Co-Patronage Patterns in Arts-Related Leisure Activities

ABSTRACT - Analysis of two studies of performing arts patronage indicate considerable co-patronage among the arts. The greater the attendance frequency at one art form the greater the attendance at other entertainment events (other arts, sports, motion pictures, rock concerts). Demographics were similar for theater and symphony patrons which were both more upscale than the general population.


Russell W. Belk, Richard J. Semenik, and Alan R. Andreasen (1981) ,"Co-Patronage Patterns in Arts-Related Leisure Activities", in SV - Symbolic Consumer Behavior, eds. Elizabeth C. Hirschman and Morris B. Holbrook, New York, NY : Association for Consumer Research, Pages: 95-100.

Symbolic Consumer Behavior, 1981     Pages 95-100


Russell W. Belk, University of Utah

Richard J. Semenik, University of Utah

Alan R. Andreasen, University of Illinois


Analysis of two studies of performing arts patronage indicate considerable co-patronage among the arts. The greater the attendance frequency at one art form the greater the attendance at other entertainment events (other arts, sports, motion pictures, rock concerts). Demographics were similar for theater and symphony patrons which were both more upscale than the general population.

In a recent review of 270 published and unpublished audience studies conducted since 1970, DiMaggio, Useem, and Brown (1977) were able to find only eight studies examining the co-patronage of various cultural arts forms, i.e. , the extent to which patronage of one arts form is associated with patronage of another, Nevertheless, it is generally assumed by arts manager that they are drawing from a common pool of aficionados and that in this sense they may be competing with other arts forms. While the research to date generally tends to support the assumption of a high degree of overlap in arts audiences, it also suggests some possible exceptions which, if more broadly supported, would make the concept of "cultural arts patron" an over-simplification. The present research examined such crosspatronage patterns in come detail using the data from two large and disparate studies in order to establish a broader base from which to generalize findings.


One superficial but highly consistent indication of arts audience overlap has been the similarities found in the demographic characteristics of arts audiences. With rare exceptions, the studies reviewed by DiMaggio, Useem, and Brown (1977) found that audiences for theater, dance, ballet, classical music, opera, art museums, history museums, and science museums are all more highly educated, higher in occupational status, higher in income, and less likely to include racial minorities than the population as a whole. Even though ail of these audiences tended to be "up-scale" compared to the general population, there were many variations in the audience profiles in the various studies, including some which appear to be related to the art form. These differences suggest that museum patrons are lower in income and occupational status than performing arts patrons (although this was less true of the somewhat older arts museum patron) and that theater audiences are slightly less educated, less wealthy, and less likely to be composed of professionals than are the audiences at other performing arts events.

The hypothesis that theater audiences are in some way different from audiences of other performing arts may also be supported by the studies of Baumol and Bowen (1966) and the Ford Foundation (1974) which found less overlap between theater audiences and audiences of other performing arts than between the other performing art forms. For example, in the Ford Foundation study, 63 percent of those attending theater in the past year had attended none of the other art forms, whereas this was true of only 25 percent of opera attenders, 36 percent of symphony attenders, and 20 percent of ballet attenders.

Another finding of several prior studies which may suggest that the designation "cultural patron" may be overly simplistic is the observation that there tends to be much greater overlap across art forms among frequent arts attenders than among occasional arts attenders. There is also some evidence that frequent arts attenders are more likely to be involved in a social network of arts patrons and to hear about arts events from their friends (National Research Center of the Arts, 1975), However a partial explanation for these findings may lie in the fact that frequent attenders are over-represented relative to infrequent attenders in studies which use the audience at a particular arts event or institution as a sample.

A more general approach to considering whether there is a "Cultural arts patron" is to examine the frequency with which individuals participate in a variety of leisure activities and then consider whether the correlations between arts events participation are all higher than the correlations between participation in some of the arts activities and participation in other leisure activities. The studies taking this approach have used a variety of measures, activities and analytical methods, and have obtained an equal variety of conclusions. Using a set of 47 leisure activities, Hawes, Talarzyk, and Blackwell (1975) performed a factor analysis of participation in these activities and found that attending concerts cr plays was the second highest positive loading for a factor which also was represented by camping (for females) or by hiking, volunteer work, golf, job-related study, and water sports (for males). Holbrook and Lehmann (1980) examined frequency of participation in 50 leisure activities and found some support for the a priori expectation that the "audience activities" J popular concerts, classical concerts, TV sports, movies, X-rated movies, sports events, car races, and museum attendance would be more related to each other than to other leisure activities. Sexton and Britney (1980) conducted a factor analysis of a more limited set of 10 leisure activity attendances and obtained three factors which might perhaps be seen as "high culture" (ballet, opera, and symphony), museums (of art, history, and science), and "low culture" (modern dance, jazz, plays, and movies). And several studies such as the one by the National Research Center of the Arts (1975) have found that frequent attenders of arts events and institutions are also frequent participants in a variety of activities including sporting events, movies, creative activities, and even the circus. Thus a clear answer to the question of whether attendance at various arts activities tends to cluster together and/or with other leisure activities does not emerge from past research.


In an effort to clarify the issues raised concerning cc-patronage of the arts, a reanalysis of two quite different studies of performing arts and related activities was undertaken. The hypotheses to be tested were those drawn from the literature just reviewed, and may be summarized as follows.

1 . In the context of leisure activities, attendance at the cultural activities of theater, symphony, ballet, opera, and museums will overlap and cluster together with each other more than with other leisure activities.

2. Attendance at these arts events and institutions will also be positively related to attendance at motion pictures, rock concerts, and sports events.

3. The greatest overlap in the attendance hypothesized to be related in the first two hypotheses will be among the heavy attenders of such activities; occasional attenders will show little if any of the hypothesized patterns of co-patronage.

4. Theater and museum attendance will be less related to ballet, opera, and symphony attendance than is true of other possible pairings.

5. Attenders of theater, museums, ballet, opera, and symphonies will be higher in education, income, and occupational status, and more often non-minority members than the non-attenders of these events.

6. Among these arts events, theater and museum patrons will be lower in education, income, and occupational status, and more often members of minority races than ballet, opera, and symphony patrons.

7.- Frequent attenders of arts activities will be more likely to have obtained arts information from friends than will. non-attenders and occasional attenders.


The first of the two studies to be used in hypothesis testing was a study of potential patrons of the performing arts in the southern United States. The second was an investigation of opera audiences at several locations throughout the United States. While the details of these studies are reported elsewhere (Belk and Andreasen 1980; Semenik and Young 1980), Table 1 shows a summary of the methods involved. As may be readily seen, these were two quite different studies in terms of populations sampled and data collection methods. However, because of the similarity in data collected, the two studies offer a unique opportunity for hypothesis testing in that the same relationships may be tested on two substantially different populations. This opportunity for establishing corroborative evidence based on parallel analyses wherever possible (e.g., using the same age and income units of measurement, the same definitions of heavy and light attendance, and the same analyses and statistics) also provides the potential for resolving some of the apparent conflicts from prior studies.


Clustering of Arts-Related Leisure Activities, Interests, and Opinions

Because of greater time constraints, the opera study was only able to collect activity, interest, and opinion (AIO) information on a subset of the items included in the theater and symphony study. The following analysis is thus based only on the latter study, but where parallel items exist the opera study does offer consistent support for these results.

The AIO items of interest were a subset of 50 items concerning leisure activities, interests, and opinions. The intent of this analysis was to derive what may be seen as "leisure lifestyles" and consider whether the hypothesized clustering of arts activities formed the basis for one or more of the resulting lifestyle groups. To this end, a Q-type factor analysis of responses to the 50 item was performed seeking to group respondents into mutually exclusive categories on the basis of the patterns of their responses. Using a program developed by Johnson (1970), a six group solution was adopted after comparing the stability of several solutions when applied to split halves of the sample. These six leisure lifestyle groups are described in Table 2 based on the AIO items on which they were significantly different.

The hypothesized clustering of arts-related activities, interests, and opinions very clearly emerged in the group dubbed "arts patrons". The group is characterized by a general positive arts orientation and a negative television orientation. Contrary to some of the findings of Baumol and Bowen (1966) and the Ford Foundation (1974), symphony, theater, and museum patrons appear to be from among the same lifestyle group rather than from distinct groups. (This counter-hypothesis is considered more fully in the next section.) It is also worth noting that several of the other lifestyle groups such as the active sports enthusiasts show an antipathy or at least an apathy toward the performing arts. This may suggest that the arts are an all-or-nothing proposition in adopting leisure lifestyles, but since the clusters force individuals into one group or another this possibility will be considered more fully in a later section.

Overlap in Arts Audiences and the Effect of Attendance Frequency

Both studies contained data on the number of times respondents had attended rock concerts, museums, ballet, theater, and symphony performances in the previous 12 months. To split the samples into heavy and light attenders for these events and institutions, considerations of consistency and sample sizes led to the definition of "light attendance" as attendance at one or two events and "heavy attendance" as attendance at three or more events. This allowed separate comparisons of three levels of attendance at one event (heavy, light, and none) to attendance at another event. The results of cross-classifications of the three different attendance levels at one event with the three different attendance levels at another, showed that of the 15 pairs of events all but rock concert/ballet and rock concert/symphony were significantly related in the opera study, and all but rock concert/ballet were significantly related in the theater and symphony study.

The nature and level of these relationships may be seen in Table 3 which summarizes the proportions of heavy attenders, light attenders, and non-attenders of each art form who had attended at least one performance or exhibit of each of the other art forms in the past year. Two general patterns emerge consistently from the two data sets. The first is that in both data sets virtually all relationships are positive; attendance at one art form is likely to be coupled with attendance at another art form. The few exceptions which exist are for ballet and, to a lesser degree, museums. But only for heavy attenders of ballet does the average likelihood of a heavy attender to attend the other art forms drop below 50 percent, and this is true only for the theater and symphony study. In general then, the past year's attendance at one art form is highly related to the past year's attendance at other art forms. This supports previous research findings except that the positive relationships to rock concert attendance may be something of an intuitive surprise.







The second general pattern to emerge in Table 3 is that, going from the non-attender groups to the light attender groups to the heavy attender groups, there are increasingly higher incidences of attending other art forms. Comparison of the heavy and light attender groups supports the hypothesis that the greatest copatronage of arts occurs among heavy attenders. However the comparison of the light and non-attender groups also shows that light attenders are more likely to patronize other arts. Differences in the level of co-patronage between heavy and light attenders thus may derive from the fact that the former simply have a higher base rate of overall attendance frequency.

One further result in Table 3 requiring comment is the general similarity of the co-patronage percentages of the two studies except for rock concerts and symphonies. Although the patterns of these percentages in the two studies is very similar, the level of co-patronage of rock concerts is much higher for the theater and symphony study, and the level of co-patronage of symphonies is much higher for the opera study. An immediate but superficial explanation lies in the differences in base rates for attendance at these two events by the two samples. Rock concerts were attended by 35 percent of the theater and symphony study sample versus 8 percent of the opera study sample; and symphonies were attended by 61 percent of the opera study sample versus 14 percent of the theater and symphony study sample. However, a more complete explanation must consider the nature of the samples. Although not presented in Table 3, analysis showed that heavy opera attenders in the opera study co-patronized rock concerts at a rate of less than 7 percent compared to their average co-patronage of the other arts at a rate of nearly 67 percent. Thus drawing a sample of opera attenders has greatly suppressed the co-patronage levels possible for rock concerts. Since these opera attenders co-patronized symphony at near their average 67 percent level, the sample did not have the same limiting effect there. It should also be noted that the theater and symphony study was conducted in the Southeastern United States which has the lowest arts attendance levels of any U.S. region (National Research Center for the Arts, 1975). This resulted in the low symphony attendance level in this sample and caused a suppression of the levels of co-patronage of the symphony which mirrored that which occurred for rock concerts in the opera study. Both of these occurrences suggest that base rates should be carefully examined in considering the absolute levels of co-patronage found in any particular study, but that comparisons of relative patterns of co-patronage among heavy, light, and non-attenders should be substantially free of such base rate effects.

Finally, the hypothesis that theater and museum attenders will engage in less co-patronage than attenders to other art forms receives little support from the data in Table 3. Both data sets reveal that theater attenders and museum attenders are well integrated in their attendance to other art forms. The data from the opera study indicates that 83% of heavy attenders to theater were heavy attenders at museums, ballet, and symphony. The theater and symphony study shows 58% of heavy theater attenders were co-patrons of other arts offerings. The same pattern holds true for museum attenders where over 50% of the museum attenders in both studies were attenders at other artistic offerings.

Arts Audiences and Attendance at Other Events

In general, the data from the current studies support the hypothesis that attendance at arts events is positively related to attendance at motion pictures, live sporting events, and rock concerts. Both the opera study and the theater and symphony study found that a greater percentage of attenders at arts events were heavy attenders of motion pictures than the non-arts attenders.

However, with regard to live sports events, the results are mixed. The opera study data revealed that theater, symphony, and museum goers were also patrons of live sports events (33%, 35%, and 34% respectively) to a greater degree than non-goers to these art forms. Conversely, the theater and symphony study found a negative relationship between ballet and symphony attendance and attendance at sporting events.

Finally, both studies showed a positive relationship between arts attendance and rock concert attendance (Table 3). The theater and symphony study is most supportive here because of the larger number of rock concert attenders in the sample. A greater percentage of heavy and light attenders of all the arts offerings studied attended rock concerts than non-arts attenders.

Arts Attendance and Demographic Characteristics

The hypothesized relationships between arts attendance and non-attendance regarding education, income, occupational status and non-minority racial status were supported by the data from both studies. Attenders of theater, museums, ballet, opera, and symphony were higher in income, education, and professional employment status. The notable exception was the income status of the ballet attenders from the theater and symphony study. Further, these arts offerings attracted a predominately white audience although ballet did have 20.7 percent black attenders.

In comparing demographic profiles of attenders across arts events, Baumol and Bowen (1966) and the Ford Foundation (1974) found that theater and museum patrons were less "upscale" in education, income, and occupational status, and more often non-white than ballet, opera, and symphony patrons. The results of the two current studies provide little support for these earlier results. Patrons to all the art offerings studied tended to be upscale to the same degree on education, income and occupational characteristics with the exception of the ballet attenders' income status as stated earlier. Further, non-white attendance was found more often for ballet than for theater or museums.

Arts Attendance and the Influence of Friends

The final hypothesis regarding the nature of co-patronage to arts events suggests that frequent attenders of arts activities will be more likely to have obtained information from friends than non-attenders. The data from both the opera study and the theater and symphony study do not support this allegation. The results of both studies are unambiguous in finding that frequent attenders rely less on friends for information about arts events than the non-attenders and occasional attenders. This is not to say that frequent attenders are not influenced by friends. Rather, the influence is not as prevalent as it is for non-attenders and occasional attenders. We should perhaps distinguish verbal and non-verbal influence from friends however. Even though heavy attenders report relying less on friends than do light and non-attenders, they also consistently report having more friends who like symphony and theater and having more friends who expect them to attend these art forms. Thus while it may be true that they rely less on friends for verbal information about the arts, they may still be greatly influenced by the behavior of their friends.


The analysis of two quite different studies presented here was undertaken in an attempt to contribute to an understanding of the issues surrounding co-patronage of the arts. To this end, the following conclusions can be drawn regarding the hypotheses set out earlier.

1. In general, it was reaffirmed that arts attenders are quite homogeneous in demographic characteristics.

2. Differences in demographic characteristics for museum and theater patrons were not discovered in the current analysis.

3. The co-patronage behavior of theater patrons was found to be well integrated contrary to previous research findings.

4. Heavy attenders of artistic offerings do, in fact, engage more heavily in patronage of other events (motion pictures, live sporting events, rock concerts) than non-heavy attenders of artistic offerings.

5. Heavy attenders of individual art forms demonstrate the greatest amount of co-patronage to other art forms.

6. The current analysis does not support the allegation that heavy attenders at artistic events are more likely to hear about such events from friends, although they may be influenced by their behavior.


William J. Baumol and William G. Bowen, Performing Arts--The Economic Dilemma (Cambridge: M.I.T. Press, 1966).

Russell W. Belk and Alan R. Andreasen, "De Gustibus Non est Disputandum: A Study of the Potential for broadening the Appeal of Performing Arts," in Jerry Olson (ed.), Advances in Consumer Research, Vol. 7, 1980.

Paul DiMaggio, Michael Useem, and Paula Brown, Audience Studies of the Performing Arts and Museums: A Critical Review (Washington, D.C.: National Endowment ior the Arts, 1978).

Ford Foundation, The Finances of the Performing Arts, Vol. 1 (New York: Ford Foundation, 1974).

Douglass Hawes, Wayne Talarzyk, and Roger Blackwell, "Consumer Satisfaction from Leisure Time Pursuits," Mary Jane Schlinger (ad.), Advances in Consumer Research (Ann Arbor: Association for Consumer Research, 1975).

Morris B. Holbrook and Donald R. Lehmann, "Patterns of Allocating Discretionary Time: Some Issues in Assessing Complementarity Among Activities," mimeographed paper, Columbia University, 1980.

Richard M. Johnson, "Q Analysis of Large Samples," Journal of Marketing Research, 7 (1970), 104-105.

National Research Center for the Arts, Inc., Americans and the Arts (New York: National Committee for Cultural Resources, 1975).

Richard J. Semenik and Clifford E. Young, "Correlates of Season Ticket Subscription Behavior," in Jerry Olson (ad.), Advances in Consumer Research, Vol. 7, 1980.

Donald E. Saxton and Kathryn Britney, "A Behavioral Segmentation of the Arts Market," in Jerry Olson (ed.), Advances in Consumer Research, Vol. 7, 1980.

Gordon L. Wise, "Differential Pricing and Treatment by New-Car Salesmen: The Effect of the Prospect's Race, Sex, and Dress," Journal of Business, 47 (1974), 218-230.



Russell W. Belk, University of Utah
Richard J. Semenik, University of Utah
Alan R. Andreasen, University of Illinois


SV - Symbolic Consumer Behavior | 1981

Share Proceeding

Featured papers

See More


Love is Blind: How Sensory Liking Impacts Perceptions of Unbranded Products

Jennifer L Stoner, University of North Dakota
Maria A Rodas, University of Minnesota, USA

Read More


I'm Scared, Want to Listen? Fear's Influence on Self-Disclosure

Anupama Mukund Bharadwaj, University of Washington, USA
Lea Dunn, University of Washington, USA
Joey Hoegg, University of British Columbia, Canada

Read More


Thank You = Trust Me: When Gratitude Expressions Help Promote New Products

Maria Ng, University of Houston, USA
Dejun Tony Kong, University of Houston, USA
Vanessa Patrick, University of Houston, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.