Working Paper Series, 20
An introduction to the literature on arts impact studies
Taking the Measure of Culture Conference
June 7-8, 2002
The author thanks Paul DiMaggio and Steve Tepper for their guidance and suggestions, and Jesse Mintz-Roth for his fine research assistance. Also, thanks are due to the Rockefeller Foundation for its generous support of this project. Please do not cite without permission. Direct any comments for the author to email@example.com.
As private and public agencies seek innovative ways to employ the arts to improve and strengthen communities, they have become increasingly interested in assessing the impact of their investments. In this context, arts advocates and researchers have made a variety of ambitious claims about how the arts impact communities. These claims, however, are made problematic by the many complications involved in studying the arts. Just consider the possible definitions of the phrase, “the arts impact communities.” When speaking of “the arts,” do we refer to individual participation (as audience member or direct involvement?), to the presence of arts organizations (non-profit and for-profit?) or to art/cultural districts, festivals or community arts? When speaking of “impact,” do we refer to economic, cultural or social impact; do we refer exclusively to direct community-level effects or do we also include individual- and organizational-level ones? By “communities,” do we mean regions, cities, neighborhoods, schools or ethnic groups?
Of course, there are no authoritative answers to these questions, since different research questions require different definitions. And as one might expect, arts impact studies employ these heterogeneous definitions in a variety of combinations. Given this array of definitions, how would we go about measuring the impact of the arts on communities? One problem is that researchers and arts advocates rarely seem to consider such complications when making claims about the broader impact of the arts, and seldom discuss the implications of making particular theoretical and methodological choices.1
In this paper, I will lay out some of the issues that need to be addressed when thinking about and studying how the arts impact communities, in addition to providing an introduction to the literature on arts impact studies. I begin discussing the mechanisms through which the arts are said to have an impact. Following this is a
To be fair, many studies are not intended to examine the impact of arts programs on the broader community, but only at a relatively limited number of participants. Nevertheless, the findings of these studies are often used by arts advocates to support more ambitious claims about the impact of the arts on communities.
discussion of key theoretical and methodological issues involved in studying the impact of the arts. I conclude by suggesting areas for further research and reflecting on the limitations of past research.
The arts have been heralded as a panacea for all kinds of problems Arts-integrated school curricula supposedly improve academic performance and student discipline (Fiske 1999; Remer 1990). The arts revitalize neighborhoods and promote economic prosperity (Costello 1998; SCDCAC 2001; Stanziola 1999; Walesh 2001). Participation in the arts improves physical and psychological well-being (Baklien 2000; Ball and Keating 2002; Bygren, Konlaan and Johansson 1996; Turner and Senior 2000). The arts provide a catalyst for the creation of social capital and the attainment of important community goals (Goss 2000; Matarasso 1997; Williams 1995).
Given these claims, the question arises of how to elaborate the causal mechanisms through which the arts have an impact (i.e., the intervening factors that connect a particular arts activity with a specific outcome). Below is a grid that lays out two dimensions that will help in thinking about this.2 The rows represent three aspects of the arts typically highlighted in the literature: direct involvement in arts organizations, especially that which entails personal engagement in some form of creative activity (most often associated with community arts programs and the use of the arts in education); participation in the arts as an audience member (mostly associated with cognitive ability, cultural capital and health improvement arguments, as well as economic impact studies of the arts – i.e., whether the arts have an economic impact by drawing audience dollars from outside the community); and the presence of arts organizations in a community (mostly associated with economic impact studies and social capital arguments).
This grid expands and builds upon a typology of arts effects developed in a research proposal to the Wallace-Readers Digest Funds by Kevin McCarthy (2002) of the RAND Corporation.
The columns represent types of impact and are divided into individual and community levels. Individual-level effects are relevant for the purposes of community impact studies to the extent that the impact of the arts on individuals aggregates to the community. (For example, some individual-level impacts, such as ‘personal enjoyment,’ may not have any consequences on community life.) The three types of individual impacts are material (mainly health), cognitive/psychological and interpersonal. Types of community-level effects, which are roughly homologous to individual-level ones, are economic, cultural and social. The cells of the table contain, where relevant, specific impacts claimed in the literature.
The grid helps to assess how different levels and types of artistic inputs are related to different types of outputs. It can be taken as axiomatic that, other things being equal, the more widespread and/or intense the participation of community members (who are not involved as professionals), the greater the impact the arts will have on cultural and social factors.3 However, direct involvement is more intense than audience participation, whereas audience participation is more widespread than direct involvement. (To the extent that community arts programs are geared towards producing some kind of public ‘show’ [art show, play, reading, festival, etc.], they will tend to optimize both dimensions of participation.) Greater concentrations of artists and arts-related organizations lead to higher degrees of arts participation among residents, directly and as audience members (Stern and Seifert 2000). There is also often a trade-off between different types of arts activities in terms of the kinds of benefits they are most likely to produce. For example, a well-respected theater employing a professional staff is more likely to draw visitors and tourists from outside the community than is a local community arts project exhibition, and hence it will have a greater economic impact. But, since the level of participation among community members lacks intensity in the case of the theater, it has less potential for
Note that this does not apply to economic impacts, since those rely primarily on bringing revenue from outside the community. In this example, the type of participation is ‘widespread’ and the degree is the ‘intensity.’ building social capital and a sense of collective efficacy. Both the theater and the community arts project may enhance community pride and self-image.
It should be noted that, with the exception of economic impact studies, almost all other research focuses on the benefits that accrue to individuals and organizations involved in the arts, rather than the direct impact of the arts on a community as such.4 I will discuss this problem of aggregation later in the paper, but for now I bracket it in favor of explicating mechanisms that connect well-defined arts activities to well-defined outcomes.5 The following discussion is organized by claims about the impact of the arts. I focus on three types of claims: first, claims that the arts build social capital; second, claims that the arts improve the economy; and third, claims that the arts are good for individuals. These three broad claims capture virtually all of the more specific assertions about the impact of the arts.
Claim: The arts increase social capital6 and community cohesion
Claims under this heading encompass the last two columns of the table –community-level cultural and social impacts – as well as interpersonal effects. Virtually all studies that make this claim examine the effects of community arts programs on the participants and organizations involved (Costello 1998; Dolan 1995; Dreeszen 1992; Fritschner and Hoffman 1984; CDA 2000; Krieger 2001; Landry et al. 1996; Matarasso 1997; Matzke 2000; Murphy 1995; Ogilvie 2000; Preston 1983; Stern et al. 1994; Stern and Seifert 2002; Trent 2000; Williams 1995; Wollheim 2000). The following discussion draws on all of these studies.
Although quite varied, community arts programs are grassroots organizations that attempt to use the arts as a tool for human or material development (Costello 1998). Community arts programs almost universally involve community members in a
4 One notable exception is Stern (1999; 2001), who demonstrates that a greater concentration of arts organizations in a neighborhood leads to longer-lasting ethnic and economic diversity in that neighborhood.
5 By aggregation, I refer to the process by which effects on individuals, taken together, can combine to have an influence on the broader community.
6 Scholars often fail to define precisely what they mean by social capital. According to Robert Putnam’s influential definition, “social capital refers to connections among individuals – social networks and the norms of reciprocity and trustworthiness that arise from them,” which may facilitate coordination and cooperation for mutual benefit (2001: 19) creative activity leading to a public performance or exhibit. As defined by the Ontario Arts Council (2002), “Community Arts is an art process that involves professional artists and community members in a collaborative creative process resulting in collective experience and public expression. It provides a way for communities to express themselves; enables artists, through financial or other supports, to engage in creative activity with communities; and is collaborative – the creative process is equally important as the artistic outcome.” (Note that this is different from such things as local, neighborhood knitting groups.) Community arts programs often involve people who are disadvantaged in some way (at-risk youth, ethnic minorities, people in a poor neighborhood) and are designed in the context of some larger goal, such as neighborhood improvement (typically aesthetic) or learning and teaching about diverse cultures (multiculturalism). These goals are usually the basis for claims about the politically transformative potential of community arts projects (e.g., see Williams 1997). Regardless of the ultimate purpose(s) to which social capital is to be put, community arts programs are said to build social capital by boosting individuals’ ability and motivation to be civically engaged, as well as building organizational capacity for effective action. This is specifically accomplished by:
• Creating a venue that draws people together who would otherwise not be engaged in constructive social activity.
• Fostering trust between participants and thereby increasing their generalized trust of others
• Providing an experience of collective efficacy and civic engagement, which spurs participants to further collective action
• Arts events may be a source of pride for residents (participants and non-participants alike) in their community, increasing their sense of connection to that community.
• Providing an experience for participants to learn technical and interpersonal skills important for collective organizing
• Increasing the scope of individuals’ social networks
• Providing an experience for the organizations involved to enhance their capacities. Much of this comes when organizations’ establish ties and learn how to work, consult and coordinate with other organizations and government bodies in order to accomplish their goals.
A case study from Williams’ (1995: 101-106)) research in Australia provides an example of these mechanisms. The study was conducted on a sample of recipients of community-based arts grants provided by the Australia Council. One of these grants was given to a small group of women residents of Longlea, a suburb of Brisbane. Their goal was to beautify their blighted community center, which involved local residents in the creation of artworks around the community center. This drew together townspeople who might otherwise have stayed at home to engage in a constructive social activity. As people worked collaboratively on the project and got to know each other better, their mutual trust increased. Their success in negotiating with the municipal bureaucracy in order to accomplish the task gave participants a newfound sense that they could accomplish other goals. The community group and individuals coordinating the efforts learned organizing skills, learned how to navigate the bureaucracy and built relationships with the municipal and regional government. Finally, the people involved felt an increased sense of pride and appreciation of their town.
Claim: The arts have a beneficial impact on the economy
Economic impacts are perhaps the most widely touted benefits of the arts. The literature on economic impact studies of the arts tends to fall into two categories: on the one hand, advocacy studies based on quick appraisals that often exaggerate the impact of the arts (Azmier 2002; Bryan 1998; Eckstein 1995; Perryman 2001; SCDCAC 2001; Singer 2000; Walesh 2001). On the other hand are more rigorous studies — which, overtime, show increasing methodological refinement (Cohen 1994; Costello 1998; CPC 2002; Cwi 1980a; Cwi 1980b; Cwi and Lyall 1977; DiNoto and Merk 1993; Frey 1998; Gazel 1997; Kling, Revier and Sable 2001; Mitchell 1993; O’Hagan and Duffy 1987; Port Authority of New York and New Jersey 1983; Radich 1987; Rolph 2001; Sable and Kling 2001; Seaman 1997; Stern and Seifert 2000; Throsby 2001; Travers, Stokes and Kleinmann 1997). In the following discussion, I have tried to rely on these more rigorous studies.
• The arts attract visitors (art as ‘export’ industry):
Tourists visit a community primarily in order to attend an arts event (alternatively, tourists may prolong a trip in order to attend an arts event). They will spend directly on the arts event and may also shop, eat at a local restaurant and/or stay at a hotel in the community. To the extent that these tourist dollars are spent by the arts organization – as well as the stores, restaurants and hotels – on local goods and services, the dollars brought in to the community for an arts event will have indirect multiplier effects on the local economy.
• The arts attract residents and businesses:
The density of arts organizations and prevalence of arts events may play a role in attracting residents and businesses to (re)locate to a community by improving its image and making it more appealing. This is especially true for attracting highly skilled, high-wage residents, who will have a larger economic impact than less-skilled people. Businesses, especially those that employ highly trained mobile personnel, may consider the presence of art venues when making (re)location decisions (Cwi 1980b: 18-19). The presence of the arts (i.e., improved image of an area) may work to enhance the impact of tax incentives for business location decisions (Costello 1998: 147-9).
High concentrations of artists and/or high-skilled workers may produce agglomeration effects, where businesses (especially those in the fast-growing ‘creative industries’ (Walesh 2001)) are drawn to an area because of the availability of creative talent and/or high-skilled workers, and vice versa.
• The arts attract investments:
By improving a community’s image, people may feel more confident about investing in that community. So for example, people might be
7 An indirect multiplier is based on the idea that a portion of each dollar spent on some good or service is then used by the recipient to pay for more goods and services. To the extent that the money circulates within a community (e.g., a city), it ‘multiplies’ within that community. So for example, if you spend $20 on a ticket to a play, the playhouse turns around and spends $15 of that for set design supplies from local markets. The employees also spend locally some portion of their income that is derived from that $15 to pay for more goods and services; and the stores from which they bought supplies in turn use some of that money to pay their workers and buy more supplies, and so on. This ‘multiplies’ the value of the initial $20. more likely to buy property in an area that they feel is “up-and-coming” because of the presence of the arts. Or, banks may be more likely to lend to businesses in areas perceived as more secure and stable, and so on.
One problem with determining the impact of the arts is distinguishing between revenue from locals vs. revenue from tourists, and among the latter determining the extent to which the arts drew them to visit the community. Expenditures by locals should not be included in studies of the economic impact of the arts, because the arts may simply represent an alternative outlet for spending (rather than an additional outlet), thus representing no net differences on the local economy (assuming equal multiplier effects among outlets). In terms of private and public subsidies for the arts, it is difficult to determine the opportunity costs of investing the money in other things (i.e., whether investing the same amount of money in something else would have a stronger impact on the economy). There is scant evidence on whether money spent on the arts is more likely to circulate locally than money spent in other areas (though see Palmer 2002 for a comparison of arts performances versus sports arenas).
As an example of how the arts may have an economic impact, let us examine a summer theater festival that a small town puts on every year(Mitchell 1993). This festival draws thousands of visitors who come – some from far away, but most from the surrounding area – in order to attend the performance. These visitors spend money on tickets as well as restaurants, hotels, parking and retail shopping. (In this sense, the arts are said to be an ‘export’ industry to the extent that they bring in money from outside the local economy.) This spending has a direct positive impact on the town’s economy. Indirectly, this spending has what is called a “multiplier effect” to the extent that those dollars re-circulate in the local economy as a result of spending on local goods and services by the festival and the other business.
Claim: The arts are good for individuals
Claims that the arts are good for individuals take many forms. The arts have been said to improve health, mental well-being, cognitive functioning, creative ability and academic performance.
• The arts improve individual health.
Either engaging in creative activity or simply attending some kind of artistic event appears to improve physical health (Angus 1999; Baklien 2000; Ball and Keating 2002; Bygren, Konlaan and Johansson 1996; HDA 2000; Thoits and Hewitt 2001). This could be due in part to its ability to relieve stress. Also, arts engagement widens and strengthens social bonds, which also improves health (Baklien 2000: 250-51; Ball and Keating 2002). On a more physiological level, Bygren, Konlaan and Johansson (1996: 1580) explain: “we know that the organism responds with changes in the humoral nervous system–for example, verbal expression of traumatic experiences through writing or talking improves physical health, enhances immune function, and is associated with fewer medical visits.”
• The arts improve psychological well-being.
Here we have to distinguish between passive and active participation. Attending arts events may be stimulating and relieve stress, hence leading to improved happiness/ life satisfaction. Active participation in the arts leads, in addition, to improved self-concept and sense of control over one’s life. There are different reasons why this might be so. Lots of the anecdotal evidence comes from community arts programs, some of which are geared towards poor, marginal or ‘at-risk’ populations (Lynch and Chosa 1996; Seham 1997; Weitz 1996; Williams 1995). This is backed up by the little – and poor quality – survey data that do exist. To the extent that the creation and completion of some arts project provides an opportunity to such participants to succeed and gain some positive public recognition, it will improve their sense of control over their life and self-concept (Fiske 1999; Jackson 1979; Randall, Magie and Miller 1997; Seham 1997; Weitz 1996). To date, there has been no systematic comparison between community arts programs operating in different socio-economic climates to see whether such effects appear to be uniform.
• The arts improve skills, cultural capital and creativity.
Here again we have to distinguish between passive and active participation. Audience members may gain some new knowledge or cultural capital8 by attending arts events. There is also the so-called Mozart effect showing that children who listen to Mozart (and other similar stimuli) show improved performance on visuo-spatial reasoning tests – although the effect may not last (Chabris et al. 1999; Hetland 2000). Individuals directly involved in creating or organizing artistic activity may learn skills that they did not previously have and may demonstrate greater creativity (Fiske 1999; Randall, Magie and Miller 1997; Rolph 2001; Seham 1997; Sharp 2001; Weitz 1996). On the whole, education studies show that kids engaged in an arts class will do better in other subjects and that an arts-integrated curriculum improves school performance (Albert 1995; Fiske 1999; Jackson 1979; Remer 1990; Weitz 1996; Winner and Hetland 2000). The basic reason for this may be that children find learning through artistic/creative activity much more enjoyable, and so they will have an easier time engaging with the material. It is important to point out, however, that most studies do not control sufficiently for self-selection into arts activities and the effects are not as dramatic as boosters would claim.
The Coming Up Taller report (Weitz 1996) provides concrete examples of some of these mechanisms. The report identifies arts-training programs targeted at at-risk youth and seeks to understand why these programs work. At least two of the programs involved working with sentenced juvenile offenders. One program taught musical theater; the other painting. Both programs appeared to enhance the self-esteem of their participants, because they learned new skills, found that they had undiscovered talents, and received positive recognition from peers and others when they perform or exhibit their work. Learning new skills may also improve their position on the job market. For example, in addition to learning singing, dancing and acting, participants in the music theater program also learn about the technical side of producing a play, such as lighting, set-design and sound. Also, performing a play or doing other kinds of artistic activity can provide a means of learning that children find much more fun and engaging. As a result they will learn and absorb the material better.
8 I use the most restricted definition of cultural capital as simply knowledge of the fine arts. For example, in taking an arts class, one learns something about aesthetics and art appreciation and perhaps about art history. Such knowledge has been linked to better school performance and improvement of other life outcomes (DiMaggio 1982; DiMaggio and Mohr 1985).
THEORETICAL & METHODOLOGICAL ISSUES
As I pointed out at the start, the phrase “arts impact communities” admits of many possible definitions. Specifying these definitions is an important task that researchers often ignore. Here, I briefly sketch some dimensions along which these terms can be defined.
Defining “the arts” – Different research projects rarely define “the arts” in the same way, and often the same study will include diverse activities and organizations, including professional opera companies, neighborhood cultural centers, community arts programs and in some cases even major league sports. There are several dimensions along which definitions of the arts might be specified: genre or art-form (whether the activity is painting, singing, acting, etc.); sector (whether the organization involved is non-profit, commercial or governmental); time (duration of the arts activity or involvement); place (where does the activity/performance take place); group participation (whether the activity is done alone, in small groups or in large groups); medium (whether the arts is live, recorded or Web-based); and mode of participation (whether involvement is active art-making, organizational volunteering or audience participation).
This last dimension provides a distinction useful for classifying prior studies. Some studies look at the effect of participation in the arts on those who are directly involved, especially when they are engaged in art-making. Such studies often examine the impact of community arts programs (CDA 2000; Landry et al. 1996; Matarasso 1997; Matzke 2000; Murphy 1995; Trent 2000; Williams 1995; Wollheim 2000) or arts-centered teaching programs (Albert 1995; Fiske 1999; Jackson 1979; Remer 1990; Seham 1997; Sharp 2001; Weitz 1996; Winner and Hetland 2000), usually on the participants themselves but sometimes on the local community. Other studies look at arts attendance, occasionally examining the impact of the arts on their audience (Bygren, Konlaan and Johansson 1996; Chabris et al. 1999; Hetland 2000; Landry et al. 1996; Matarasso 1999; Williams 1995), but most often focusing on the audience’s impact on the local economy (Bendixen 1997; DiMaggio, Useem and Brown 1978; Frey 1998; Gazel 1997; Laing and York 2000; Mitchell 1993; O’Hagan and Duffy 1987; SATC 1998).9 A third major focus of arts research is on the presence and density of arts organizations, looking sometimes at how these factors affect involvement in the arts and other local organizations (Stern 1999; Stern and Seifert 2000), but typically emphasizing the impact of arts organizations on the local economy (Cohen 1994; Costello 1998; Cwi 1980a; DiNoto and Merk 1993; Port Authority of New York and New Jersey 1983; Stern 2001; Stern and Seifert 2000; Travers, Stokes and Kleinmann 1997). Here I have simply provided a quick survey of the definitional terrain of arts studies. The broader point to be made is simply that it is crucial to define precisely what are “the arts” that one is studying, because different arts activities are likely to lead to a different set of outcomes. Furthermore, the use of vague and diverse definitions of the “arts” makes comparability and accumulation across studies very difficult.
Defining “impact” — As this discussion illustrates, defining the scope of what is meant by the “arts” goes some length towards delimiting their potential impact. (For example, a school arts program is not likely to have an appreciable impact on the economy of a city.) Like the arts, there are also a number of dimensions along which the scope of the impact(s) ought to be clarified: whether the impact is on individuals, institutions/organizations, communities or the economy; whether it is direct or indirect (e.g., does it indirectly affect communities by affecting individuals?); whether the impact is short-term or long-term; whether impacts are greater for some groups and individuals than for others; and whether the impact is social, cultural, psychological, economic, and so on. These dimensions are often under-specified, and as a result findings can be easily inflated or over generalized (e.g., a small, short-term impact on a subgroup of people might be viewed as an enduring impact on a broader
9 Dollars spent in a community by cultural tourists are only one way in which the arts are said to have an economic impact. class of residents). Furthermore, as Cwi (1987) notes, the policy relevance of most arts program evaluations studies is limited, because of their failure to adequately specify the impact that the program is intended to have.
Defining “community” – Community can be defined in a variety of ways: as a geographic region, municipality, neighborhood (itself open to a variety of definitions), or ethnic group. In general, researchers use one of two criteria in defining community: propinquity and group membership. With the first criterion, researchers define community in terms of people’s proximity to one another and study things like neighborhoods, schools, cities or SMSAs. For example, the Social Impact of the Arts Project (SIAP) usually uses census ‘block groups’ as part of its definition of neighborhood, and also historically institutionalized, widely recognized neighborhoods, such as Germantown in Philadelphia or the ‘south side’ of Chicago (Stern 2001). Another common way to define community is as a legally distinct area, such as a town, city or state (Cwi 1980a; Cwi 1982; Cwi and Lyall 1977; DiNoto and Merk 1993; Gazel 1997; Mitchell 1993; NALAA 1994; Perryman 2001). Studies using this criterion usually focus on the economic impact of the arts, so examining a well-defined tax base makes sense. Alternatively, researchers may study community defined by group membership, categorizing people on the basis of race/ethnicity, national origin, gender, sexual orientation, occupation and so forth.
Researchers may use one of two methods for classifying people into communities: one method defines community on the basis of criteria imposed by the researcher; the other defines community in accord with individuals’ self-identification (see Stern et al. 1994 for an example of this). Note that the basis of people’s self-identification can come from many sources. It may be coterminous with proximity- or legally-based definitions (e.g., “I’m from Germantown,” “I’m from Robert Taylor Homes” or “I’m from Atlanta).” People may also self-identify on the basis of group membership. Some community-based arts programs are organized around such communities. For example, one program studied by Williams (1995) was designed to have aboriginal children in a rural Australian town express their culture. It is important to distinguish between researcher-imposed vs. self-identified definitions of community. It is possible, for example, that in order to understand if and how the arts contribute to such subjective outcomes as increased trust of others, greater pride in one’s community and motivation to work towards collective ends, then one needs to take an inductive approach to this question of community (e.g., using definitions that members themselves put forward). And if there is a disjuncture between the researcher’s definition of a community and the self-identifications of its members, then the researcher may fail to find evidence of, for example, social solidarity (because s/he would be looking in the wrong places for evidence).10
Whether researcher-imposed or not, clearly specifying the scope of the community is crucial when trying to think about how the arts impact a community directly, as well as the related problem of aggregation.
The Problem of Aggregation
One of the more vexing issues confronting anyone wishing to understand the impact of the arts on communities is the question of how to link micro-level effects on individuals to the more macro level of the community. Except for economic impact studies, virtually every arts impact study examines how the arts affect individuals (though see Stern 1999; 2001), whether by improving their health (Bygren, Konlaan and Johansson 1996; Costello 1998), their self-esteem (Weitz 1996), their skills, talents and knowledge (Fiske 1999; Winner and Hetland 2000), or their tolerance of other cultures (Matarasso 1997; Williams 1995). In some cases, researchers have also argued that the creation of arts programs (usually made possible by government or private grants) increases the capacities of arts organizations, for example by enhancing their ability to work with local government agencies (Stern and Seifert 2002; Williams 1997). In this case, the problem becomes one of aggregating organizations rather than individuals.
10 I am grateful to Paul DiMaggio for suggesting these last two points.
Note that defining the scope of the community in question is critical to the problem of aggregation. For example, other things equal, a small community arts program is more likely to have an impact on people in the neighborhood in which it operates than on people living on the other side of town. But without having to define community, at least five general ways in which individual/organizational-level effects might aggregate can be distinguished:
1. Most obviously, one could simply talk in terms of the percentage of individuals/organizations in a population that are affected. Social capital is typically conceived of in such a manner, where a community with a higher percentage of individuals participating in civic groups and/or a greater density of such groups is considered to have greater social capital. Hence, if arts programs get more individuals involved in community groups, then they increase the community’s social capital.
2. Closely related to this is the idea that there may be threshold levels or ‘tipping points’ (Gladwell 2000) at which individual/organizational-level effects begin to have community-level consequences. In this case, as in number 1 above, an unresolved issue is determining the level at which these effects can properly be said to have an impact on the ‘community.’
3. The presence of the arts and/or participation by community members may have an impact on community norms or the “opinion climate.” For example, the presences and performances of a multicultural theater may reinforce norms about multiculturalism and diversity or free expression..
4. To the extent that arts organizations serve as a catalyst in the creation of ties between dispersed individuals and organizations (who would not otherwise establish ties), these networks, may then be used to accomplish other community goals.
5. Communities may be affected when a few key individuals and/or organizations are affected. For example, a successful community arts program may influence the perceptions of key government officials and make them more likely to support such programs in the future. Or successful arts-based neighborhood revitalization programs targeted at particular crime-ridden neighborhoods or juvenile offenders may lower the overall crime rate.
6. Finally, individuals and groups involved in the arts can be said to affect the community by creating public goods.11 The value of arts as a public good (its contingent valuation) is usually measured by willingness-to-pay surveys12 (CPC 2002; Kling, Revier and Sable 2001; Sable and Kling 2001; Seaman 1997; Throsby 2001).
As with much social research, arts impact studies typically suffer from selection bias problems, which make it difficult to identify clearly the causal role of the arts.13 This problem is usually expressed by the truism that ‘correlation is not causation.’ For example, research indicates that people who participate in the arts are healthier and happier (Bygren, Konlaan and Johansson 1996; Costello 1998; Thoits and Hewitt 2001). But, does this mean that arts involvement makes people healthier and happier, or that such people are more likely to get involved in the arts? Do arts programs build social capital, or are communities with higher social capital more likely to initiate arts programs? Usually, the answer to such questions is ‘both.’ On average, healthier people are more likely to volunteer in arts programs, but that activity likely improves their health as well (Thoits and Hewitt 2001). Communities with greater social capital are more likely to initiate arts programs, but those programs may further promote the building of social capital. Most likely, health or social capital
11 Outdoor sculpture is a good example of public goods, since many people can enjoy it. But, people don’t necessarily need to use/enjoy art for it to be a public good.
12 Willingness-to-pay surveys ask respondents how much they would be willing to pay (usually in taxes) to support some artistic activity (e.g., “How much would you be willing to pay in taxes to support the NEA?”). People who don’t patronize the arts still report that they are willing to pay to fund them, and this is interpreted to mean that the arts are valuable to them.
13 Generically, selection bias means that the sample (i.e., the people and/or organizations that one is studying) is not representative of the entire population, leading to conclusions that are not valid. In arts research, the most pernicious of these is self-selection bias: since people who choose participate in the arts may be different from others, that difference may explain the observed outcome rather than the arts activity. would not have improved in the same way and to the same degree had the arts programs been absent. When seen from this perspective, selection issues – when recognized and handled appropriately – arguably do not present an intractable problem to arts impact studies.
Lack of Appropriate Comparisons
From a policy perspective, however, the issue is no longer whether the existence of the arts has a beneficial impact, but whether money spent on arts programs will have more of an impact than other programs. Indeed, one flaw with the literature on arts impact is the lack of studies that compare the arts with other programs or industries. The key question for policy-makers (or grant-givers) is this: given some pre-defined goal (improving the economy, attracting tourism, improving education, reforming at-risk adolescents, etc.), how can that goal be most effectively reached? Thus, the issue changes from ‘did this program work at all’ to ‘did this program work better than another?’ Instead of ‘what are the benefits of the arts,’ the question becomes ‘what are the opportunity costs14 of using this money to fund the arts’? For example, are arts programs for at-risk youth more effective than the Boy Scouts or midnight basketball? Do arts programs draw people away from other high-impact activities in which they would otherwise be involved, such as environmental activism or charity; would public money be better spent on things like transportation infrastructure or police? Determining whether a program is more ‘effective’ than another is of course no simple matter and demands precise definition of the goal of the program, but none of the studies I reviewed adequately addressed this issue. The difficulty of the comparison is compounded by the fact that many of the benefits we associate with the arts, like increased creativity or feelings of well-being, are ‘intangible’ and therefore difficult to measure. However, to the extent that the arts do potentially provide something unique, the lack of comparative studies make it that much more difficult to concretely demonstrate the unique contribution of the arts.
14 Opportunity costs basically mean that when you spend your money or time in one activity/investment, there is a cost of not being able to use that time or money in some other activity/investment.
In addition to ignoring opportunity costs, arts impact studies typically ignore the potentially negative impacts of the arts. For example, given the broad definition of the ‘arts’ found in many studies, the negative impact of such events as raves or rock concerts – for example noise pollution and delinquency – largely goes ignored (though see Gazel  for an economic impact study of a Grateful Dead concert in Las Vegas that took into account the city’s extra expenditures on security for the event). Or, if an arts’ program builds social solidarity among some ethnic group, could this lead to greater balkanization of the community? Zukin’s (1989) study of New York City shows that the presence of arts activities and artists in a poor neighborhood may be a harbinger of gentrification (though see Stern  for evidence from other cities that the presence of arts organizations leads to lasting diversity). To the extent that studies do examine failed programs, they tend to focus on the causes of failure rather than its consequences (Matarasso 1997; Williams 1997). In short, those who investigate the impact of the arts need to be more aware of potential negative as well as positive impacts.
Lack of Adequate Data
Most arts impact studies are based on cross-sectional data, making inferences about selection and the causal role of the arts exceedingly difficult. The lack of over-time data also makes it impossible to see how long the effects of an arts program persist.15 Furthermore, the sample sizes of many studies are too small for making proper statistical inferences.16 In many instances, researchers employ multiple or comparative case study approaches, for example by studying several different community arts programs. Despite the strengths of this type of analysis for describing
15 Williams’ (1997) study in Australia did follow some of the communities she studied for several years after the initial program; this enabled her to draw inferences about what factors lead to sustained impact.
16 Statistical inferences (for example, determining with what degree of confidence we can say that children in arts programs do better in school) are based on the premise that the sample is representative of the entire population. The representativeness of small sample sizes cannot be guaranteed with a high level of confidence in detail the supposed consequences of particular arts programs on particular individuals, these studies are limited in a number of ways:
First, they tend to rely exclusively on the subjective accounts of people involved in the art programs or audience members in order to support their claims – in short, they tend to be anecdote-rich and evidence-poor (though perhaps there’s an argument to be made that a mountain of anecdotes serves as some kind of evidence). The fundamental question here is whether impact can be measured solely or largely on the basis of these accounts, especially considering that participants almost always self-select into participation. What would happen if people were randomly assigned into an “arts treatment” group? This is closely related to another problem with these, as with other arts impact studies, which is that they tend to sample only treated groups. For example, questionnaires go only to people who are centrally or closely involved in a particular arts program, rarely asking community members what consequences the program had on them (though see Matarasso 1997). Also, evaluation studies only look at organizations or communities that won the supporting grant (whose impact the study is intended to measure), never comparing it with a similar community that didn’t win a grant, let alone one that never even applied. No doubt it is especially difficult to create a quasi-experimental design in applied social science, but arts impact research seldom makes an effort to achieve this goal. (One problem, of course, has been lack of adequate funding to undertake such an effort.) More generically, the problem with in-depth case studies is that they are rarely representative of the overall population.
Specification of Context Effects and Intervening Factors
Researchers studying the impact of the arts are rarely sensitive to contextual or intervening factors that influence the outcomes they find. This is important for generalizing from the findings of a specific study. To take a simple example, many studies claim that the arts have a beneficial economic impact. However, it is likely that this impact varies depending on the size of the community under discussion and the size and density of arts organizations/events. Thus, in order for the arts to make an appreciable (and perhaps measurable) impact on the economy of a large city, it will likely require the development of an arts district (such as the Temple Bar in Dublin, see Costello 1998). An annual drama festival is likely to have little economic impact on a large city (though it may have an appreciable impact on the neighborhood in which it is located), but may be a decisive factor in the economy of a small town (Mitchell 1993). And local community arts projects are likely to have little economic impact. The National Association of Local Arts Agencies study is one of the best to date in selecting arts activities of various sizes across a wide range of municipalities (Cohen 1994). The point is simply that arts impact researchers need to begin to think more seriously about the conditions under which their results do – or do not – generalize.
Research on the how the arts impact communities is a burgeoning and wide-ranging field of research. Despite the variety of research subjects and methodologies alive and well in the field, there are a number of avenues this literature has yet to explore. For example, researchers study formal groups and organizations to the exclusion of more informal groups, such as local neighborhood knitting groups and the like. Case studies tend to focus on arts programs developed for marginal populations (like at-risk children); it would be interesting to see what could be learned from comparing these programs to ones where most of the participants are middle- or upper-middle class. Also, researchers often study community arts programs that have some kind of political or social goal: what might be learned by comparing these organizations to those that have no such goal? And in terms of determining their relative economic impact, we need to know whether arts organizations tend to spend more money in the local economy and on locally-produced goods than do other organizations/businesses. These examples point to a larger problem with the research in this field, especially those that use multiple, in-depth case studies: the cases are generally not chosen in such a way as to gain much empirical ‘leverage’ from the comparison. Cases appear to be selected on the basis of capturing the widest diversity of programs possible – sometimes with an implication that this will ensure representativeness. The most that comes from this sort of comparison is a list of some factors that appear to affect the relative success of the programs. Researchers need to think more about the logic driving their case selection, so that they can get more from their comparisons.
The criticisms that I have enumerated in this paper could apply to most bodies of social research. But, the field of cultural policy studies is young and resources are scarce. Therefore, it is perhaps more important than in other fields that small investments in research yield strong results that can be leveraged to advance public policy and private philanthropy. As a result, it is especially incumbent upon arts researchers to carefully specify their definitions and think critically about the theoretical and empirical issues confronting them when attempting to take the measure of culture.
Albert, Maria. 1995. “Impact of an arts-integrated social studies curriculum on eighth graders’ thinking capacities.” Pp. xi, 344 leaves ; p., 29 cm. Lexington, Ky.: University of Kentucky.
Angus, John. 1999. An Enquiry concerning Possible Methods for evaluation Arts for Health Projects. Bath, UK: Community Health.
Azmier, Jason J. 2002. “Culture and Economic Competitveness: An Emerging Role for the Arts in Canada.” Canada West Foundation.
Baklien, Bergljot. 2000. “Culture is Healthy.” International Journal of Cultural Policy 7.
Ball, Susan, and Clare Keating. 2002. “Researching for Arts and Health’s Sake.” in 2nd Conference on Cultural Policy Research. Wellington, NZ.
Bendixen, Petere. 1997. “Cultural Tourism – Economic Success at the Expense of Culture?” International Journal of Cultural Policy 4.
Bryan, Jane. 1998. The economic impact of the arts and cultural industries in Wales. Cardiff: [Cardiff Business School?].
Bygren, Lars O., Boinkum B. Konlaan, and Sven-Erik Johansson. 1996. “Attendance at cultural events, reading books or periodicals, and making music or singing in a choir as determinants for survival: Swedish interview survey of living conditions.” British Medical Journal 313:1577-1580.
Chabris, Christopher F, Kenneth M Steele, Simone Dalla Bella, Isbelle Peretz, Tracy Dunlop, Lloyd A Dawe, G Keith Humphrey, Roberta A Shannon, Johnny L Jr Kirby, CG Olmstead, and Frances H Rauscher. 1999. “Prelude or requiem for the “Mozart Effect”.” Nature 400:826-828.
Cohen, Randy. 1994. Arts in the local economy: final report. Washington: National Assembly of Local Arts Agencies.
Costello, Donal Joseph. 1998. “The Economic and Social Impact of the Arts on Urban and Community Development.” Pp. 1333-A in Dissertation Abstracts International, A: The Humanities and Social Sciences. Pittsburgh: University of Pittsburgh.
CPC. 2002. “Contingent Valuation of Culture Conference.” Cultural Policy Center at the University of Chicago. http://culturalpolicy.uchicago.edu/cvmconf.html.
Cwi, David. 1980a. The economic impact of ten cultural institutions on the economy of the Springfield, Illinois SMSA. Springfield, Ill.: Center for the Study of Middle-size Cities Sangamon State University.
—. 1980b. The role of the arts in urban economic development. Washington, D.C.: Economic Research Division Economic Development Administration.
—. 1982. The arts in New Jersey : status, impacts, needs. Trenton, N.J.: The Council.
Cwi, David, and Katharine Lyall. 1977. A model to assess the local economic impact of arts institutions : the Baltimore case study. Baltimore: Center for Metropolitan Planning and Research the Johns Hopkins Univ.
DiMaggio, Paul. 1982. “Cultural Capital and School Success: The Impact of Status Culture Participation on the Grades of U.S. High School Students.” American Journal of Sociology 47:189-201.
DiMaggio, Paul, and John Mohr. 1985. “Cultural Capital, Educational Attainment, and Marital Selection.” American Journal of Sociology 90:1231-1261.
DiMaggio, Paul, Michael Useem, and Paula Brown. 1978. “Audience Studies of the Performing Arts and Museums: A Critical Review.” Research Division, The National Endowment for the Arts.
DiNoto, Michael J, and Lawrence H Merk. 1993. “Small Economy Estimates of the Impact of the Arts.” Journal of Cultural Economics 17:41-54.
Dolan, Teresa. 1995. Community Arts: Helping to Build Communities? Taken from a Southern Ireland perspective. London: City University.
Dreeszen, Craig. 1992. “Intersections: Community Arts and Education Collaborations.” Journal of Arts, Management, Law and Society 22:211-240.
Eckstein, Jeremy. 1995. The Contribution of the Cultural Sector to the UK economy. London: Policy Studies Institute.
Fiske, Edward B. 1999. Champions of change : the impact of the arts on learning. Washington, DC: Arts Education Partnership President’s Committee on the Arts and the Humanities.
Frey, Bruno S. 1998. “Superstar Museums: An Economic Analysis.” Journal of Cultural Economics 22:113-25.
Fritschner, Linda Marie, and Miles K. Hoffman. 1984. “The Community and the Local Art Center.” in American Sociological Association.
Gazel, Ricardo. 1997. “Beyond Rock and Roll: The Economic Impact of the Grateful Dead on a Local Economy.” Journal of Cultural Economics 21:41-55.
Gladwell, Malcolm. 2000. The Tipping Point: How Little Things Can Make a Big Difference. New York: Little, Brown and Company.
Goss, Kristin. 2000. Bettertogether : the report of the Saguaro Seminar on Civic Engagement in America. Cambridge, MA: Saguaro Seminar Civic Engagement in America John F. Kennedy School of Government Harvard University. http://www.bettertogether.org/bt%5Freport.pdf.
HDA. 2000. Art for health: a review of good practice in community-based arts projects and initiatives which impact on health and wellbeing. London: Health Development Agency. http://www.hda-online.org.uk/downloads/pdfs/arts%5Fmono.pdf.
Hetland, Lois. 2000. “Listening to Music Enhances Spatial-Temporal Reasoning: Evidence of the ‘Mozart Effect’.” Journal of Aesthetic Education 34:105-148.
Jackson, Ernest. 1979. “The impact of arts enrichment instruction on self-concept, attendance, motivation, and academic performance.” Pp. v, 135 leaves : p., forms. New York: Fordham University.
Kling, Robert W., Charles F. Revier, and Karin A. Sable. 2001. “Estimating the Public Good Value of Preserving a Local Historic Landmark: The role of non-substitutability and information in contingent valuation.” Fort Collins, CO: Colorado State University.
Krieger, Alex. 2001. “Community builders.” Architecture 90.
Laing, Dave, and Norton York. 2000. “The Value of Music in London.” Cultural Trends 10.
Landry, Charles, Lesley Greene, François Matarasso, and Franco Bianchini. 1996. The Art of Regeneration: Urban Renewal through Cultural Activity. Stroud: Comedia.
Lynch, Ruth Torkelson, and Deanne Chosa. 1996. “Group-oriented community-based expressive arts programming for individuals with disabilities: Participant satisfaction and perceptions of psychosocial impact.”
Matarasso, François. 1997. Use or ornament? : the social impact of participation in the Arts. Stroud, Glos.: Comedia.
—. 1999. Magic, Myths and Money : The Impact of the English National Ballet on Tour. Stroud: Comedia.
Matzke, Christine. 2000. “‘Healthy community arts, healthy communities’: Community Arts Conference, Exhibition Fair, and Festival.” Research in Drama Education 5:131 – 139.
McCarthy, Kevin. 2002. “Building an Understanding of the Benefits of Participation in the Arts.” Unpublished proposal submitted by the RAND Corporation to the Wallace-Reader’s Digest Funds.
Mitchell, Clare J. A. 1993. “Economic Impact of the Arts: Theatre Festivals in Small Ontario Communities.” Journal of Cultural Economics 17:55-67.
Murphy, Eileen M. 1995. “The impact of cultural organizations on urban revitalization projects as exemplified in the Pennsylvania Avenue Development Corporation.” Pp. iv, 85 leaves ; p., 29 cm. Washington D.C.: American University.
NALAA. 1994. Arts in the local economy: final report. Washington: National Assembly of Local Arts Agencies.
OAC. 2002. “Community Arts Organizations Program Guidelines.” Ontario Arts Council.
Ogilvie, Robert S. 2000. Community building : increasing participation and taking action : prepared for the 7th Street/McClymonds Neighborhood Initiative. Berkeley, Calif.: University of California Berkeley Institute of Urban and Regional Development.
O’Hagan, John W., and Christopher T. Duffy. 1987. The performing arts and the public purse: An economic analysis. Dublin: The Arts Council/An Chomhairle Ealaion.
Palmer, John P. 2002. “Bread and Circuses: The Local Benefits of Sports and Cultural Businesses.” C.D. Howe Institute Commentary 161.
Perryman, M Ray. 2001. “The Arts, culture, and the Texas economy: The catalyst for creativity and the incubator for progress.” Baylor Business Review 19:8-9.
Port Authority of New York and New Jersey, Cultural Assistance Center. 1983. “The Arts as an Industry: Their Economic Importance to the New York-New Jersey Metropolitan Region.” New York, NY: Port Authority of New York and New Jersey.
Preston, James C. 1983. “Patterns in Nongovernmental Community Action in Small Communities.” Journal of the Community Development Society 14:83-94.
Radich, Anthony J. (Ed.). 1987. Economic impact of the arts: a sourcebook. Denver, CO: National Conference of State Legislatures.
Randall, Paula, Dian Magie, and Christine E. Miller. 1997. Art works! : prevention programs for youth & communities. Rockville, MD: National Endowment for the Arts and the Center for Substance Abuse Prevention.
Remer, Jane. 1990. Changing schools through the arts : how to build on the power of an idea. New York: ACA Books.
Rolph, Stephen. 2001. “Impact of the Arts: A study of the social and economic impacts of the arts in Essex in 1999/2000.” Pp. 42. Chelmsford: Essex County Council.
Sable, Karin A., and Robert W. Kling. 2001. “The Double Public Good: A Conceptual Framework for “Shared Experience” Values Associated with Heritage Conservation.” Journal of Cultural Economics 25:77-89.
SATC. 1998. 1997 opera in the outback : economic and social impact study. Adelaide, S. Aust.: South Australian Tourism Commission Arts SA.
SCDCAC. 2001. “The arts and culture in San Diego : economic impact report, 2000.” San Diego, Calif.: City of San Diego Commission for Arts and Culture.
Seaman, Bruce A. 1997. “Arts Impact Studies: A Fashionable Excess.” Pp. pages 723-55 in Cultural economics: The arts, the heritage and the media industries. Volume 2. Elgar Reference Collection. International Library of Critical Writings in Economics, vol. 80. Cheltenham, U.K, edited by Ruth ed Towse. and Lyme, N.H.: Elgar; distributed by American International Distribution Corporation Williston.
Seham, Jenny C. 1997. “The effects on at-risk children of an in-school dance program.” Pp. iv, 101 leaves ; p., 29 cm.: Adelphi University.
Sharp, Caroline. 2001. “Developing Young Children’s Creativity through the Arts: What Does Research have to Offer?” Pp. 36. Slough: National Federation for Educational Research.
Singer, Molly. 2000. “Culture Works: Cultural Resources as economic development tools.” Public Management 82:11-16.
Stanziola, Javier. 1999. Arts, government and community revitalization. Ashgate: Aldershot, U.K.; Brookfield, Vt. and Sydney.
Stern, Mark J. 1999. “Is All the Worl Philadelphia? A multi-city study of arts and cultural organizations, diversity, and urban revitalization.” Philadelphia: University of Pennsylvania School of Social Work.
—. 2001. “Social Impact of the Arts Project: Summary of Findings.” Philadelphia: University of Pennsylvania School of Social Work.
Stern, Mark J, Laura Amrofel, Gina Dyer, and Alison Wok. 1994. “The Embeddedness of Community Cultural Institutions.” Philadelphia: University of Pennsylvania School of Social Work.
Stern, Mark J, and Susan C Seifert. 2000. “Cultural Participation and Communities: The Role of Individual and Neighborhood Effects.” Philadeliphia: University of Pennsylvania School of Social Work.
—. 2002. “Culture Builds Community Evaluation Summary Report.” Pp. 62. Philadelphia: University of Pennsylvania School of Social Work.
Strom, Elizabeth. 1999. “Let’s put on a show! Performing arts and urban revitalization in Newark, New Jersey.” Pp. 423-35 in Journal of Urban Affairs.
Thoits, Peggy A., and Lyndi N. Hewitt. 2001. “Volunteer Work and Well-Being.” Journal of Health and Social Behavior 42:115-131.
Throsby, David. 2001. Economics and Culture. Cambridge: Cambridge University Press.
Travers, Tony, Eleanor Stokes, and Mark Kleinmann. 1997. “The Arts & Cultural Industries in the London Economy.” London: London Arts Board.
Trent, Allen W. 2000. Community : a collaborative action research project in an arts impact elementary school. Columbus, Ohio: Ohio State University.
Turner, Francesca, and Peter Senior. 2000. A Powerful Force for Good: Culture, health and the arts- an anthology. Manchester: Manchester Metropolitan University.
Walesh, Kim and Doug Henton. 2001. “The Creative Community–Leveraging Creativity and Cultural Participation for Silicon Valley’s Economic and Civic Future.” San Jose, CA: Collaborative Economics.
Weitz, Judith. 1996. Coming up taller: arts and humanities programs for children and youth at risk. Washington, DC: President’s Committee on the Arts and the Humanities.
Williams, Deidre. 1995. Creating social capital : a study of the long-term benefits from community based arts funding. Adelaide, S. Aust.: Community Arts Network of South Australia.
—. 1997. How the arts measure up : Australian research into social impact. Stroud: Comedia.
Winner, Ellen, and Lois Hetland. 2000. “The Arts and Academic Improvement: What the Evidence Shows.” Special Issue of the Journal of Aesthetic Education 34.
Wollheim, Bruno. 2000. “Culture makes communities.” York, England: Joseph Rowntree Foundation.
Zukin, Sharon. 1989. Loft Living: Culture and Capital in Urban Change. New Brunswick, NJ: Rutgers University Press.