannot be generalized to all of the
states. They simply measure the correlation that one observes when looking at various pairs of
variables across this particular set of ten states.
These caveats notwithstanding, it is our hope that with the analyses contained in these pages we
have begun a fruitful inquiry into the geographic variation in participation across the United
States. Perhaps the SPPA does not afford the ability to produce the definitive analysis that might
be desirable, but it does provide a solid base of data on which future research and inquiry can be
The full report is available from Seven Locks Press; P.O. Box 25689; Santa Ana, CA 92799.
Telephone: 714-545-2526 or 1-800-354-5348. Paper $10.95. 182 pp.
National Endowment for the Arts · an independent federal
1100 Pennsylvania Avenue NW
Washington, DC 20506