for the Alan R. Andreasen Research Division Report #24 National Endowment for the Arts Expanding the Audience For the Performing Arts Expanding the Audience For the Performing Arts Alan R. Andreasen Research Division Report # 24 National Endowment for the Arts Seven Locks Press Washington Expanding the Audience for the Performing Arts is Report #24 in a series on matters of interest to the arts community commissioned by the Re- search Division of the National Endowment for the Arts. Library of Congress Cataloging-in-Publication Data Andreasen, Alan R., 1934- Expanding the audience for the performing arts / by Alan R. Andreasen p. 53 cm. — (Research Division report; #24) Includes bibliographical references. 1SBN0-929765-01-X 1. Performing arts — United States — Audiences. 2. Performing arts — United States — Marketing. I. Tide. II. Series: Research Division report (National Endowment for the Arts, Research Division); 24. PN2226.A53 1991 791'.06 , 980973— dc20 90-23150 CIP Manufactured in the United States of America Typeset by Edington-Rand, Inc. Printed by McNaughton & Gunn, Saline, MI Seven Locks Press is a Washington-based book publisher of nonfiction works on social, political and cultural issues. It takes its name from a series of lift locks on the Chesapeake and Ohio Canal For more information or a catalog: Seven Locks Press P.O. Box 27 Cabin John, MD 20818 (301) 320-2130 Table of Contents INTRODUCTION 1 PART I: DEFINITIONS AND BASIC MODEL 3 PARTE: METHODOLOGY 7 Childhood S ocialization 1 Adult Socialization in the Arts 1 1 Life- style 12 General Leisure Activities 12 Work Status 13 Family Life Cycle 13 Other Socioeconomic Characteristics 14 Barriers to Attendance 14 PART III: RESULTS 16 Childhood Socialization 17 Adult Socialization 17 Life- style 18 Work and Television Viewing 23 Age and the Family Life Cycle 24 Socioeconomic and Background Characteristics 26 Barriers 28 Arts Attended and Arts Sought 30 Summary 32 Discriminant Analysis 34 PART IV: CONCLUSIONS AND RECOMMENDATIONS 37 Recommendations for Performing Arts Managers 39 Insights from the Performing Arts Adoption Process Model 40 Empirical Insights 40 Research Recommendations 45 NOTES 48 APPENDIX A: Candidate Variables for Discriminant Analysis 52 About the Author 53 v INTRODUCTION Organizations involved in producing or funding the performing arts have been heartened by evidence of the growing popularity of their offerings. This is reflected in steadily growing attendance at opera and symphony concerts and at legitimate theater, especially regional theater, since 1970. 1 Undoub- tedly, performing arts organizations have benefitted from a long-term im- provement in the socioeconomic status of the general population, since attendance at these types of events is closely correlated with income, educa- tion and occupation. 2 It is equally clear, however, that the recent growth in attendance can be at least partly attributed to carefully planned interventions by arts organiza- tions. These have been both indirect and direct. Indirect interventions have included the proliferation of programming of the arts on television, the increased availability of videocassettes of arts performances, formation of cable television arts networks, and dramatically increased quality of recorded music reproduction (e.g. compact disks). Direct interventions have included audience-building marketing efforts such as imaginative advertising, com- plicated and enticing subscription programs, ticket-selling booths in places such as shopping malls and railroad stations, telemarketing, informative preprogram lectures, and shirtsleeve performances. In the present political and economic environment, producers and sup- porters of the performing arts face rapidly mounting pressures to expand audiences for their performances. Since this is essentially a marketing problem, arts managers have increasingly turned to the marketing literature for help. Unfortunately, the manager is likely to find that, with very few exceptions, the literature is extremely sparse and often not very useful. Stated very simply, arts managers who wish to build an arts audience have at their disposal a number of marketing instruments that can be used at some level of cost to achieve audience development goals. These instruments include marketing's traditional "four Ps," product, price, promotion, and place of offering, plus other tools such as public relations. The manager typically would like help with four kinds of decisions: 1. How to segment the market into mutually exclusive target populations that can be the focus of one or more marketing interventions; 2. How much to allocate to marketing effort in each segment, including no effort; Alan R. Andreasen 3. On what to expend marketing effort (e.g. reducing price, increasing advertising, changing the offer); and 4. When to approach a segment with a particular strategy. The question of when to approach particular market segments has re- ceived inadequate attention in the marketing literature. It is possible, however that at any point, consumers will vary in their readiness-to-change with respect to any particular behavior, such as attendance at the performing arts. Further, different consumers at a given stage of "readiness," or at different stages of readiness, may vary in the kinds of interventions they will respond to. 3 If a marketer is to be effective and cost-efficient in using resources to develop markets, it is essential to learn more about which consumers are ready for which kind of intervention to change their behavior. The need to be more effective and cost-efficient is especially serious in the performing arts wheie budgets are extremely limited. Research that helps arts managers understand how and when to approach particular segments of the market would be valuable. If marketers are to change consumer behavior with respect to performing arts attendance, they must understand more about the process by which people come to attend performances. It is quite obvious that people do not become deeply involved in the performing arts overnight; it is a gradual process. Some become very active and committed, some only marginally active, and the majority not involved at all. Thus, when we speak of consumers' readiness-to-change, we are really talking about consumers' readiness to become more committed to attending the performing arts. Given this framing of the issue, the difficulty one faces in making recommendations for the future development of arts audiences is that we do not have any clear understanding, conceptually or empirically, of that process by which someone becomes a committed, involved arts attender. The present paper represents an attempt to provide a beginning under- standing of that process. It outlines a hierarchical model of the audience development process and then utilizes a set of secondary cross-sectional data to begin to understand the transitions a consumer makes from being un- interested to being highly committed to the arts. The paper describes con- sumers at various stages in this process, attempts to learn what seems most related to transitions between stages, and then makes recommendations for both managerial action and further research based on the model and the study's primary findings. Expanding the Audience for the Performing Arts I: DEFINITIONS AND BASIC MODEL Individuals do not become members of the audience for the performing arts by chance. Attendance is the result of a developmental process by which each individual progresses from a lack of awareness and interest through several stages of consistent, growing involvement, eventually becoming an active participant. Although participation in the arts may comprise many other things besides attendance — performing, or working backstage, enjoying the arts through other media such as books, radio, recordings, and television — here we are considering only attendance at live performances. Indeed, other kinds of participation may be complements or substitutes for live attendance. Each individual's progression towards deeper commitment to perform- ing arts attendance can be described in terms of two measures: interest and attendance. Further, it can be assumed that those who are now the most actively committed to arts attendance once upon a time were uninvolved and disinterested. For some, involvement and interest in the arts started very early in life. For others, it started later. And, of course, for the majority of Americans, interest and involvement has never grown or has grown very little. Both interest and attendance can be thought of as a hierarchy with a great many potential positions that can be occupied over a lifetime. There is a substantial body of literature in the social sciences that suggests that, in acquiring any new behavior, individuals pass through a number of definable stages. These steps are often referred to as the "adoption process." 4 Adapting this approach to the present case permits one to postulate// ve major transition points across six stages as individuals progress from lack of interest and involvement to active participation and interest in the performing arts. The six stages outlined in Figure 1 are given labels borrowed directly from the adoption literature. The transition points may be described as follows: 1. Changing from being a nonattender and disinterested to first becoming interested in attendance, that is, moving from Stage I, Disinterest, to Stage II, Interest; 2. Changing from being merely interested in the arts to attending a first arts event, that is, moving from Interest to Stage III, Trial; 3. Changing from attending a first arts event to being interested in further attendance, that is, moving from Trial to Stage IV, Positive Evaluation; 4. Changing from being merely interested in further attendance to attending Alan R. Andreasen DISINTEREST I. INTEREST I. TRIAL IV. POSITIVE EVALUATION i r V. ADOPTION r VI. CONFIRMATION Figure 1. The Performing Arts Adoption Process many events, that is, moving from Positive Evaluation to Stage V, Adoption; 5. Changing from attending many performing arts events to also being interested in attending many more, that is, moving from mere adoption to Stage VI, Confirmation. This adapted hierarchical model will be called the performing arts adoption process. The model is, of course, an idealized progression. It is not expected that individuals will progress neatly through all six stages to reach the end point. Because of changes in circumstances, individuals who have passed to a relatively advanced stage may regress to an earlier stage before moving on to the next stage. Some may move through most of the stages and then decide Expanding the Audience for the Performing Arts to reverse the process and permanently reduce their attendance. It is also possible that some individuals will skip some stages, or simply stop at some point and progress no farther. An opportunity to understand the performing arts adoption process is offered by the availability of data from the 1982 Survey of Public Participa- tion in the Arts (SPPA). 5 The study was prepared and administered to a nationwide sample of adult Americans by the Bureau of the Census. It contains information on involvement in various aspects of the arts, both directly and through the media, and detailed background information on each household, including standard demographics and several dimensions of household life- style. The present paper represents a new analysis of the SPPA data to answer four related questions: 1 . What characteristics describe individuals at each of the six stages of the performing arts adoption process? 2. What characteristics discriminate between those individuals occupying adjacent stages in the performing arts adoption process? 3. What do these discriminating variables explain about the process by which an individual moves from one stage to the next? 4. What do these findings recommend to those managers seeking to expand audiences for the performing arts? Implicit in the conceptualization underlying the present analysis are three theoretical assumptions: (a) Audience development is best facilitated by recognizing that target consumers are at different stages with respect to progression from complete disinterest to extensive involvement; (b) The performing arts adoption process is a reasonable specification of the stages and the sequence of progression; and (c) The most reasonable objective for audience development is to move each segment to the next stage rather than to move each segment directly to the stage of greatest involvement. The variables used to describe and explain progression through the performing arts adoption process are those measured or approximated by the data in the Survey of Public Participation in the Arts. The set of variables explored here is based on the notion that increased attendance is likely to occur if: Alan R. Andreasen 1. The individual has somehow learned to appreciate the performing arts; 2. The individual has a life-style that includes or is compatible with attending the performing arts; 3. The individual does not satisfy the need for entertainment entirely from other media; 4. The individual is not yet satiated with attendance; and/or 5. The individual does not perceive major barriers to increased attendance. Each major variable will, in turn, be affected by other factors. For example, learning to appreciate the arts can take place early in life, and this is presumably affected by whether one's parents were well educated and/or participated in the arts themselves. Further, whether an individual perceives certain barriers as important impediments to further attendance will be a function of such factors as income and marital and family status. Some of the variables to be explored here are at least partially controllable by arts managers; most are not. Some, such as eliminating specific barriers, are controllable in the short run. Others, such as introducing more people to the arts at an earlier age, are only "controllable" in the long run. The focus here is limited to attendance at six performing arts activities: jazz, classical music, opera, musical plays, ballet, and legitimate theater. We are explicitly not considering other arts attendance such as visits to museums. There are several reasons for making this distinction. The most basic reason is that these six live performing arts are assumed to be part of the same market: They involve being entertained by the performances of live human beings, they are almost never free, and they typically involve considerable preplanning and additional expenses for babysitting, dinner, transportation and the like. 6 Finally and most importantly, these art forms are usually seen as substitutes for each other. 7 6 Expanding the Audience for the Performing Arts II: METHODOLOGY The basic model for expanding the audience for the performing arts rests on a simple premise: becoming involved in the arts is not a one-step process, but is a progression through six stages called the "performing arts adoption process." Although the model appeals to common sense, it should not become the basis for policy or action unless it has been tested further with other databases and other methodologies. One could argue that the model conforms to the "reality" of this database. What one needs is a different approach to validate the "stages" and the basic notion of "progression." To test the model and explore its implications, data from the Survey of Public Participation in the Arts (SPPA) are used. The database consisted of responses from 2,678 persons who were surveyed during the months of November and December 1982. Two changes were made in this database. Potential target members who could not advance along the performing arts adoption process either because of ill health or a handicap were excluded from the analysis. Excluding these 69 respondents and two others for whom there was incomplete data brought the final sample size to 2,607. Another group of 370 respondents could not advance along the hierarchy because a specific art form was unavailable where they lived. These in- dividuals were retained in the sample, but their responses about the unavail- able art form are excluded. This revised sample is assumed to be representative of the U.S. adult population. Although the responses contain random "noise" due to question- naire design, administration, and tabulation, there is no systematic bias evident. The SPPA surveyed individuals at one point in time at a specific stage in the performing arts adoption process. Assuming that these individuals have moved there from the preceding stage allows us to explain transitions by assuming that those in an earlier stage will move to the next stage //"they adopt certain characteristics that distinguish them from those in the next group. As noted earlier, it is recognized that this is a major assumption and that it is very likely that some members in each of the six categories may have "moved" there from several stages away or from a later rather than an earlier stage. A true test of the model must await the existence of tracking data on the same households at several points over time. The first task in the present analysis was to categorize respondents into the six stages of the performing arts adoption process. After excluding cases Alan R. Andreasen where art forms were unavailable, respondents were categorized into the six stages as follows: I. DISINTEREST: Has not attended any of the six performing arts forms in the last twelve months and is not interested in attending; II. INTEREST: Has not attended any of the six performing arts forms in the last twelve months but is interested in attending; IE. TRIAL: Has attended one performance of one or more of the six performing arts forms in the last twelve months and is not interested in attending more often; IV. POSITIVE EVALUATION: Has attended one performance of one or more of the six performing arts forms in the last twelve months and is interested in attending more often; V. ADOPTION: Has attended two or more performances of one or more of the six performing arts forms in the last twelve months but is not interested in attending more often; VI. CONFIRMATION: Has attended two or more performances of one or more of the six performing arts forms in the last twelve months and is interested in attending more often. Since the six performing arts are assumed to be substitutes for each other, increased attendance can be measured both as more attendance at the same performing arts form or as attendance at several different arts forms. This assumption is consistent with initial analyses of this database by Robinson et al., where they found high correlations among the six performing arts forms. It is also assumed that the lack of interest in attending more often on the part of individuals assigned to the Trial stage is temporary and not evidence of a negative trial experience. Table 1 shows the distribution of the entire sample across the six stages. Approximately one half of the sample is in Stage I, Disinterest. Many of those at this stage, probably the majority, are unlikely to progress to the next stage because those in Stage I are very much unlike those in any other stage. Table 1 indicates that a further one-fifth of the sample has only progressed to Stage II, Interest. Thus, 70 percent of the sample was found to be at the very beginning stages of the performing arts adoption process. While this may seem discouraging at first glance, the fact that 22 percent of the sample is at Stage II suggests a significant untapped potential among nonattenders. The individuals who comprise the remaining 30 percent of the sample have attended at least one event in the past twelve months. One-half of them 8 Table 1 Sample Distribution Across Stages of the Performing Arts Adoption Model Stage in Process Number Percent I. Disinterest 1,225 47.0% II. Interest 579 22.2% III. Trial 172 6.6% IV. Positive evaluation 307 11.8% V. Adoption 98 3.8% VI. Confirmation 226 8.7% Total 2,607 100.0% have only attended a single event while the others have already attended two or more in the last year. Before moving to an analysis of each stage and the transitions between them, one must ask whether the data offer general empirical support for the validity of the model, the performing arts adoption process. The model has face validity in that it describes a logical progression based on common sense, the experience of those in the field, and conceptualizations and findings of those in other social science fields. But does the model predict relationships that do in fact appear in the data? Three sets of associations are offered as predictive support. In the literature on performing arts attendance and in the initial analysis of both the 1982 and 1985 SPPA, researchers have found that participation in the performing arts is positively associated with (a) education, (b) participation in a wide range of other activities (referred to as the "more-more" phe- nomenon) and (c) consumption of the arts through other media. 8 If the present model meets the test of predictive validity, there should be systematic increases in levels of education, general activity and arts media consumption measures from one stage to the next. Table 2 shows that the model meets this test of predictive validity. Table 2 reports data on average years of education and the percent attending college for individuals at each stage of the process. It also records two measures of activity: the average number of different out-of-home activities in which those at each stage participated, and the average number of total activities including at-home activities, hobbies and crafts, and working backstage on Table 2 Education, Leisure Activity, and Media Participation by Stage in Model Mean Number Mean Percentage of Art Activities Years of Attending Stage in Process Education College In-home Out-of-home Via media 1. Disinterest 11.0 18.8% 2.88 3.22 .52 II. Interest 127 39.7% 4.10 5.29 .94 III. Trial 13.0 47.7% 4.31 6.17 1.42 IV. Positive evaluation 13.6 57.0% 4.55 6.62 1.55 V. Adoption 14.2 65.3% 4.63 7.02 2.29 VI. Confirmation 14.7 ' 74.3% 4.86 8.00 2.65 the arts (see Table 5 for listing of out-of-home and in-home activities). Finally, it shows the average number of other media used to appreciate the arts. In each case, the progression from Stage I to Stage VI is systematic, linear and continuous, although the transitions between each stage are not always statistically significant. In the present analysis it is assumed that an individual's present stage in the performing arts adoption process is a function of seven sets of variables: childhood socialization, adult socialization, life-style, work status, family life cycle, barriers to participation, and other socioeconomic characteristics such as race, sex, and income. Childhood Socialization The stage of a sample member at the time of the study is strongly influenced by his or her early exposure to the arts. Several studies have found that this is, indeed, an important predictor of attendance at the performing arts. 9 This exposure could come about in one of two ways. First, individuals could have parents who encouraged them to take an interest in attending the arts. In the SPPA, two questions explored this factor. First, respondents were asked whether, when they were growing up, parents or other adult members of the household listened to classical music or opera, or took the respondent to plays, dance, or classical music performances. Respondents were asked whether this happened never, occasionally or often. For the present analysis, an "early encouragement index" was constructed by weighting answers of 10 Expanding the Audience for the Performing Arts "often" twice as much as answers of "occasionally." Thus, this index could range from zero to four. Second, individuals could have been encouraged by their parents to attend classes in an art form. In the SPPA, respondents were asked whether they took lessons or classes in four categories: music (either voice training or playing an instrument), acting or theater, ballet, or music appreciation. These questions were asked for four age ranges: under 12, 12 to 17, 18 to 24 and 25 or older. A "childhood arts education index" was constructed by counting how often an individual reported having had classes in each of the four categories in each of the two age groupings under 18. Thus, an indi- vidual's "childhood arts education index" score could range from zero to eight. In addition to these two indexes, we shall also observe the proportion of respondents at each stage who were both encouraged by adults to participate in the arts and given classes as a child. Adult Socialization in the Arts It is also possible that individuals will develop or extend an interest in the arts as adults. Many individuals obtain their first exposure to the perform- ing arts in college either through direct exposure or through the influence of fellow students. To assess this possibility a dichotomous variable for atten- dance at college was created, i.e. whether a given respondent did or did not attend college. There are two other possibilities for adult socialization. First, adults could take the same kinds of classes they took (or could have taken) as a child. Second, they could be exposed to the performing arts on radio, television, or recordings. To permit analysis of these factors, an "adult arts education index" was constructed in the same manner as the childhood arts education index discussed earlier, with values ranging from zero to four. In addition, a dichotomous measure of whether a respondent had any adult education in the arts was also constructed. For the three electronic media, separate indexes were constructed for exposure to arts such as jazz, classical music, opera, musical stage plays or operettas, nonmusical stage plays, and ballet on television (range of the index is zero to six), radio (range of the index is zero to five) and records or tapes (range of the index is zero to four). The problem with adult classes and media exposure, however, is that they may be an effect of attendance at live performances rather than a cause. 11 Alan R. Andreasen Further, it may be that exposure in other media could be a substitute for live attendance. Life-style General Leisure Activities Interest and participation in the arts do not occur in isolation. It is part of an overall "life-style" which, in the marketing literature, is typically meas- ured by an individual's activities, interests, and opinions. These are some- times called AIO measures. 10 Bureau of Census interviewers took measures on a number of life-style dimensions. One set of 14 questions asked about general leisure activities. Respondents were asked whether in the previous twelve months they had: 1. Attended movies, sports events, zoos or gardens, or amusement parks; 2. Pursued hobbies such as games, collecting, preparing special meals or gardening; 3. Engaged in physical activities such as exercising or jogging, playing sports, or camping; 4. Read books, magazines, or novels; 5. Did volunteer work; and/or 6. Worked on home improvements or vehicle repairs. In addition, the SPPA asked respondents to indicate how many hours they watched television (presumably in their leisure time) "on an average day." A second set of questions asked about activities of a "cultural" nature other than attendance at the six types of art forms that are the major focus of this analysis. Respondents were asked whether they had: 1. Visited an art museum, a science museum, a historic site, or an arts and crafts fair; 2. Read or listened to poetry; 3. Undertaken arts and crafts activities; 4. Worked behind the scenes at performing- arts performances; and/or 5. Engaged in creative writing. Because it is essential to understand whether particular individual ac- tivities show up as important to consumers presently at specific stages of the 12 Expanding the Audience for the Performing Arts process, each activity is analyzed separately. 11 Understanding the activities that are more often associated with a particular stage may help to round out our understanding of the people who are currently there. Once this analysis of individual activities is completed, we will turn our attention to general summary indexes of the activities based principally on logic and preliminary analyses of the data. Work Status A second important dimension of one's life-style is work status; that is, whether a respondent works and, if so, in what type of occupation and for how many hours a week. It is particularly important to know how many hours an individual works, because this will indicate the amount of leisure time available for activities such as attending the performing arts. Further, certain occupations might encourage or "require" such attendance. It is assumed that professional or managerial occupations fit this category. The present analysis, therefore, explicitly includes each respondent's total num- ber of working hours and whether the respondent was in a managerial or professional occupation. Family Life Cycle An individual's life-style is inextricably linked to family status. Engaging in certain leisure activities (or avoiding them) will often be determined by whether one is young and single, has children living at home, or is elderly and retired. One study has noted: From the point of view of consumption, there are several other factors that vary systematically as the individual progresses through such a life cycle. Besides age and income, the individual's needs and tastes change. Responsibilities for other family members change with the size of the family and the self-sufficiency of its members, and there are systematic changes in accumulated ex- perience, accumulated and desired durable goods, and accumu- lated savings and other liquid assets. Finally, when children and spouse are present, their needs, preferences and abilities are also changing as is the pattern of interaction for the family as a whole. 12 A substantial number of studies in the marketing literature have con- firmed this point. 13 13 Alan R. Andreasen In the present analysis, individuals were categorized into six life-cycle groupings based on age, marital status, and whether they have children and the ages of children as follows (percentages of total sample indicated in parentheses): I. Young and single. Divorced or never married and under 35 years of age (11.8%). II. Young, married and no children (6.5%). HI. Young children at home. Married, single, or divorced with one or more children under 6 years of age (18.5%). IV. Older children. Married, single or divorced with children, none of whom is under 6 years of age (21.4%). V. Older, no children. Age between 35 and 64 and no children (26.5%). VI. Elderly. Age 65 and older (14.9%). . Other Socioeconomic Characteristics There are a number of other socioeconomic characteristics in addition to education, occupation and family life cycle that may be related to arts attendance. 14 Among them are income, race, place of residence (specifically, whether the individual resides in a Standard Metropolitan Statistical Area or not), household size, and gender. Several of these may be related to a respondent's stage in the performing arts adoption process. However, an earlier study by Andreasen and Belk found that, while socioeconomic characteristics were related to performing arts attendance, this relationship was mainly due to their also being correlated with such factors as life-style and early socializa- tion in the arts, although the latter could also be related to socioeconomic status. In a multivariate analysis, life-style and early socialization were shown to be much more important determinants of attendance. 15 Barriers to Attendance Individuals in Stages n, IV and VI of the performing arts adoption process expressed interest in attending the arts more often, but are not doing so. In such cases, SPPA interviewers asked their reasons for not attending. In addition to illness or unavailability of an art form, a great many other reasons were indicated by respondents. Five predominated and will be used in the present analysis. These are: cost, lack of time, problems associated 14 Expanding the Audience for the Performing Arts with travel ("too far to go" or "transportation/traffic/parking problems"), personal barriers ("feel uncomfortable," "don't have anyone to go with," "babysitter problems/must care for children," or "crime or fear of crime") and lack of motivation. 16 A Note on Statistical Significance Classical statistics recommends stringent criteria when reporting something as "significant." Typically, this means that something should be ignored if the chance of a particular finding really being non-significant is greater than five percent. Several of the findings reported here meet this criterion. Since this is an exploratory study, we have adopted less stringent tests for reporting some of the data, namely that the chance of being wrong is no greater than ten percent. Since there is a very small base of research in this area that is useful to marketers, we decided to let the data reveal as much suggestive material as possible by allowing a 10% chance of error. In general, the narrative does not distinguish between findings that meet the stricter test; however, the tables clearly report significance levels for those readers who wish to reinterpret the findings using narrower tolerances. 15 Alan R. Andreasen III. RESULTS The data on the characteristics of respondents at different stages of the performing arts adoption process reveal one dramatic, persistent pattern that affects both our conclusions and the manner in which the analyses proceed. As seen in Tables 3 through 9, respondents in Stage I, those disinterested and not attending the arts, are dramatically different from those in the other five stages. For example, they are much less likely to have been exposed to the arts as children or to have attended college. They are significantly less likely to be exposed to the performing arts in other media such as radio, television, or tapes and records, to take classes in the arts, or to engage in related crafts and, in point of fact, they are less likely to engage in any of the activities measured in this study. It is very clear that the distance between those at Stage I and those in Stages II through VI of the performing arts adoption process is not just significant. It represents a chasm. The size of this gulf implies that those in Stage I are the least ready of any of the six groups to change their involvement Table 3 Selected Measures of Childhood Socialization in the Arts by Stage in Model I II III IV V VI Dis- Positive Childhood Socialization Measure interested Interested Trial Evaluation Adoption Confirmation Early encouragement index .41 .80 f .69 1.06 f 1.46 f 1.42 Child arts education index .47 .89 T .98 1.15 1.20 1.69 f Total socialization index .88 1.69 f 1.67 2.21 f 2.66 3.1 1 + Any parental encouragement 29.1% 48.0%* 44.8% 57.3% 65.3% 69.9% Any classes as child 34.8% 62.7% r 62.8% 67.1% 72.4% 77.9% Parental encouragement j. AND classes 25.4% 34.9% T 36.6% 44.0% 53.1% 59.3% Parental encouragement X OR classes 47.6% 75.8% T 70.9% 80.8% 84.7% 89.8% Attended college 18.8% 39.7% f 47.7% 57.0% 65.3% 74.3% Mean years of education 11.0 12.7 f 13.0 13.6 14.2 1 " 14.7 "^Significantly different from the preceding stage at the .05 level of significance or better. 16 Expanding the Audience for the Performing Arts in the arts. Attempts to move consumers across the chasm are likely to be very expensive and not very successful. Because those in Stage I represent a market segment that from an efficiency standpoint should be of lesser interest to performing arts managers, our focus in the analyses to follow is primarily on the remaining five stages. This is not to say that some managers might not justifiably choose, from a cultural equity standpoint, to pursue efforts to reach those in Stage I, only that it would be less efficient and is not the focus of this paper. Childhood Socialization It was hypothesized that an adult's present status in the performing arts adoption process will be strongly influenced by socialization in the arts as a child. Table 3 indicates two important patterns with respect to the last five stages of the performing arts adoption process. First, there is a clear increase in the amount of childhood socialization between people at Stage II and those at Stage VI, but childhood socialization apparently becomes important only when one moves beyond the trial stage. Early childhood socialization appears to have the important effect of producing individuals who quickly become interested in the performing arts, moving quickly to an interest in attending two or more events (Stage IV) or actually doing so (Stages V and VI). Adult Socialization Table 4 reports data on the extent to which stages in the performing arts adoption process are associated with a college education and being involved in the performing arts in some other form, either through taking classes or through exposure in other media. It shows that all of the measures increase as one moves forward in the performing arts adoption process. It is reasonable to conclude that a college education speeds one along the process. However, with respect to adult classes and consumption through other media, it is not possible to conclude from cross-sectional data whether the positive relationships represent cause, effect, or simultaneity. With respect to consuming the arts through other media, the data do permit us to conclude that this is not a substitute for attendance at live performances. Table 4 also indicates that consumption through the media increases dramati- cally at Stage IV, the point at which consumers positively evaluate the performing arts. Thus it seems reasonable to speculate that increased media consumption may be a part of a pattern of generally increased interest in the performing arts rather than a cause or effect. 17 Table 4: Selected Measures of Adult Socialization in the Performing Arts by Stage in Model I II III IV V VI Dis- Positive Adult Socialization Measure interested Interested Trial Evaluation Adoption Confirmation Adult arts education index .11 .31* .41 .49 .73* .92 Percent taking classes as adult 13.7% 36.3%* 38.4% 45.9% 54.1% 72.6%* Number of art activities via media: radio .18 .59* .54 .68 .99* 1.26 records/tapes .18 .57* .59 .89* 1.07 1.58* television .43 1.28* 1.38 1.89* 2.34* 2.68 Total, all media .79 2.44* 2.51 3.46* 4.40* 5.53* "•"Significantly different from the preceding stage at the .05 level of significance or better. The data on adult classes show a different but understandable pattern. Consumption does not dramatically increase until consumers are actually attending multiple events rather than just being interested in doing so (i.e., Stage V). Taking classes, like attending performances, requires a commit- ment of time and effort. Increasing both at the same point would seem perfectly consistent. On the other hand, increasing one's interest (Stage IV) is a relatively passive step forward, not requiring a change in behavior. Furthermore, one may expect to find significant increases in exposure to the arts via the media at this stage because that, too, requires very little effort and commitment. This is, indeed, the case. Life-style The data in Table 5 report the proportion of those at each of the six stages of the performing arts adoption process engaging in 30 different activities within the last twelve months. Table 6 then summarizes these data into five indexes and an overall "general activity" measure for the sample. These groupings are based on logical considerations and preliminary analyses of the data rather than factor analyses. 17 18 Table 5: Percentage Engaged in Leisure Activities by Stage in Model Activity Dis- interested Interested III Trial IV V VI Number Positive Reporting Evaluation Adoption Confirmation Activity Out-of-home Individual Go to movies 48.3% Go to sports 36.6% Camp/hike 29.8% Play sports 27.5% Exercise, jog 37.8% Do charity work 18.2% Out-of-home Family Visit art museum 6.7% Visit science museum 1 0.4% Visit historic site 21.7% Visit art fair 24.4% Visit amusement park 40.0% Visit zoo 21.3% 36.8% 29.1% t 46.8%* j t 58.2%* 79.1% 71.3% 52.3%' 57.6% t 44.2% t 46.5%' 55.2% t 39.6%* 20.4%* 35.5%* 24.7%* 29.1% 43.5% f 52.6%* 39.4% 54.7% f 62.8% 39.2% 82.6% 64.8% 47.9% 51.1% 57.8%* 62.8% 71.7%* 38.4% 35.2% 34.2% 60.2% 60.7% 64.6% 49.3%* 78.1% 67.4% 53.1% 55.2% 66.7% 60.4%* 53.1%* 42.7% 58.3% 65.6% 59.4% 45.8% 86.2% 75.0% 51.8% 59.4% 78.6% 58.9% 64.6% 49.1% 75.0%* 77.2%* 56.7% 61.2%* t 1660 1279 966 1042 1361 765 567 574 1014 1084 1314 886 Creative Do crafts Do needlecrafts Paint, sculpt Write poems, stories Take art class Do photography Play musical instrument Act, sing, dance In-home Play games Read books, magazines Read novels, poetry Collect stamps, coins Cook gourmet meals Repair home, vehicle Garden Read, listen to poetry Backstage Music, play, ballet Jazz, classical performance 8.8% 27.4% 5.7% 2.3% 4.8% 5.7% 3.3% 3.3% 51.5% 70.6% 38.8% 10.0% 18.8% 50.0% 49.2% 9.0% .7% 0.0% 15.9%* 38.2%* 12.3% f 6.4%* 12.8%* 10.2%* 2.4%* 4.5%* 74.4% 90.5% 64.1% 16.6% 31.4% 69.7% 63.4% 1.7%* .7%* 20.3% 41.3% 16.3% 9.3% 11.0% 14.5% 2.3% 5.2% 76.6% 94.8% 68.4% 22.7% 32.6% 65.5% 71.5% t $ 21.6% + 24.4% 4.7%* 1.2% 14.4% 47.2% 11.1% 9.2% 16.4% 16.4% 13.1% 3.6% 83.6%* 98.0%* 78.2%* 20.3% 40.8%* 67.2% 70.1% 32.1%* 4.9% 1.6% 17.7% 38.5% 21.9% 17.7% 25.0% 19.8% 13.3% 13.3% 77.1% 95.8% 78.6% 27.1% 50.0% 77.1% 65.6% 46.9% t 6.3% 4.2% 14.7% 40.6% 21.0% 21.9% 21.0% 24.1% 9.3% 10.6% 81.3% 97.3% 88.0% 23.7% 53.6% 73.2% 72.8% 50.4% 10.3% 3.1% t 326 890 270 175 272 276 107 124 1702 2156 1473 398 760 1562 1530 533 70 22 "^Significantly different from the preceding stage at the .05 better; ^Significantly different from the preceding stage at the or better. level of significance or .10 level of significance 19 Table 6 Indexes of Leisure Activities by Stage in Model i II III IV V VI Dis- Positive Activity interested Interested Trial Evaluation Adoption Confirmation Out-of-home, individual 1.98 2.94 f 3.38 1 3.54 3.72 4.06 Out-of-home, family 1.24 2.33 f 2.73 t 3.02 3.19 3.81 f Creative .61 1.03 T 1.20 1.22 1.64 1 1.61 In-home 2.38 3.40 1 " 3.66 f 3.89 f 3.88 4.13 Backstage .01 .02 f .06 f .07 .10 .13 All activities 6.69 10.32 1 " 11.62 f 12.23 13.06 14.15* "^Significantly different from the preceding stage at the .05 level of significance or better. ^Significantly different from the preceding stage at the .10 level of significance or better. Tables 5 and 6 share the pattern found in earlier findings in that the level of activity in all categories is higher at Stage II than at Stage I. On the other hand, the pattern of differences across the other five stages varies consider- ably across the 30 activity categories reported in Table 5. In three cases, there are no clear conclusions that can be drawn because there are too few participants in Stages II through VI to permit statistically valid conclusions: working backstage at jazz or classical performances, playing a musical instrument, and acting, singing or dancing. In the remain- ing 27 categories, three patterns appear: 1. In seven cases there are some statistically significant differences in selected transitions, but the overall pattern appears to be one where there are not major differences across the five stages. These seven activities are: visiting amusement parks, doing crafts, doing needlecrafts, playing games at home, reading books or magazines, repairing home or vehicles, and gardening. 20 Expanding the Audience for the Performing Arts Six of the seven activities are usually carried out around the home, and in five cases they are done alone. 2. In three cases there is a significant increase between Stage II and Stage VI. However, the increase appears to be gradual. These three categories are: • going to sporting events, • visiting science museums, and • doing photography. 3. In the remaining 16 categories there are increases between Stage II and Stage VI. However, the stage at which respondents demonstrate a marked increase in commitment to the activity varies considerably. There are ten cases in which there is a significant increase in activity at the Trial Stage (Stage III). These are: going to the movies, camping and hiking, playing sports, doing charity work, visiting art museums, visiting historic sites, visiting art fairs, collecting stamps and coins, gardening, and working backstage at musicals, plays, opera or the ballet. It should be noted that eight out of these ten categories involve getting out of the home and doing something active. It would appear that moving to the trial stage is strongly associated with a willingness to spend leisure time out of the home. Since those in Stage III have attended only one event in an arts category in the previous year and have not expressed an interest in greater attendance, the performing arts may play a relatively minor role in these activities. Those in Stage III are outgoing, and one of the things they do when going out is attend the performing arts. There are seven cases where activity increases at Stage IV (Positive Evaluation), the stage at which consumers have tried the arts and are interested in attending more. These activities are: • exercising and jogging, • visiting zoos, 21 Alan R. Andreasen reading or listening to poetry, playing games at home, reading books or magazines, reading novels or poetry, and cooking gourmet meals. Five of these activities are clearly home-centered, and exercising/jog- ging may also be done at home. This raises the possibility that those who have tried the arts and want to attend more often are held back because their home-centered life-styles have "forced" them to. The most obvious reason for this would be the presence of children at home. Six activities appear to increase at Stage V (Adoption), the point at which consumers begin to go to the performing arts relatively frequently but express no interest in going more often. The six activities are: doing charity work, painting or sculpting, writing poems or stories, taking arts classes, reading or listening to poetry, and repairing home or vehicles. Four of these activities reflect a commitment to arts activities done alone. It may be that those who have reached this stage have, in effect, fashioned a life-style that encompasses a multifaceted active arts life of which attendance at multiple events is simply one part. At Stage VI (Confirmation), five activities significantly increase. These are: visiting art museums, visiting historic sites, visiting art fairs, visiting zoos, and reading novels and poetry. The first four of these are family-oriented activities. This last stage apparently includes two kinds of people: those who are generally excited about the arts and those whose expressed desire to attend more often is simply a reflection of that excitement, and those whose desire to attend more may be an expression of frustration that other things in their lives (e.g. family commitments) may be keeping them from doing so. If this 22 Expanding the Audience for the Performing Arts is the case, only the former can be tapped by the arts community for increased attendance. The summary indexes of activities reported in Table 6 clearly illustrate the above patterns. The Trial stage reflects increased activity more or less across the board. The stage of Positive Evaluation shows an increase in in-home activities; Adoption, an increase in creative activities; and Confir- mation, an increase in out-of-home, family activities. These broader indexes appear to add little to the richness provided by the individual-level analysis. Work and Television Viewing It might be hypothesized that work and television viewing would compete with attendance at the performing arts. The more one works, the less time one would have for any kind of leisure activity, including attending the performing arts. In addition, the more one watched television, the less one is likely to be found at the performing arts, not because television viewing keeps one from the arts but because it may be preferred to attending live performances. Data in Table 7 indicate that neither hypothesis is consistently true. There is a significant difference between respondents in Stage I and Stage II in whether they work. However, contrary to the hypothesis, those interested in attending the performing arts are more likely to be working than those who are not. Further, among those who do work, those at Stage I work fewer hours than those at Stage II. As we have seen throughout these data, those not Table 7 Employment Characteristics by Stage In Model Employment Characteristic i Dis- interested II Interested III Trial IV Positive Evaluation V Adoption VI Confirmation Percent employed Hours worked per week Percent in professional/ managerial occupations TV hours per day 52.4% 38.0 10.2% 3.80 66.5% f 39.8 f 22.6% f 3.18 64.5% 36.6 f 24.4% 3.21 67.7% 36.2 29.6% 3.11 65.3% 37.8 32.7% 4.11 67.7% 39.1 42.5% 3.17 "^Significantly different from the preceding stage at the .05 level of significance or better. 23 Alan R. Andreasen attending and disinterested in the arts have markedly different life-styles in both their leisure and work pursuits. It may be fair to say that their most overwhelming trait is that they are simply less active in most dimensions of their lives. They do less, including working and attending the arts. Encourag- ing them to attend would not be merely a matter of changing the kinds of activities in which they engage, but of changing their overall level of activity. There is a decline in work hours between Stages II and III that would be consistent with the hypothesis that those attending are more likely to do so because they have the time. On the other hand, it may be noted that by Stage VI, the average number of work-hours of those who do work has approached the levels of those in Stage II. Table 7 also shows that as one moves through the performing arts adoption process, consumers are increasingly likely to be in managerial or professional occupations. With respect to watching television, Table 7 reveals no significant differences across groups. Those in the disinterested group (Stage I) do watch about 20 percent more television than four of the five remaining groups. Only the Stage V group exceeds their level of viewing. Age and the Family Life Cycle Data in Table 8 on mean age for each stage of the performing arts adoption process show that those with the least interest in the arts are the oldest. One striking age pattern in the data is the finding that those contented with their present status (Stages I, III, and V) are older than those expressing interest in attending more (Stages II, IV, and VI). This suggests that those who are older (even when they are current attenders) may not be particularly good prospects for programs intended to increase attendance. Of course, as social scientists have noted for some time, age may not best describe an individual's progress through life. A richer conception, the family life cycle, 18 incorporates age with marital status and the presence or absence of children to yield a set of "normal" life stages. The six stages and their relationship to the performing arts adoption process are indicated in Table 8. The results confirm the speculation that those over 65 (married or living alone) are much less likely to be interested in attending the performing arts more often. Does the presence of children make a difference? Three competing hypotheses may be offered: 24 Table 8 Stage in the Family Life Cycle by Stage in Performing Arts Adoption Model I II III IV V VI Dis- Positive Characteristic interested Interested Trial Evaluation Adoption Confirmation Mean age 46.0 38.5 f 43.1 40.8 44.8 f 41.2 Family Life Cycle Young single 9.0% 14.3% 15.1% 10.7% 17.3% 16.8% Young, married, no children 5.1% 8.6% f 4.1% f 10.4% f 4.1% j 6.2% Infants at home 17.6% 24.5% f 12.8% f 19.5% 21.2% 13.3% Children 6 plus 20.7% 21.9% 23.8% 21.8% 17.3% 23.0% Older, no children 26.6% 22.5% f 27.9% t 26.4% 31.6% 33.2% Elderly 20.8% 7.4%* 1 4.5% t 10.4% 17.3%* 7.5% Not categorized .2% .8% 1.8% .8% .2% 0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Mean Household size 3.04 3.11* 2.90 2.95 2.64 2.72 "^Significantly different from the preceding stage at the .05 level of significance or better. ^Significantly different from the preceding stage at the .10 level of significance or better. • The presence of children decreases attendance but not interest be- cause children impose increased costs if they are taken along or if babysitters are needed or because scheduling a time to actually get away to live performances is more difficult; • The presence of children decreases arts attendance and interest because children cause a shift in priorities (i.e. more attention devoted to the work of childrearing) or because the children impose their own values on how the family spends its leisure time; • The presence of children increases arts attendance because adults interested in the arts make extra efforts to expose their children (perhaps as they were) to the performing arts. The data in Table 8 suggest that both the first and second hypotheses may have merit. First, it may be noted that those who are young and single or just married with no children are much more likely not to be in the disinterested group. Two thirds of both groups have progressed beyond Stage I to develop 25 Alan R. Andreasen at least some interest in the arts. Further, married individuals are more likely to have attended an arts performance and expressed an interest in attending more (Stage IV). With the appearance of young children, there is a substantial increase in those reporting themselves "disinterested." (This is not a function of aging since the two groups with children at home are not defined as either young or older.) With children under six years at home, we find that only one-quarter of the respondents have attended even one art performance, although 29.5 percent would like to. Of the respondents with older children, 31.8 percent have not only expressed interest, but have actually attended a performance. This trend continues to the point where the respondents are older and the children are gone. Here we see that 34 percent have attended in the last year. More significantly perhaps, the proportion having attended several times rises simultaneously. The effect of the life cycle is perhaps best seen by looking at the proportions at each life-cycle stage who are at Stages V or VI: Young, single 17.9% Young, married, no children 10.7% Children under six at home 8.7% Children six or older 12.4% Older, no children , 15.4% Elderly 8.8% Here we can see again that the elderly are clearly less active. But during the rest of the family life cycle, the relationship appears to be curvilinear. Those who are either young and single or older and with no children are those most likely to attend multiple events. Next most likely are those married with no children or married with children over six. Least likely to attend multiple events are those with children under six. The conclusion that young children at home inhibit rather than motivate involvement in the arts seems highly plausible. Finally, the data on average household size indicates relatively little difference across the various stages, with the slightly larger household size in Stage II consistent with the explanation above. Socioeconomic and Background Characteristics We have already analyzed several important socioeconomic charac- teristics including age, occupation, education and the family life cycle. There are several additional measures available in the Survey of Public Participa- 26 Expanding the Audience for the Performing Arts tion in the Arts that, according to other hypotheses, may be related to attendance and interest. These are race, ethnicity, gender, residence, and income. As indicated in Table 9, we do not find any statistically significant effects across the stages of the performing arts adoption process related to race and ethnicity. In part, this is a function of the small sample sizes. 19 Gender, however, exhibits a curious pattern. Unlike most other variables in this study, there is no difference between the disinterested/non-attending group (Stage I) and the rest of the sample. Further, as Table 9 indicates, there are no differences in the proportion of females in Stages I, II, HI, and V. However, the table does indicate that those in Stages IV and VI are sig- nificantly more likely to be female. Both of these stages represent cases where respondents want to attend more performances but perceive barriers to doing so. As we shall outline below, there are important differences in the kinds of barriers men and women perceive to attending more arts performances. Residence. Just as gender affects whether one is interested in attending the performing arts more, so does whether one lives in a Standard Metropolitan Statistical Area (SMSA). 20 As noted in Table 9, residence in an SMSA affects interest in attending more performances at all three stages, II, IV, and VI. The most obvious explanation suggested by this finding is that a "penal- ty" of living in an SMSA for those who have developed some interest in Table 9 Selected Socioeconomic Characteristics by Stage in Model Characteristic I Dis- interested II Interested III Trial IV Positive Evaluation V Adoption VI Confirmation Female Black Hispanic In SMSA Income over Mean family $30,000 income 52.3% 11.0% 6.1% 56.6% 16.5% $21,364 53.7% 9.2% 5.0% 76.3% f 33.2% f $27,134 + 52.9% 4.7% 3.5% 59.3% 30.8% $26,457 62.5% f 8.1% 3.6% 83.1%* 40.1% f $30,1 22 f 49.0% t 6.1% 4.1% 56.1 % f 38.8% $30,660 64.6% f 8.0% 1.8% 85.8%* 41.2% $31,502 "•"Significantly different from the preceding stage at the .05 level of significance or better. ^Significantly different from the preceding stage at the .10 level of significance or better. 27 Alan R. Andreasen attending the performing arts is that one is exposed to a great deal of opportunity. Those already attending either one (Stage III) or many (Stage V) arts events are as likely to be from an SMSA as are those who profess interest in the arts but did not attend in the previous year. It appears that living in an SMSA has a perverse effect: It does not make one any more likely to be an active arts attender, but it probably makes one much more likely to want to attend more. Income. It has often been suggested that the performing arts are the play- ground for the well-to-do. We have already noted (Table 4) that education increases systematically through the stages of the performing arts adoption process. We have also seen in Table 7 that there is a slow increase in the proportion of respondents in managerial or professional occupations up through Stage VI. The effects of income, however, are different. When we observe either mean family income or the proportion of households with income over $30,000, we see two points of change in Table 9. As expected, the disinter- ested (Stage I) have significantly lower incomes than those with any involve- ment with the performing arts. There is little difference between Stages II and III, but a jump occurs as one moves from merely trying the performing arts to being interested in becoming a more active arts consumer. This would suggest that a necessary precondition for moving to the stage where one wishes to attend multiple performances may be a reasonable level of discre- tionary income. The increasing importance of income occurs at the positive evaluation stage, and after that stage there is little difference in average income or in the proportion of those who have incomes over $30,000. Barriers To this point we have investigated primarily personal or family attributes associated with progress from stage to stage of the performing arts adoption process. Because of the nature of the database, we have had to infer the likely factors that may be causing or inhibiting progress. However, one question in the Survey of Public Participation in the Arts directly addresses this issue. At three of the stages of the performing arts adoption process, individuals reported an interest in moving on to the next stage (i.e. attending more). In each case they were asked what barriers kept them from moving forward. Their own explanations were categorized in the SPPA under 16 headings, three of which we have excluded from this analysis: art form unavailable, respondent handicapped, or respondent ill or infirm. Barriers reported in five 28 Table 10 Percentage Citing Barriers to Attending More Performing Arts by Gender, Residence and Family Life Cycle Barrier Cited: Characteristic Cost Travel Time Personal Motivation Non-SMSA SMSA 33.9% f 41.5% 36.7% f 17.7% 36.7% f 49.2% 18.1% 22.6% 11.3% 15.3% Male Female 38.2% 41.3% 19.0% 23.3% 54.2% t 41.3% 11.0% f 29.3% 16.4% 13.1% Young, no children Children at home Older, no children 38.0% f 47.9% 31.2% 20.4% 20.1% 24.1% 54.4% 47.9% 37.0% 10.4% f 27.6% 22.0% 16.0% 12.1% 16.4% Significantly different from the preceding stage better. at the .05 level of s ignificance or key categories representing 88 percent of all barriers mentioned are reported in Table 10. The five categories are cost, problems associated with travel (i.e. too far to go or transportation, traffic, or parking problems), time, other personal reasons (i.e. "would feel uncomfortable," "don't have anyone to go with," babysitter problem/must care for children, or crime or fear of crime), and lack of motivation. Analysis of these perceived barriers revealed very little. Whether one has not attended in the last year, attended once, or attended more than once has no relationship to whether one mentions travel, time, other personal reasons, or lack of motivation for not going more. In only one case does a difference appear. Those who did not attend during the previous year were less likely to mention cost as a reason. The most plausible explanation for this may be that those who have not attended simply do not have much knowledge of costs at this stage. The finding would certainly imply that arts managers seeking to attract those who have not recently attended may not have to worry as much about addressing the cost problem. Although only one difference in barriers is mentioned across stages in the performing arts adoption process, there are differences related to resi- dence, gender and family life-cycle. The relevant data are presented in Table 10, which shows patterns that are quite plausible. Those in SMS As are much 29 Alan R. Andreasen more likely to mention cost and less likely to see travel as a problem. They also mention not having enough time, which would be consistent with our speculation that life in the "big city" is rife with opportunity for an active life-style, of which the performing arts are but one part. With respect to gender, men are somewhat more likely to mention not having enough time, while women are significantly more likely to mention personal reasons for not attending. When the latter is broken down into its components, we find that women are somewhat more likely to cite fear of crime, babysitter problems, or lack of someone to go with when asked identical questions about barriers to attendance. The findings with respect to the family life cycle are intriguing. Those with children are much more likely to mention cost, presumably because going to performances represents a much greater total outlay either for multiple tickets or for babysitting costs. With respect to time, the major difference is that the elderly less often complain about not having enough time. Those who are younger with no children probably have time problems because of an active life-style, while those with children may feel more family time pressures. The finding of differences due to personal concerns masks diversity among the family life-cycle groups. Those who are young without children rarely mention these problems. They seldom lack dates, they don't need babysitters, and they don't worry about crime or feeling uncomfortable. Those with children, especially those with children under six, mention babysitting problems, while those who are older and have no children are much more likely to mention fear of crime or the fact that they have no one to go with. Several of these findings support our earlier speculation that arts par- ticipation is curvilinear with respect to the family life cycle because the presence of children imposes both financial burdens and other respon- sibilities on these households. Arts Attended and Arts Sought To this point we have been treating the six categories of performing arts as interchangeable. Examining the particular art form currently attended and those forms sought out for future attendance provides further insight into the characteristics of those at each stage of the performing arts adoption process. Relevant data on current attendance are presented in Tables 11 and 12. Table 11 reports the proportion of those in Stages III through VI attending each performing art. Table 12 then reports attendance in terms of what 30 Table 11 Percentage Attending Performing Arts by Stage in Model in IV Positive V VI Performing Art Trial Evaluation Adoption Confirmation Jazz 27.3% 20.5% 46.9% 43.8% Classical music 18.6% 25.4% 50.0% 54.4% Opera 2.3% 2.0% 14.3% 10.2% Musical plays 47.7% 55.4% 64.3% 72.1% Plays 25.6% 26.1% 40.8% 52.2% Ballet 2.3% 4.9% 17.3% 24.8% Total events attended (number) (213) (412) (229) (582) marketers call "market share." The latter controls for the fact that those in Stages V and VI attend many events while those in Stages HI and IV attend only one or two. Table 12, therefore, reports the percentage each category represents of all categories mentioned by the group in the stage. Thus, for example, Table 1 1 shows that 27.3 percent of those in the Trial Stage (Stage III) attended jazz performances, while Table 12 shows that jazz events represent 22.1 percent of all events attended by those at this stage. 21 As expected, those who attend multiple events show increases in absolute levels of attendance in all six categories. Obviously, those in Stages V and VI are better targets for each of the performing arts than those in Stages III and IV. However, Table 12 shows a subtle shifting of relative emphasis among categories. There is little difference across the four stages for attend- ance at classical music performances. Those at the trial stage are relatively more likely to patronize jazz or plays. By contrast, those attending multiple events (Stages V and VI) are relatively more likely to attend ballet and opera. Table 13 confirms two patterns among those who wish to attend more often. Those at the "interest" stage mention many fewer categories (although they could have mentioned all six) than those attending multiple events. This clearly confirms the assumption that movement along the performing arts adoption process represents deepening involvement in the arts. On the other hand, Table 13 does not show equal increase in interest in all categories. There is virtually no deepening of interest in jazz attendance across the three categories. There is an increase in the other five categories from those who didn't attend last year (Stage II) and those who attend only one event (Stage IV). However, of those who have attended several events, there is no change 31 Table 12 "Market Share" for Each Performing Art by Stage in Model hi IV Positive V VI Performing Art Trial Evaluation Adoption Confirmation Jazz 22.1% 15.3% 20.1% 17.0% Classical music 15.0% 18.9% 21.4% 21.1% Opera 1.9% 1.5% 6.1% 4.0% Musical plays 38.5% 41.3% 27.5% 28.0% Plays 20.7% 19.4% 17.5% 20.3% Ballet 1.9% 3.6% 7.4% 9.6% Total 100.1% 100.0% 100.0% 100.0% Total events attended (213) (412) (229) (582) in interest in attending more stage plays, and a decline of interest in attending musical plays. Greater interest, however, is shown for the more "serious" art forms (classical music, opera, and ballet). In general, the patterns in Tables 11,12 and 13 show a correlation between increased attendance and increased interest in more serious art, a finding that validates the basic model. Summary A review of the analyses of individual variables and indexes in the preceding sections indicates that in every case except one there were sig- nificant differences between those individuals not attending and not inter- ested in the arts (Stage I) and those who have a beginning interest in the arts (Stage II). The one exception is being in a family with children over six years old. The latter is one of three measures where there are no major changes across all six stages. The other two are race and ethnicity, where the lack of significant effects across all six stages is probably due to small sample sizes. If one looks only at the four transitions beyond the first two stages, there are six variables where there are virtually no differences across the remaining stages. These measures are all life-style activities: • visiting amusement parks, • doing crafts, • sewing and knitting, 32 Table 13 Performing Arts Sought by Those Wanting to Attend More by Stage in Model ii IV Positive VI Performing Art Interest Evaluation Confirmation Jazz 31.3% 29.0% 29.6% Classical music 24.4% 31.9% f 42.9% Opera 10.9% 12.1% 18.1% Musical plays 51.3% 71.7% + 56.6% f Plays 36.6% 50.2% T 50.0% Ballet 14.0% 19.2% f 28.3% t "^Significantly different from the preceding stage at the .05 level of significance or better. • working backstage at a jazz or classical music performance, • acting, singing, or dancing, and • playing a musical instrument. The lack of differences for the last two measures is in part due to small sample sizes. Of the remaining measures, eight show no significant differences be- tween any two stages, but do show a systematic increase stage-by-stage over the five stages. These measures are: number of years of education, having attended college, being a professional or manager, total consumption of the arts through other media, going to sports events, visiting science museums, and doing photography. Many of the remaining variables also rise as one moves from Stage II to Stage VI. However, there are occasions where there are spurts or lags at certain stages that create significant differences between stages. This generally describes all of the remaining life-style measures and most of the other socialization, media, and socioeconomic measures. There are three cases, 33 Aian R. Andreasen however, in which the patterns involve oscillations. These are: (a) SMSA, where those interested in greater attendance (Stages n, IV and VI) are more often found in SMS As than those who are content with their present level of attendance (Stages I, HI, and V); (b) gender, where females are more likely to be interested in attending more multiple events (Stages IV and VI) and men are more content with their level of attendance (Stages HI and V); and (c) the elderly who were more likely to be content with their present levels of attendance (Stages HI and V) than wanting more (Stages IV and VI). Discriminant Analysis To this point, our analysis has identified and provided prehminary validation for the model of the performing arts adoption process, uncovered a number of dramatic differences across stages, and identified factors that appear to be related to transitions between the stages. Except for the oc- casional use of indexes the analysis has considered each of the major sets of variables in the SPPA study separately. However, many of the variables are related to each other. This would include both traditional socioeconomic correlations such as occupation and education, and also many of the life- style measurements specific to this study. Many of the factors shown to be significant may only be so because they are associated with other variables. It is important, therefore, to investigate the proposed performing arts adoption process model considering the variables simultaneously. Since our primary interest is in understanding movement through the performing arts adoption process, the questions we need to answer are the following: 1. At each stage of the process, which set of variables in combination best distinguishes those at this stage from those at the immediately preceding stage? 2. What is the relative importance of each of the variables in the final set? 3. How successful is the entire set of variables in separating those at one stage from those at the earlier stage? The appropriate technique for this task is the two-group discriminant analysis which has had a relatively long history in social science and marketing research. The approach uses the technique analytically to find the linear combination of candidate discriminator variables that best separates a 34 Expanding the Audience for the Performing Arts given population into two groups where group membership is already known. A typical application in marketing would be to identify the key distinguishing characteristics of those who buy a firm's brand or patronize a firm's outlet or service from those who do not. 22 It has also been used to separate buyers and nonbuyers of new products. 23 Where there are more than two groups involved, it is also possible to extend two-group discriminant analysis to a multiple-group case. 24 In the present analysis, discriminant analysis is employed to learn which combination of variables best separates respondents at each transition point in the performing arts adoption process.* As expected from the results of other studies, several of the variables found to be significant in the bivariate analyses (analyses using only two variables) were not significant in the discriminant analysis when their inter- actions with other variables were considered. Still others showed reversals of signs suggesting that real relationships may be obscured in the bivariate analyses by the presence of hidden third or fourth variables. In the five discriminant equations, a number of patterns emerged. First, the single most important discriminating variable is place of residence, which appeared in all five equations and typically carried the heaviest weight. It is clear that living in a major metropolitan area creates a strong interest in attending more performing arts events, even among those who haven't attended in the past year. Second, life-style factors appear in all of the equations. And, while the specific factors differ by equation, they are almost always life-style activities that involve considerable effort and activity and take place outside the home. Interpretation of each of the last four equations in light of the variables also found significant in the bivariate analysis is as follows: • Trial. With other variables held constant, moving to the trial stage is not associated with parental encouragement in the arts as a child, but is positively associated with attending college. Those moving to the trial stage are likely to be older, not recently married or not with children under six. Having children under six clearly inhibits actual attendance, even at a first event. Those who move to the trial stage and show no interest in moving farther are likely to engage in other activities such as visiting museums, hiking or camping, or going to * The full set of predictor variables is listed in Appendix A. In general, the candidate set of variables was as disaggregated as possible. Thus, for example, the set includes the individual life-style activities rather than summary indexes. 35 Alan R. Andreasen the movies. They also balance their arts attendance with working backstage at plays and painting and sculpting on their own. Positive Evaluation. Developing an interest in increased arts atten- dance after a trial appears to be a matter of both motivation and inhibition. Parental encouragement and living in an SMS A lead one to want more attendance, and having an income over $30,000 ap- parently helps provide the means for getting more deeply involved. However, being newly married or having children under six appar- ently discourages attendance at multiple events (Stage IV). As a consequence, this group listens to records and tapes and visits his- torical sites, presumably with young children. Home life apparently permits less painting, sculpture, or crafts, and encourages board games. People at this stage also keep up their interest in the arts by attending adult arts classes. Adoption. Moving on to attend multiple events is again associated with parental encouragement and with going to art museums and painting and sculpting, but not with visiting historical sites or exer- cising or jogging. Adoption is more likely to be associated with charitable work and not associated with being female. Females are more likely to want to move beyond present attendance levels, but are inhibited by other obligations. 25 Movement to the adoption stage is also associated with a greater interest in arts on the radio and less interest in arts on records or tapes. Confirmation. Movement beyond multiple attendance to wanting more involvement is again a function of motivators and inhibitors. Childhood socialization is again important, in this case in the form of early arts classes in school rather than parental encouragement. As in the previous transition, life-cycle factors are no longer significant (although we know that the elderly in general are not interested in increased arts activity). Movement to the confirmation stage is as- sociated with a decline in attending art classes and an increase (again) in visiting historical sites and exercising or jogging. Of more sig- nificance is the fact that women are more likely than men, and SMS A residents more likely than non-SMSA residents, to have a long term commitment to the arts. 36 Expanding the Audience for the Performing Arts IV: CONCLUSIONS AND RECOMMENDATIONS The preceding bivariate and multivariate analyses constitute an ex- ploratory investigation of a new model outlining six steps through which an individual progresses from being a disinterested non-attender at the perform- ing arts to being an attender at multiple events who wants to attend more. This model is responsive to the needs of performing arts managers who wish to expand their audiences, but who recognize that active intervention must be based on a clear understanding of the underlying process to be influenced. Audiences differ in "where they are" with respect to attending the performing arts, and marketers must realize that it is an incremental process. The research reported above appears to offer strong support for the model both as a way of describing the process of becoming involved in arts patronage and as a source of insights into the factors that encourage or inhibit movement through the process. The model of the performing arts adoption process not only has face validity, but also predictive and explanatory power. The bivariate analysis indicates that movement along the stages is clearly associated with factors such as education, occupation, life-style, and early childhood socialization, found to be important in other studies. Movement from one stage to another is characterized by more involvement and by deeper involvement in the sense of greater interest and more attendance. The strongest differences in the entire process are between the first two stages. Those who express no interest in the arts are dramatically different from those who have even the most minimal interest. More importantly, these differences are so pronounced that people at Stage I should, from an effi- ciency standpoint, be given very low priority in future audience development programs. Their socioeconomic characteristics are just not like those at any other stage in the process. Furthermore, they are simply less active in all pursuits, including work. Thus encouraging their interest in the arts would be not only a matter of changing their activities, but changing their inactive life-style. This would seem a formidable, if not an impossible task. This study found that movement from Stage II to Stage VI is probably influenced by the following socioeconomic variables: • Early childhood socialization whether through parental encourage- ment or art classes is strongly associated with movement along the performing arts adoption process. 26 However, the discriminant analy- sis suggests that the influence of this variable may not come about until after the adoption stage. 37 Alan R. Andreasen Residence in a Standard Metropolitan Statistical Area presents more opportunities to participate in the arts and presumably leads to a stronger desire to attend more performing arts no matter what one's present level of attendance. At the same time, while an SMS A reduces travel problems as a barrier to greater attendance, it increases eco- nomic costs and apparently offers so many other opportunities (and perhaps a more hectic life-style) that SMS A residents complain more often of not having enough time to attend more. One's stage in the performing arts adoption process is associated with one's stage in the family life cycle. The elderly are less likely to attend multiple events and more content with their present level of attendance, whatever it is. Attendance at multiple events is greatest by those who are young and single and those who are older and have no children living at home. It seems probable that the presence of children under six strongly inhibits movement through the stages. According to the discriminant analysis, and contrary to other studies, discretionary income is not uniformly associated with movement along the process. It sets those interested (Stage II) apart from those who are not (Stage I), probably because it is associated with at least a minimal interest in the arts. Beyond that point, income only appears to be an important discriminator at Stage IV. One apparently has to have at least a minimum amount of discretionary income to be interested in attending multiple events. As other studies have shown, measures of life- style provide useful insights into movement across various stages of the process. How- ever, in contrast to the earlier studies, we found that, once other factors are controlled, only those life-style activities involving active commitments of time and effort away from home appear to offer consistent explanatory power: People who seldom leave home during their leisure time are unlikely to increase attendance at live perfor- mances. The multivariate analysis shows that these activities appear most strongly at the Trial stage where a passive interest in the arts is first associated with active effort to actually get out of the house and go to a performance. At the stages of the process where individuals may perceive barriers 38 Expanding the Audience for the Performing Arts to increased attendance, there were three major findings. First, with one exception, we found no major differences in barriers whether one has never attended, attended one or two times, or attended multiple events. The one exception was that cost was less important to those who had not attended in the last year. Second, cost seems less an inhibitor for those early in the process than one might have assumed. Finally, we found that the types of barriers one mentioned were related to SMSA (as noted above), gender, and family life cycle, although in ways that were quite predictable (e.g. those with young children complained about babysitting problems). • Other uses of time such as working or watching television appeared not to inhibit arts attendance, nor did consuming arts on other media reduce attendance. On the contrary, consumption of art through other media appeared to become important when respondents had gone beyond the trial stage, had positively evaluated a live performance, and wanted to attend more. Finally, with respect to methodology, the bivariate analyses identified some variables that appeared to be associated with movement along the process, but were subsequently found not to be significant or to have a different effect when other factors with which they were correlated were introduced. Clearly, bivariate analyses should be approached with caution if they are not accompanied by the application of more powerful techniques such as discriminate analysis. Recommendations for Performing Arts Managers We have provided strong support in the preceding sections for the performing arts adoption process as a valid descriptive and explanatory model of the sequence of steps through which consumers pass in then- progress from disinterest to active involvement in the live performing arts. The model must still be considered exploratory, but it seems reasonable to argue that it is useful to both theorists and researchers. A final test for the model is whether it is useful to performing arts managers. This audience is, in fact, the one whose development problems motivated the present study. We propose the following as reasonable managerial insights based on the model itself and on the empirical investigation of target segments of the population defined by that model. 39 Alan R. Andreasen Insights from the Performing Arts Adoption Process Model • Becoming a committed, performing arts attender requires progres- sion through several stages of commitment and behavior. Transform- ing a presently disinterested nonattender into an enthusiastic patron of the arts is a gradual, one-step-at-a-time process. • Transition from step to step varies in the challenge it presents. Getting interested for the first time presumably requires acquisition of much more information than is the case for those who tried the performing arts and need to be brought to the point where they attend more. Clearly, the marketing strategies that a manager uses for those at one stage must be different from those used at a different stage. • Since the transitions facing a consumer at each of the first five stages of the model differ, it is reasonable to expect that the type of consumer one finds at each stage will differ (a major empirical finding of the analysis). This has further implications for the marketing strategies to be employed at each stage. Empirical Insights As noted, the results clearly indicate that consumers at each of the five stages of the model differ markedly in who they are and what their interests and backgrounds are. In this study, it was difficult to judge whether people at some stages were more ready to change than others — a major concern for managers. However, the empirical evidence clearly suggests that the gap between those at Stage I (Disinterest) and at Stage II (Interest) is more than a gap; it is a chasm. It seems quite clear that efforts to move the 47 percent of the market at Stage I to the next stage would be uneconomical and probably would have a very limited chance of success. Ways to influence the 53 percent of the population at the other five stages in the process are suggested by empirical data which describe those at each stage and the nature of the transitions between them. This allows managers to develop an effective segmentation strategy to address each stage separately. One approach is direct targeting. It involves clearly identifying those with traits that make them very likely to be prime candidates for making a particular transition and then focusing one's resources on them. The second is self-selection. It involves contriving messages that speak directly to the particular segment and inducing them to think "Oh, yes, they're speaking to 40 Expanding the Audience for the Performing Arts me. Let me hear what they have to say." Self- selection is more expensive than direct targeting in that it requires blanketing the market rather broadly in the hopes of catching many in the targeted group. It is also risky in that many who are not in the targeted group will be reached by the message and some may, in fact, be offended by it. To develop a basis for using either strategy, one must identify the traits that distinguish consumers at each transition point, and then emphasize the characteristics of those at the next stage to which one wishes those in the earlier stage to move. The assumption is that those in the earlier stage possessing these traits are the best candidates for moving on. The data, therefore, suggest the following segmentation approaches for each of the four major transitions: Inducing Trial 1. Direct Targeting. Prime candidates would be those who: a. Are not on an arts organization's subscription list but have indicated some interest {e.g. asking for a brochure, taking a guided tour, or stopping by an arts fair exhibit); b. Have worked backstage at a play or musical comedy; c. Have visited an art museum (i.e. are on their mailing list); d. Belong to some organized sports team or gym; e. Hike or camp (e.g. belong to a hiking club or visit specific recreational sites); f. Attend movies (perhaps of a particular type?); g. Are in an SMS A; h. Are alumni of colleges; i. Paint or sculpt (e.g. show at arts fairs, take classes, rent studios); j. (Possibly) Are involved in charities; k. (Possibly) Elderly (e.g. belong to Golden Years organizations). 2. Self-Selection. Messages directed to this segment should portray them as having one or more of the traits just outlined. Further, they should not be portrayed as young and married with no children or with children under six. Messages that might "speak" to this audience could contain some or all of the following language: • "You've always been interested in the arts. Now that the kids are in school and your life is more settled, here is your chance ..." 41 Alan R. Andreasen • "There are certain people who are always on the go. They hike, they enjoy sports, they go to art museums. They're curious and they like new experiences. For those of you who have always wanted to try the performing arts as part of your active life-style, here's your chance . . ." Encouraging Positive Evaluation 1. Direct Targeting. The obvious target groups would be those who: a. Are first- time attenders at any event; b. Purchase records or tapes of arts performances; c. Visit historic sites or zoos; d. Take art classes; e. Are in SMS As. 2. Self-Selection. Messages appealing to this group might contain some or all of the following: • "This is often a child-centered time for many households. Times for taking the kids to a zoo or a historic site like . Or sometimes just to stay home and play board games and listen to your tapes or records. But remember the time when you went to the performing arts? Don't you think that's something you ought to try again — maybe after the kids are a little older? Or maybe it's something to take the kids to now, just as your Mom and Dad took you. Now that you've got a little more money, don't you think you owe it to yourself — and to them?" • "We know you spend a lot of time at home, but you're still a very active person. Don't you think you should enjoy the live performing arts as a regular part of your active life-style?" At this stage, your objective is to get the target audience interested in attending more, not in actually doing so. Encouraging Adoption 1. Direct Targeting. The groups to appeal to here are those who: 42 Expanding the Audience for the Performing Arts a. Have attended a performing arts event recently and expressed an interest in attending more, perhaps by picking up a brochure at the site; b. Visited art museums; c. Are active in charitable work; d. Listen to some classical music radio stations; e. Are Black; f . Paint or sculpt; g. Are male. 2. Self-Selection. Messages for this segment may be designed around the following themes: • "For a long time you've been thinking about making the performing arts a permanent part of your active life- style. Here is your chance to subscribe to a package of great events." • "Your parents were right. Making the performing arts a major part of your life is rewarding. Even for the busy person who thinks he or she can't fit in another thing, the arts are a great diversion — something really different! They can stimulate, enrich, relax the most jaded and overworked person." • "We know how exciting it is to live in a big city and to have so many choices. Maybe you're a little frustrated that there is so much out there that you can't do it all and just don't know where to begin. Here's our guide for those who want to 'Grow with the Arts'." Inducing Confirmation 1. Direct Targeting. The targets here are those who: a. Are already subscribers to arts series or who can be found on multiple mailing or subscriber lists; b. Visit historical sites; c. Participate in exercise classes or health clubs; d. Are female; e. Are in SMSAs. 43 Alan R. Andreasen 2. Self-Selection. Messages for this group would either reinforce their existing attendance {e.g. subscription) behavior or encourage them to further broaden their experiences. A typical message might be: • "As an active person and a serious patron of the arts, you know what an enriching experience the arts can be. It is something you will want to keep up for a lifetime. Remember, our city contains many resources in the performing arts. Here's a way you can continue to expand your involvement in the arts . . ." Although the data do not permit estimates of how much to allocate to each of these four transitions, two considerations should be kept in mind. First, the segments differ in size. Almost two-thirds of those beyond Stage I are in Stages II and IV where they have expressed positive interest in increased attendance. Second, the segments beyond Stage I probably differ in their susceptibility to efforts to move them to the next stage. It would seem logical to suppose that it is easier to get those in (the larger) Stages II and IV to translate their interest into action than to change those in Stages in and V who are satisfied with their level of attendance. If more resources are put into Stages II and IV, further segmentation of those markets in terms of the barriers they face seems warranted. Thus, in addressing women, one should encourage them to come alone or should provide babysitting. In addressing those with children at home, one should stress cost savings by offering a family discount package. Although the data do not suggest it, an approach to moving people at one stage to the next could be to identify those at the later stage and have them motivate their friends who are at an earlier stage. For example, one might contact subscribers and encourage them to interest a friend, perhaps by offering a discount coupon (or subscription brochure) and a $10.00 discount at a restaurant good only for a party of four. Finally, the role of early childhood socialization in the later stages of the process should be noted. This again emphasizes the value to the arts com- munity generally of continued, substantial investments in programs in schools and elsewhere to encourage children to begin a lifetime involvement with the performing arts. These strategies allow managers to focus on targeted markets at particular stages and to inch them toward an extended commitment to attendance (and perhaps to other forms of financial and vocal support). In carrying out these strategies, managers need to develop systems to identify those at each stage 44 Expanding the Audience for the Performing Arts and then to track their progression to subsequent stages over time. Such systems would allow knowledge to accrue on how to stimulate movement in particular markets and how the analysis here applies to different regions and art forms. Research Recommendations Although we believe the approach of this explanatory study is both valid and useful, much remains to be done to extend the validation, improve our understanding of the transitions, and increase the model's managerial use- fulness. The following are some of the research possibilities to explore in the near term. 1. Clearly, the two most problematic aspects of the present study are that it is cross-sectional and based on secondary data. Several assumptions were necessary, particularly those assigning individual respondents to each stage of the performing arts adoption process. A replication of the present study with more careful questioning about respondents' entire past performing arts behavior would be a valuable next step. 2. Two alternative research approaches that could be used to measure changes in arts behavior year-to-year for the same consumers would be the following: • A retrospective study would simply ask a cross- section of consumers about their current behavior and their behavior a year earlier. While such a design would be subject to memory and telescoping biases, gross measures (e.g. asking if they are now attending "more, less, or the same") can be trustworthy. • A true panel would ask a fixed sample of individuals to report their current behavior at specified intervals (e.g. once a year). This design requires patience and can be expensive. It is also subject to biases due to sample "mortality" and pre-sensitization. However, it does yield relatively objective data on changes in behavior which can be tied to specific individuals. Further, costs can be kept within bounds by using lower-cost methods such as mail-back diaries after the first wave. 45 Alan R. Andreasen 3. If and when a new study is undertaken, measures of the following would add important new insights: • Attendance and interest in the performing arts on the part of other members of the household; • Satisfaction with recent attendance (especially among those who have attended only one event in a category in the past year); • Factors considered when deciding to attend a performance (to move from Stage II to Stage III) or to expand attendance to multiple events (to move from Stage IV to Stage V); • Magazines or newspapers read, radio stations listened to, television programs watched (to give guidance to future media strategies); • Recent changes in life status such as divorce, job change, geographic relocation, or birth of a child (to indicate whether such changes precipitate changes in performing arts involvement). 4. Other important assumptions or issues to be explored in future research are the following: Are the six art forms really substitutes for each other? Are there identifiable subsets that compete only with each other? Are there other going-out options that ought to be included as alternatives in future research designs (e.g. going to the movies or to an upscale restaurant)? • In what ways are metropolitan areas different? Is the substantial array of alternatives there motivating or frustrating? Are the life-styles of arts attenders in major urban areas markedly different from the life-styles of arts attenders in non-SMS As? Are there regional differences in the contribution of explanatory variables? Is cost more important in the South or Midwest? Is time pressure a greater problem on the East Coast? Are young people different in California? • Do individuals pass through all stages of the model, or do they leap 46 Expanding the Audience for the Performing Arts across stages under certain circumstances? Under what conditions does this happen? Under what circumstances do consumers "regress" in the process? To what extent are family life-cycle factors important in this regard? What can arts organizations do to inhibit or prevent such regression? 47 Alan R. Andreasen NOTES 1. U.S. Bureau of the Census, Statistical Abstract of the United States 1986 (Washington, D.C.: U.S. Government Printing Office, 1987), p. 236. Note that attendance at Broadway and regional theater declined in 1983 and 1984 after growing more or less steadily since 1976. 2. Paul DiMaggio, Michael Useem, and Paula Brown, Audience Studies of the Performing Arts and Museums. National Endowment for the Arts, Research Report #9 (Washington, D.C., 1977). 3. Alan R. Andreasen, "Life Status Changes and Changes in Consumer Preferences and Satisfaction," Journal of Consumer Research 11: 1 (Spring 1984), pp. 784-794. 4. Everett M. Rogers, Diffusion of Innovations (Glencoe, IL: The Free Press, 1983). This is the seminal work in the area. Also, Thomas S. Robertson, "Marketing's Potential Contribution to Consumer Behavior Research: The Case of Diffusion Theory," in Thomas C. Kinnear, ed. Advances in Consumer Research, Vol. XI (Provo, Utah: Association for Consumer Research, 1984), pp. 484-489. 5. John P. Robinson, Carol A. Keegan, Terry Hanford, and Timothy A. Triplett, Public Participation in the Arts: Final Report on the 1982 Survey (Washington, D.C.: Research Division, National Endowment for the Arts, October 1985). Data from a similar 1985 study were not used in the present analysis since in the later study all of the measures used in this analysis were never collected at one time for one sample of respon- dents. 6. As will be noted below, these assumptions deserve attention in future explorations of the present model. 7. This assumption is well documented in the 1982 and 1985 SPPA studies. See Robinson, et ai, Public Participation in the Arts. 8. See Robinson et ai, Public Participation in the Arts. 9. See, for example, Alan R. Andreasen and Russell W. Belk, "Predictors of Attendance at the Performing Arts," Journal of Consumer Research 7: 2 (September 1980), pp. 1 12-120; Richard J. Orend, Socialization and Participation in the Arts. National Endowment for the Arts, Research Report #21 (Washington, D.C., 1989). 10. William D. Wells, ed., Life Style and Psycho graphics (Chicago: Ameri- can Marketing Association, 1973). The usual strategy in marketing research is to factor-analyze such measures in one of two ways: Either one seeks underlying dimensions that the sample as a whole appears to express in its answers, or one seeks to group individuals in terms of their 48 Expanding the Audience for the Performing Arts AIO patterns. Both approaches were used by Belk and Andreasen in a 1978 study for the National Endowment for the Arts (Andreasen and Belk, "Predictors of Attendance at the Performing Arts"). These authors analyzed leisure life-style data for a sample of respondents in four southern cities and classified six basic types, two of which — the culture patrons and social activists — were important consumers for symphony or theater. Psychographic analysis using the widely popular VALS life-style model developed by Arnold Mitchell (Arnold Mitchell, The Nine American Lifestyles [New York: MacMillan, 1983]) was used to help explain performing arts attendance in a study carried out for the Association of College, University and Community Arts Administrators. Inc. (The Professional Performing Arts: Attendance Patterns, Preferen- ces and Motives, [Madison, WL: Association of College, University and Community Arts Administrators, Inc., 1984]). This study found that those who were classified as "societally conscious" or "experientials" were by far the heaviest attenders in the total population. Although the "achiever" psychographic group bulks larger in the general population, the study predicts that the "societally conscious" will soon overtake them in total audience size. In the present study, it was not possible to use the psychographic segmentation approach developed in either of the other two studies. 11. Robinson et ai, Public Participation in the Arts. Robinson and his associates analyzed these data (with the exclusion of TV viewing and museum attendance) and drew the following major conclusions: a) Five factors appeared in the analysis. However, four of these factors had relatively small associations with key variables. b) All of the activities were positively associated with each other and that "one 'general activity' factor seemed a more apt descriptor of the data than the five dimensions that emerged from the analysis." Because of these conclusions, we decided not to begin with a reduced set of factors. A second reason for adopting this strategy is that, in the present investigation, our interest is in the stages of the performing arts adoption process. It is not clear that, even if some general factor stru dure for the entire sample could be developed, such a structure would be appropriate to the separate stages and transition points. 12. Russell W. Belk and Alan R. Andreasen, "The Effects of Family Life Cycle on Arts Patronage," Journal of Cultural Economics, Vol.6, no. 2 (December 1982), pp. 25-26. Belk and Andreasen found that attendance at symphony and theater in four southern cities declined when in- dividuals were newly married and had children at home. 49 Alan R. Andreasen 13. William D. Wells and George Gubar, "The Life Cycle Concept in Marketing Research," Journal of Marketing Research 3 (November, 1966), pp. 355-363; E. Laird Landon and William B. Locander, "Family Life Cycle and Leisure Behavior Research," in William L. Wilkie, ed., Advances in Consumer Research 6 (Ann Arbor, MI: Association for Consumer Research, 1979), pp. 133-138. 14. DiMaggio and Useem, for example, point to the persistence of income and gender as predictors of arts attendance in a large number of studies. Paul DiMaggio, Michael Useem, and Paula Brown, Audience Studies of the Performing Arts and Museums: A Critical Review (Washington, D.C.: National Endowment for the Arts, 1978). 15. Andreasen and Belk, "The Effects of Family Life Cycle." 16. For a full description of barriers reported by respondents, see Robinson et al., Public Participation in the Arts. 17. However, they are roughly parallel to those used by Robinson et al., Public Participation in the Arts. 18. Wells and Gubar, "Life Cycle Concept." 19. Paul DiMaggio and Francis Ostrower, Race, Ethnicity and Participation in the Arts: Patterns, of Participation by Black, Hispanic and White Americans in Selected Activities from the 1982 and 1985 Surveys of Public Participation in the Arts. Report to the National Endowment for the Arts, Research Division, June 1987. Using a different approach and more information from the SPPA database, DiMaggio and Ostrower have concluded that there are differences in consumption due to race and ethnicity (i.e. Hispanic origin), but that these differences are largely attributable to differences between these groups and the white majority in other socioeconomic characteristics such as education. 20. Standard Metropolitan Statistical Areas are metropolitan areas with populations of 250,000 or more. They include a population nucleus and adjacent communities which have a high degree of economic and social integration with the nucleus. After 1983, SMSAs were redefined and named as Metropolitan Statistical Areas (MSAs). 21. This calculation does not weight attendance in a category by the number of times a respondent attended in the category. 22. See, for example, Franklin B. Evans, "Psychological and Objective Factors in the Prediction of Brand Choice: Ford Versus Chevrolet," Journal of Business, October 1959, pp. 340-369; Henry J. Claycamp, "Characteristics of Owners of Thrift Deposits in Commercial Banks and Savings and Loan Associations," Journal of Marketing Research, May 1965, pp. 163-170. 50 Expanding the Audience for the Performing Arts 23. Edgar A. Pessemier, Philip C. Burger, and Douglas J. Tigert, "Can New Product Buyers be Identified?" Journal of Marketing Research, Novem- ber 1967, pp. 349-369. 24. See, for example, Thomas S. Robertson and John N. Kennedy, "Predic- tion of Consumer Innovators: Application of Multiple Discriminant Analysis," Journal of 'Marketing Research, February 1968, pp. 64—69. 25. Race in this equation could not be interpreted. 26. See Andreasen and Belk, "Predictors of Attendance," and Orend, So- cialization and Participation. 51 Alan R. Andreasen APPENDIX A Candidate Variables for Discriminant Analysis Life-style Characteristics Outdoor, Individual Go to movies Go to sports Camp/hike Play sports Exercise, jog Do charity work Out of Home, Family Visit art museum Visit science museum Visit historic site Visit art fairs Visit amusement parks Visit zoo Creative Do crafts Do needlecrafts Paint, sculpt Write poems, stories Take arts class Do photography Play musical instrument Act, sing, dance In-Home Play games Read books, magazines Read novels, poetry Collect stamps, coins Cook gourmet meals Repair home, vehicles Garden Read, listen to poetry Backstage Musical, play, ballet Jazz, classical performance Family Life Cycle Young single Young, married, no child Children under six at home Children six or older Older, no child Elderly Arts on Other Media Television Radio Records, tape Employment Employed Number of hours worked Professional/managerial Other Socioeconomic Characteristics Age Years of education Household size Total family income Income over $30,000 Female InSMSA Black Hispanic Attended college Other Hours of TV viewing Parental encouragement as child Childhood classes in the arts Adult classes in the arts 52 Expanding the Audience for the Performing Arts About the Author Alan R. Andreasen is professor and head of the Department of Marketing at the University of Connecticut. He is an internationally known consultant and author on marketing for non-profit organizations. Other Publications of Interest Readers of this report may wish to obtain more information about the details of the study and about related research projects conducted for the Research Division of the National Endowment for the Arts. The following reports are available at libraries, bookstores or from their publishers: Socialization and Participation in the Arts Richard J. Orend Research Division Report #21 54 pages National Endowment for the Arts (1989) Available from the American Council on the Arts, 1285 Avenue of the . Americas, New York, NY 10019 Who Reads Literature? Nicholas Zill & Marianne Winglee Research Division Report #22 104 pages, 0-932020-86-0 Seven Locks Press (1990) $9.95 The Audience for American Art Museums J. Mark Davidson Schuster Research Division Report #23 60 pages, 0-929765-00-1 Seven Locks Press (1991) $10.95 In addition the following reports are available through the Education Research Information Center (ERIC) system: Dan Abreu, " Survey of Public Participation in the Arts, Musical Theater, Operetta, and Opera Attendees." April 1, 1987, ERIC Identification Number: ED 289 760. Carol Keegan, "Public Participation in Classical Ballet: A Special Analysis of the Ballet Data Collected in the 1982 and 1985 Survey of Public Participation in the Arts." April 30, 1987. ERIC Identification Number: ED 288 756. David Waterman, "Public Participation in the Arts Via the Media.' ■ September 1987, ERIC Identification Number: ED 290 674. Jerry West, "Public Participation in the Arts: Demands and Barriers.' ' ERIC Identification Number: ED 287 764. Harold Horowitz, "The American Jazz Audience." ERIC Identification Number: ED 280 757. John Robinson, et al., "Public Participation in the Arts: Final Report of the 1982 Survey." Survey Research Center, University of Maryland, January 1986. ERIC Identification Number: ED 264 168. John Robinson, etal., "Survey of Public Participation in the Arts: 1985 Volume I, Project Report." Survey Research Center, University of Maryland, March 1987. ERIC Identification Number: ED 289 763. Judith R. Blau, "The Geography of Arts Participation: Report on the 1982 and 1985 Surveys of Public Participation in the Arts." March 1987. ERIC Identifica- tion Number: ED 289 762. Paul DiMaggio, "Race, Ethnicity and Participation in the Arts: Patterns of Par- ticipation by Black, Hispanic and White Americans in Selected Activities from the 1982 and 1985 Surveys of Public Participation in the Arts." June 1987. ERIC Identification Number: ED 293 759. J. Mark Davidson Schuster, "An Inquiry into the Geographic Correlates of Government Arts Funding." ERIC Identification Number: ED 298 023. The documents are the original research reports as prepared by the investi- gators. They contain extensive information about methods, and numerous tables and figures. The ERIC collection is available at hundreds of libraries in the United States and abroad, as well as "on-line" from computerized information services. Requests for information about the purchase of microfiche or photocopies of these reports should be sent to: ERIC Document Reproduction Services, Consumer Service, P.O. Box 190, Arlington, VA 22210. 56 Expanding the Audience for the Performing Arts Becoming involved in the arts is not a one-step process but a progression through several stages. Upon this simple, yet valuable, premise, rests Expanding the Audience for the Performing Arts. According to the author, a person travels from disinterest to active attendance as they approach becoming part of the audience for the performing arts. For arts administrators seeking to expand their audiences, understanding where on this continuum rest the people they seek to attract is critical to their success. A six-stage model is proposed in this study and is tested using data from the Survey of Public Participation in the Arts. The findings validate the usefulness of this approach for both arts managers and future researchers in the field. Alan R. Andreasen University of Connecticut Research Division Report #24 National Endowment for the Arts Seven Locks Press Washington $10.95 0-929765-01-X