De Gustibus Non Est Disputandum: a Study of the Potential For Broadening the Appeal of Performing Arts

Russell W. Belk, University of Utah
Alan R. Andreasen, University of Illinois
ABSTRACT - In view of the small proportion of the population represented in audiences attracted to the performing arts, this study examines whether a number of variations in theatre and symphony offerings might attract prospects who are currently not patrons without alienating those who are. While several limited epportunities for broadening performing arts audiences are found in the four cities sampled, a long-term strategy of broadening consumer tasks beginning in childhood is suggested as a prerequisite for any substantial changes in the character of performing arts audiences.
[ to cite ]:
Russell W. Belk and Alan R. Andreasen (1980) ,"De Gustibus Non Est Disputandum: a Study of the Potential For Broadening the Appeal of Performing Arts", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 109-113.

Advances in Consumer Research Volume 7, 1980     Pages 109-113


Russell W. Belk, University of Utah

Alan R. Andreasen, University of Illinois


In view of the small proportion of the population represented in audiences attracted to the performing arts, this study examines whether a number of variations in theatre and symphony offerings might attract prospects who are currently not patrons without alienating those who are. While several limited epportunities for broadening performing arts audiences are found in the four cities sampled, a long-term strategy of broadening consumer tasks beginning in childhood is suggested as a prerequisite for any substantial changes in the character of performing arts audiences.


Audiences for live theatre, dance, symphony, and opera performances comprise a narrow and elite portion of the American population. In a recent review of over 200 audience studies in the performing arts, DiMaggio, Useem and Brown (1978), found consistent evidence that those attending these events are better educated, higher in job status and income, and less likely to include minorities than the general population. In addition, these researchers found no evidence that this elitism in the arts is declining. A major purpose of the present research was to asses whether, with reasonable alterations in various features of performing arts offerings, it may be possible to attract broader, more diversified audiences.

Toward that end, it was first necessary to develop a methodology for measuring likely consumer responses to modifications in arts offerings. Since a "broader" audience implies attracting new attenders, merely measuring preferences for and likely behaviors toward present offerings would not suffice. The intent was to develop a means of assessing likely responses to varied offerings before going so far as to implement the variations. Furthermore, since any change from existing arts offerings may affect current audiences as well as potential new patrons, the measures had to be sensitive to possible negative as well as positive changes in dispositions of different groups of consumers. In other words, a broader audience cannot be achieved by simply exchanging each newly enticed audience member for one that has been alienated.

The opportunity to conduct the present study arose from a National Endowment for the Arts research grant within a program calling for an assessment of ways to "make arts and cultural activities more widely available to millions of Americans" (NEA, 1976). Research activities were focused on the Southern United States, where arts attendance at present involves the smallest portion of the population (National Research Center of the Arts, 1976). In addition, to keep the scope of the project manageable, the present research was restricted to symphony and theatre attendance. Within this framework, the study was designed to assess whether broader audiences are possible within a limited time frame through the marketing activities of an individual theatre or symphony organization.



To insure a concrete reference point for answering questions about actual or potential symphony and theatre attendance, only communities having both resident theatres and symphonies were considered as potential research sites. Of the several dozen southern cities that met this criterion, four were judgmentally selected for investigation: Atlanta, Georgia; Memphis, Tennessee; Columbia, South Carolina; and Baton Rouge, Louisiana. While subsequent analyses showed some differences in attitudes toward the theatre and symphony between these four cities, differences did not appear in the data being considered here, which have therefore been pooled across cities for purposes of presentation.

Further restrictive sampling criteria were also employed. By conducting telephone interviews, households without telephones were intentionally excluded on the grounds that they were unlikely theatre and symphony patrons. Once a household was reached, respondents systematically selected from among those members greater than 14 years in age. A final set of restrictive criteria was embodied in screening questions intended to undersample "heavy attenders", oversample "marginal attenders", and exclude "unlikely attenders". The "heavy attender" group was of interest to detect possible alienation from some offering modifications and was defined to include those who reported attending three or more theatre or symphony performances in the past year. [Screening questions were imbedded in a list of questions about other leisure activities including those used to define the "marginal attender" group.] Based on pretest results, approximately half of the heavy attenders reached were not interviewed, [Of those screened, approximately 15 percent were identified as "heavy attenders".] leaving about 11 percent of the final sample in this category. [Of these heavy attenders, 77 percent were heavy attenders of theatre, 9 percent were heavy attenders of symphony, and the remaining 14 percent were heavy attenders of both theatre and symphony.] The "marginal attender" group was of greatest interest because they were conceived as most likely to comprise any broadened audiences that might be achieved. Intuitively and based on prior audience studies, the marginal attender group was defined as those other than heavy attenders who:

1. In the past year --

a. saw a live play

b. attended a symphony concert or other live classical music performance

c. saw a live or televised ballet performance

d. visited an art gallery or museum

e. attended a live popular or rock music concert,


f. listened at least 10 times to recorded or broadcast classical music

2. or

a. play a musical instrument

b. once worked in a theatre, music, or dance production

c. attended three or more live plays, but none of them in the past year, or

d. attended three or more symphony concerts, but none in the past year.

The "unlikely attender" group was defined as all those who failed to meet any of these criteria. This group comprised 14 percent of those screened and was excluded from the final sample.

Of 3,956 residential telephone numbers attempted via random digit dialing, 56 percent were successfully screened after five callbacks. After eliminating unlikely attenders, half of the heavy attenders, incomplete interviews, and refusals, 1491 observations remained and formed the data base for the present analysis. These were drawn in approximately equal numbers from the four cities and sampled. Compared to the general populations of these cities, the present sample is younger, better educated, higher in income, and more likely to be female. These biases probably result from drawing a sample of "heavy attenders" and "marginal attenders" of the arts and from the use of telephone interviews for data collection. Thus, even this broadened potential audience shows bias in the direction of present audience typical demographic characteristics. The fact that these biases could not be wholly accounted for by the inclusion of heavy attenders of theatre and symphony suggests that, at least in terms of demographics, potential new patrons may be more similar to present patrons than to unlikely attenders.

Interviews were conducted by continuously supervised professional interviewers located in each of the four sample cities. In addition to the screening questions, data included attitudes toward locally available theatre and symphony offerings, general and leisure-specific lifestyles, demographic and life-cycle characteristics, childhood exposure to theatre and symphonic music, and probable responses to a series of altered offerings in the theatre and/or symphony.

It is the latter probable response data that are of interest here. One final exclusionary criterion was employed prior to collecting these particular data. All respondents, whether heavy or marginal attenders of either theatre or symphony, were asked if they were very likely, somewhat likely, not very likely, or not at all likely to attend 1) a symphony concert and 2) a theatre performance in "the next year or two." Those who responded in one of the two "not likely" categories were not asked about their probable responses to the corresponding symphony or theatre offerings. This left 431 respondents for the symphony questions and 657 for the theatre questions. As would be expected, few of the previous heavy attenders were excluded by this procedure. Slightly under 20 percent of those responding to theatre questions were heavy attenders and slightly under 10 percent of those responding to symphony questions were heavy attenders. Since approximately 13 percent of those initially screened qualified as heavy attenders, it may appear that these subsequent exclusion criteria have restored the balance of sample groups responding to the offering questions to their approximate proportions in the original population sampled. However, the somewhat involved screening procedures have actually upgraded both the heavy and (especially) the marginal attender groups by removing those now considered least likely to attend and therefore presumably the least likely to respond positively to any of the changes in offerings. Nevertheless, since the proportion of those responding to these offerings who are heavy attenders now compares to the proportion initially contacted who were heavy attenders, differential weighting of their responses is not necessary. This also allowed the heavy/marginal attender distinction to be replaced in analysis by more nearly equal-sized dichotomies of those who have and have not attended each type of performance in the past year. [Of those 637 responding to theatre offerings, 40 percent had attended a theatre performance in the past year; for symphony 34 percent of the 431 responding to these offerings had attended a symphony concert in the past year.]


The potential offering modifications were selected on the basis of focus group interviews with attender and non-attender groups, an inventory of apparently successful changes in other parts of the country, and a consideration of other marketing elements that might be fruitfully altered. In the context of a telephone interview, it was not possible to investigate differing promotional strategies, but changes in the "product," context, and price were all considered.

The manipulations took the following form:

1. Product . . . Only for this group of variables did the variations differ between theatre and symphony. The three variations were in?

a. Type of performance

1. Theatre -- more . . .

i) Musical comedies such as "South Pacific'' or "Showboat"

ii) Classical plays such as "Hamlet" or "Macbeth"

iii) Well-known American dramas such as "Death of a Salesman" or "A Streetcar Named Desire"

iv) Modern comedies such as "The Sunshine Boys"

v) Original plays that have never been done before

2. Symphony -- more . . .

i) Symphonies by classical composers such as Mozart and Hayden

ii) Symphonies by romantic composers such as Brahms and Tchaikovsky

iii) Music by contemporary composers such as Stravinsky

iv) Concertos by soloists

v) Choral music

b. Quality of Performance [It is recognized that the fame of the performer and the quality of the performance are not perfectly correlated, but the former was thought to be a more likely cause of new audiences than the latter alone.]

1. Theatre -- "If famous actors and actresses appeared with the company more frequently"

2. Symphony -- "If guest conductors and famous soloists appeared with the orchestra more frequently"

c. Extent of learning opportunities

1. Theatre -- "If there was a short discussion of the play by the director after the performance"

2. Symphony -- "If there was a short introductory talk about the music by the conductor before the performance.

2. Context ---These manipulations concerned the setting of the performance and the nature of the performance itself.

a. Location and conditions of performance -- "Suppose that symphony (theatre) performances were given five times a year in a location nearer your home. The performing space wouldn't be as nice as (name of major theatre or concert hall in city) but the prices would be 20% lower."

b. Formality of atmosphere -- "If you knew that people were dressing more informally at the concert (theatre)"

3. Price -- In addition to the recognition of price in 2a. and 2c. above, six separate price variations were considered.

a. Base ticket price -- "If individual tickets were

i) reduced by $1

ii) reduced by $2

iii) reduced by $3

iv) increased by $1

v) increased by $2

vi) increased by $3"

b. Couples discounts -- "If after purchasing one ticket at regular price you could get a second ticket at 50% off"

c. Series discount -- "If you could get series tickets guaranteeing good seats for several symphony performances (plays at a . . .

i) 10% discount

ii) 20% discount

iii) 30% discount"

d. Clearance discount -- "Suppose that next year unsold tickets for performances of the (name of symphony or theatre) could be obtained at regular ticket outlets for 50% off on the day of the performance. The seats usually would not be as good as those bought in advance."

e. Effort required to secure tickets -- "If tickets could be purchased by telephone and charged to a national or department store credit card."

f. Price and performance combination -- "Suppose that your favorite kind of music (play) (as ascertained in a previous question) were presented more often during the year, but ticket prices were:

i) increased by $1

ii) increased by $2

iii) increased by $3

Apart from combinations of several offering changes reflected in certain of these manipulations, no attempt was made to consider all factorial combinations of potential changes. This would have been operationally infeasible and would also have resulted in some unlikely combinations. Those combinations included were attempts to reflect plausible trade-offs and to assess price elasticities under varied conditions.


Two basic types of projective responses to each change were solicited. For questions involving a repetitive sequence of price levels (3a, 3c, 3f), respondents were asked to indicate whether each amount of reduction (increase) would cause them to attend more (fewer) performances. If an affirmative answer was received before the end of the price sequence, the remainder of the sequence became superfluous and was omitted. All other variations were assessed by asking the respondent whether under those circumstances, they would attend theatre or symphony performances much more often, somewhat more often,

Responses to such questions cannot be interpreted literally since they will most likely show a positive acquiescence bias. There is, however, no reason to suspect that such a potential bias differs across the offering questions or between the attender and non-attender groups. Thus relative comparisons are possible concerning whether attenders or non-attenders are most affected by a particular offering and which type of offering (within each of the two response formats) most affects each group. In addition, while the magnitude of any change in attendance disposition is not estimable from these data, it is probable that the direction of such changes is accurately reflected. For analysis, the data were regarded as ordinal and tested using non-parametric statistics. However, to clarify the relevant comparisons assessing whether broader, more heavily "non-attender" audiences can be attracted by various offerings, aggregate indices of offering attractiveness were constructed and are used in presenting results. For questions not involving presentations of repetitive price sequences ("Index A"), responses were scored as follows: "much more often" = +2; "more often" = +1; "as often" = 0; and a "less often" =-1. The results of this monotonic transformation were summed over respondents and then divided first by the number of respondents to the question and then again by the average of these scores over all 12 such offerings. The resulting aggregate scores were multiplied by 100 so that 100 is then the "average" effectiveness score over all offerings and respondents. Scores higher than 100 reflect above average effectiveness while lower scores reflect below average response. For the three questions involving presentations of repetitive sequences, the construction of an index ("Index B") was analogous to that just described. Here if the lowest reduction in price (or the highest increase in price) generated a favorable response it was scored as a +3, while other responses received correspondingly lower scores. Again the index was averaged over respondents and price offerings, and again 100 is the average effectiveness score.


The aggregate scores for "Index A" offerings are shown in Table 1 for theatre and Table 2 for symphony. Table 3 presents the "Index B" scores for both theatre and symphony.







To be able to conclude that an organization's audiences can be selectively broadened in the areas sampled, l) some of the offerings should be found to be highly effective with last year's non-attenders, and 2) these offerings should be more effective with this group than with last year's attenders, [The reason for this criterion requiring that arts audiences be selectively broadened by the addition of non-attenders is that if the criterion were omitted and strategies which affect attenders and non-attenders equally were accepted, there is a strong possibility that the net composition of the audience would be unchanged! This is due to the fact that each year audiences already gain some patrons not in attendance the prior year, plus the fact that the screening criteria generating the present sample of non-attenders makes them nearly as likely to respond favorably as attenders. An examination of raw scores as well as the meager number of significant differences in the tabled index scores support this contention and necessitate the second criterion.] but 3) these offerings should not decrease the probability of attendance by last year's attenders. The results for theatre show that the first condition is met; some of the alternative offerings elicited strong positive responses from non-attenders. As reflected by high index scores, the theatre offerings of more musical comedies, more American dramas, more famous performers, and a couples discount all fulfill this criterion. Of these, however, none has a differential impact on attenders versus non-attenders. For the symphony offerings, more famous performers, a short introductory talk, a more convenient location, discounts for couples, clearance, and series and base ticket price decrease are most effective strategies for attracting non-attenders. Of these only the more famous performers are more convenient location offerings are significantly more effective with non-attenders than with attenders. Examinations of raw scores show that the famous performers offering also fulfills the third criterion of not decreasing the likelihood of attendance by prior attenders, but the location change does not do well in this regard. Thus, if the present findings are valid and generalizable, the final conclusion with respect to broadening is mixed. The theatre audiences do not appear susceptible to selective broadening by the offering changes considered and the symphony audiences appear to be susceptible only by the introduction of more famous performers and conductors. Even here, however, a more extended feasibility study would be required to justify the costs of adding more famous performers, especially on a permanent basis. This is not to rule out the viability of such a strategy, but simply to underscore the necessity of considering the supply side of the analysis as well.


Although the problem of assessing the impact of particular marketing strategies on different segments is a general one faced by marketers, it assumes special significance for arts managers. In the arts, there may be two additional imperatives beyond running a viable business. One of these arises from the conviction that art is at least in part a public good so that society benefits from the existence of arts offerings whether or not a large audience actually attends them. But this conviction may also be taken to imply that an arts organization should cater to as broad a public as possible. Another additional mandate for an arts organization is to foster, preserve, and enhance the art it represents. This third objective of "maintaining artistic integrity" may conflict, however, with the mandate to serve a broad public (as well as with the need to run a viable business), thereby posing another barrier to broadening arts audiences that is not readily apparent in the results presented.

This barrier may arise when a strategy that might effectively broaden an audience conflicts with the organization's interpretation of its artistic mission. Discussions with arts managers suggest that this is a frequent occurrence. Further analysis of the present data by leisure lifestyle groups, for example, suggested that several lifestyle groups other than hard-core culture patrons could be attracted to theatre offerings by presenting more musical comedies. However, a great majority of the theatre managers to whom this finding was presented, [At the conference on Policy Related Studies of the National Endowment for the Arts, Baltimore, Maryland, December 7-9, 1977.] considered such offerings a threat to the artistic integrity of their theatres.

Without debating the relative merits of serious versus non-serious art or the entertainment versus educational goals of potential arts attenders it may be possible to offer a resolution of such apparent goal conflicts that preserves artistic integrity, broadens arts attendance, and does both in an economically viable way. The resolution is simply to segment audiences for the performing arts and provide different offerings to different segments. The success of such methods depends, however, upon one unresearched facet of consumer perceptions of performing arts offerings. The implications may be different for the theatre and symphony, but if either type of arts organization tries to appeal to multiple market segments by offering a "mixed" program, it is possible that consumers' perceptions will diffuse rather than sharpen in realization that the offering has been modified to appeal to their particular preferences. It may be that this problem could be avoided without each performing arts organization specializing in attracting a single distinct market segment. In this case broader appeal would be sought through more extensive variations than programs alone. By presenting different types of offerings for different prices with different types of performers in different locations at different times of the day, week, or year, it may be possible to appeal to different consumer tastes without compromising or weakening the image of the organization among other segments. Whether this is in fact possible requires further research in light of the fact that multiple branding and multiple dealers do not appear to be viable strategies for maintaining distinct images in the performing arts. A further impediment may well be that the artistic personnel of the arts organization are not sufficiently numerous or flexible to be able to present such multiple programming without threatening self-perceived artistic integrity.

One final constraint on the present conclusions should be noted: the perspective presented is one of short-term audience broadening by a single arts organization in a local market that presently ignores segmentation in audience tastes. Other findings from a cross-sectional analysis of personal characteristics related to attendance among the present sample, indicate that childhood experiences with theatre and symphony are among the best predictors of present attendance. [For details see Andreasen and Belk, 1978.] This may well suggest that a long-run community strategy to broaden arts attendance should be to develop better school and non-school efforts to introduce children (or their families) to the performing arts throughout childhood. Perhaps in this way it will be possible to broaden consumer tastes rather than simply to respond to them.


Andreason, Alan R. and Belk, Russell W. (1978), Consumer Response to Arts Offerings: A Study of Theater and Symphony in Four Southern Cities, Washington, D.C.: National Endowment for the Arts, RWO-22-6N.

Andreasen, Alan R. and Belk, Russell W. (1978b), "Consumer Response to Arts Offerings", in Research in the Arts ed. David Cwi, Baltimore: Walters Art Gallery.

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

National Endowment for the Arts (1976), "Research Division Program Solicitation," February 4, 1976.

National Research Center of the Arts (1976), America and the Arts, New York: National Committee for Cultural Resources.