Eliciting the dynamics of leading a sustainable events: Key informant responses

Within the event management literature relating to network development and festival sustainability there is a paucity of research that analyzes the perception of festival sustainability by festival leaders. After an initial review of the context of sustainability, network theory, and an identification of the changing set of competencies for effective leadership, an exploratory and explanatory investigation is made to elicit and identify the critical factors that key informant festival leaders associate with sustainable festivals. The main purpose of this study is to attain a greater depth of understanding of festival leaders’ attitudes towards the dynamics of creating and directing sustainable festivals. In-depth interviews with five elite festival leaders helped to generate the elements of a repertory grid from which a “triading” method was used to elicit constructs. Of the constructs identified, the most significant relate to four areas: the event subject focus; the leadership; the funding; and the organizational culture. The research also revealed that festival leaders conceive sustainability not as an environmental concern, but as a matter of festival survival. Suggestions are then drawn as to the future role of the repertory grid method in identifying and managing stakeholder visions, and future lines of research investigation and application.


Introduction
Initially this article discusses the importance of leaders in determining festival success. It does so with reference to both sustainability as the context for festival management and with regards to emerging notions of creativity and innovation as leadership competencies. In reviewing related literature this article reports that current network theory, as applied to the wider event management research environment, does acknowledge festival leaders as significant stakeholders, but, on the other hand, does not give attention to an understanding of their view point. It is posited that in the context of festival sustainability a greater understanding of festival leader's perception of sustainability in their 316 ENSOR, ROBERTSON, AND ALI-KNIGHT own professional environment has significance for models of network management and their application. Identified thus as a gap in current festival and events research, an exploratory and explanatory investigation is made to elicit and identify the key factors that key informant (elite or expert) festival leaders relate to sustainability.
Personal construct theory (Kelly, 1955) is adopted as an appropriate methodological framework for this exploratory investigation. This framework stems from social representation theory (Moscovici, 1981), and thus the premise that the beliefs and perceptions of people affect the hierarchies and operations around them, and that an understanding of these beliefs and perceptions can aid communication and action. In-depth repertory grid (Rep Grid) interviews with elite festival leaders are employed as the most appropriate method of facilitating, forming, and recording the elements of the festival leader beliefs and perceptions of sustainability as they relate to their own professional environment. In analyzing the data, application of Honey's (1979) rating and categorization process offers an interpretative measure of the perceptions constructed. The graphic and statistical nature of the results, it is argued here, give the methodology particular resonance and versatility for further research as well as a more immediate capacity to direct communication and action more effectively in the social environment. Getz and Andersson (2008) state that the dimensions of sustainability for events and tourism are defined by their relationship with the natural resources; with community and political support; with economics and demand; and with operational management. It is posited elsewhere that it is the ability of festivals and events to create links between "endogenous resources and exogenous forces" (Quinn, 2006, p. 301) that will temper their sustainability. Both of these require as much of an appreciation and understanding of the ethical and personal dimension of sustainable development in the location in which is placed as it does a competency in technical or scientific application. To negate one would be ignoring the significance of the ethics of response in the context for which sustainability is sought (Hughes, 1995). As such, the advance of sustainability does not necessarily instigate or require an absolute measurement. It does, however, require opportunity to more fully understand its context. It is conceived here that for festivals a potentially key component of that context is event leadership.

Sustainability and Festival Leaders
While festival leaders have a very significant role in the direction and related success or failure of a festival, they receive only limited coverage in the related literature (Robertson, Rogers, & Leask, 2009). Getz (2002) cites the strength of the leadership of a festival as having a significant effect on the organizational culture and, therefore, influence the success or failure of a festival. In their discussion of national cultural influences on festival organization and processes for comparative analysis Getz, Andersson, and Carlsen (2010) refer to the work of Hofstede (1980) and cite the following influences on the management of events: power distance; uncertainty avoidance; individualism versus collectivism; masculinity versus femininity; longversus short-term orientation. The authors go on to observe that the organizational culture is either influenced by the leader or else influences the leader of a festival in their seeking "to impose or inculcate dominant values" (Getz et al., 2010, p. 34), concluding that this often culminates in points of conflict. In her analysis of the destabilizing effect of a charismatic new leader (Peter Sellers) on the 2002 Adelaide Festival, Caust (2004) echoes the work of Heifetz and Laurie (1997) and suggests that conflict may in fact act as a channel for creativity. For concision, this discussion is not developed here, other than to indicate that there is general agreement that conflict has within it the potential for a political process towards consensus and collaboration and thus creative output (Larson, 2009;Larson & Wikstrom, 2001).
However, the fact remains that in the area of festivals there is little research that focuses on the dynamics of leaders themselves. Related research focuses on the areas of ownership typology, stakeholder influence, and the lifecycle of the organization. Leadership is considered as a part of networking theory in event management (Stokes, 2006(Stokes, , 2007 and sustainable tourism events (Hede & Stokes, 2009). While each of these factors is important, ELICITING THE DYNAMICS OF LEADING A SUSTAINABLE EVENT 317 they do not gauge the perception of the director in their definition of, and their actions toward, the sustainable development of the festival which they lead.
The work of Getz and Frisby (1988), Getz (2002), Gursory, Kim, andUsyal (2004), Williams and Bowden (2007), and  are identified as relatively rare examples of research that investigate the relationship between festivals leadership and festival sustainability. Again, however, in each case the nature of the research focus does not allow for further insight into the understanding as to why the leader relates to the needs of ensuring sustainable development of an event. The authors of this work have identified this as an important gap in the field of festival and event management knowledge and theory and, furthermore, volunteer that this information may offer a better response to the more frequently recorded failure or weakness of a festival leader in ensuring network development. It could address then what will now be discussed as emerging leadership competencies. Getz and Andersson (2008) believe that most festivals are formal organizations which require competencies of management, adaptability, and an ability to ensure financial and political support to ensure sustainability. Similarly DiLiello and Houghton (2008) suggest that in the current business literature there is general agreement that individual innovation is essential in the facilitation of organizational innovation and "long term organizational success and survival" (p. 37), is likely to induce efficiencies and effectiveness as an element of uniqueness. Getz (2009) talks of festivals or event organizations as requiring a "new sustainable and responsible events paradigm" (p. 61) in which the triple bottom line approach is applied "both to the determination of the worth of events and to evaluation of their impacts" (p. 62). There is agreement that the role of a festival leader is one which requires managerial adroitness in areas which have emerged more quickly than the skills historically attributed to the direction of festivals. In much the same way as creativity and innovation are seen as competencies in the organizational structure and management of business today (Amabile, Conti, Coon, Lazenby, & Herron, 1996;DiLiello & Houghton, 2008;Ensor, Pirrie, & Band, 2006), then, so too they can be seen as vital in the management of festivals and events. Importantly creativity and innovation are different elements of a process, in which creativity is the generation of ideas and innovation is the process of their application (Ensor, Cottam, & Band, 2001). As Caust (2004) proffers, the leader of an arts festival is required to engage in their task as both managerial visionary (creator) and as negotiator and respondent to the requirements of an increasing numbers of external forces (innovator). In a rapidly changing environment, stakeholder networks have the potential to offer the elastic properties required to support and adapt to that visionary process.

Networking
There is an increasing body of literature and investigation in the role of-and the current restrictions on-networks as necessary management constructions for festival success (Hede & Stokes, 2009;Larson, 2009;Mackellar, 2006;I. Stewart & Lacassagne, 2005;Stokes, 2006Stokes, , 2007. The authors agree that networking can aid the formation of a relationship akin to the formation of an institution (Getz, 2009). Vital "knowledge and information is often embedded in social networks, which become important components in facilitating individual creativity" (DiLiello & Houghton, 2008, p. 38). This can contribute towards sustained support and, like a strong brand, inspire confidence and encourage creative and innovative progression-individually and for the collective needs of networked stakeholders. As such, a festival (or series of related festivals or events) can enable intraregional community collaboration and offer a dynamic that one person, or a group representing one interest, is unlikely to be able to match. The forms and structures of networks can, of course, create dynamics which are less positive. The levels of knowledge and respective hierarchies (preexisting and formed) incumbent within a network can have huge effects on decision making, plans, and the identity of the festival networks that emerge (Hede & Jago, 2005;Hede & Stokes, 2009). While consensus and collaboration may emerge, there are cases where the interests of various stakeholder, (e.g., arts, tourism, and community organizations), have conflicting organizational boundaries and resource objectives which delay or negate positive resolution (Carlsen, Ali-Knight, & Robertson, 2007;Long, 2000).
Although the coverage of the related literature is not comprehensive here, a previous more in-depth review indicates to the authors that the special nature of festival leadership, and the need for a personal "vision" (Caust, 2004) is given considerably less coverage than the repeated exploration of what is, or should determine, what a festival leader should be within a network. The theory of charismatic leadership advocates that there are forms of leadership in which there is a unique bond or emotional relationship between the leaders and those lower in an accepted hierarchy. These bonds are intrinsic to its success (Galvin, Balkundi, & Waldman, 2010). The organizational relationships formed in these bonds are necessarily reinforced by continued communication and interrelation with the leader (Galvin et al., 2010;Lengel & Daft, 1989;Shamir, House, & Arthur, 1993). As a positive relationship with the network is vital to the sustainability of the festival (Getz & Andersson, 2008;Hede & Stokes, 2009), so too an understanding of the perspective of the leader of an organization (a stakeholder) has a pivotal position in maximizing knowledge of the issues that as part of a network needs to be considered. Without this knowledge the long term interrelation of network members is challenged. While this is acknowledged in the wider parameters of network theory literature, there is considerably less in respect of festivals. In particular, there is a dearth of understanding as to how the proximity of a festival to, and the relationship with, a festival leader may have a fundamental effect on the viability of networked relations required for its survival. The authors undertake this exploratory research as part of a pathway towards this greater understanding.

Social Representation as Research Paradigm
Introduced in the work of Moscovici (1981) social representation is fundamentally what people think they know about a subject or situation (Gjerald & Øgaard, 2010; I. Stewart & Lacassagne, 2005) and the hierarchical relationship that will stem from this. This is to say that it is the structured judgments to which a social group identifies. As daily constructions they guide communication and behavior (Moscovici, 1981;Penz, 2006). Deery and Jago (2010) recognize it as one of three theories underpinning research in the area of social impacts of events. They identify Pearce, Moscardo, and Ross's (1996) text Tourism Community Relationships as introducing this research position to tourism literature and indicating its further role in the context of event related research. In acknowledging that people construct their own interpretations and values, qualitative research techniques such as projective procedure, elicitation activity, free association, and image associations are highly relevant methods for capturing this information (Penz, 2006;I. Stewart & Lacassagne, 2005). Repertory grid (Kelly, 1955), as discussed below, is one technique that can be used to record and understand the social realities of any given group.
Repertory grids were thus employed to allow the identification of key constructs that the sample of festival leaders found within their own professional environment. Repertory grid technique is an established psychological tool that has been used for over 40 years and has an extensive range of business application (Easterby-Smith, Thorpe, & Holman, 1996; V. Stewart & Stewart, 1981). The repertory grid technique has its foundations in the Personal Construct Theory (PCT) developed by Kelly (1955) and is described as serving to discover an interviewees' construct by way of conversation (Fransella & Bannister, 1977;Jankowicz, 1995Jankowicz, , 2003Pike, 2003;Selby, 2004). While originally developed for application to single respondents, it has great flexibility and is not necessarily limited by any particular sample size (Pike, 2003). By involving more than one person, the process serves to provide recurring constructs of knowledge based on consensus (Selby, 2004). Whether applied to individuals or larger sample sizes, it is important that the respondents are knowledgeable about the domain being investigated (Kelly, 1955). Where this can be confirmed, say when referring to recognizable communities of interest, it has the capacity to develop a perceptual taxonomy for those communities of interest.
Within the field of tourism research, this repertory grid process has been applied for evaluation of destination image in the minds of potential and im-mediate consumers (Coshall, 2000) and to illicit resident's perceptions of tourist attractions (Lawton, 2005). Pike (2003) concluded that the techniques are likely to offer constructs valuable to the decision-making process within an organization (or destination). As observed earlier, Gjerald and Øgaard (2010) apply repertory grid in a hospitality industry context to elicit an understanding of significant stakeholders-coworkers, customers (guests), and competitors-in a particular social environment.
However, in the analysis of arts and festival provision, the use of repertory grid analysis is less common. Where it has been applied, it is more commonly associated with ascertaining market knowledge. Jansen-Verbeke and van Rekom (1996) used repertory grid analysis to identify the motivation of visitors to museums (in Rotterdam) for the purpose of more effective marketing of the urban tourism product. They conclude that the motivation construct for visiting museums is convergent with the needs of visiting a city, and thus not distinct from it. Caldwell and Coshall (2002) used repertory grid analysis to measure brand association of museums and galleries. They found that there were a relatively small number of constructs, concluding that efforts by museums and art galleries to create brand associations have not been effective with customers failing "to differentiate between them on the basis of anything but functional benefits" (p. 390). Canning and Holmes (2006) used repertory grid as a consultation devise with selected community groups in Sheffield to evaluate the best form of museum and art gallery provision. Importantly, their work highlights both the robust nature of repertory grid data and the pragmatic value to both community and government policy makers in offering meaningful data, which bridges the government's desire for quantitative data, while dealing with issues suited to qualitative methodologies.
In one of a very few applications to events and festivals, Wooten and Norman (2009) used repertory grids to evaluate the motivation of tourism and festival goers. Applied to the Kentuck Festival of Arts, in the US, the findings indicate commonalities in the perception of the event and the promotional material depictions of it. However, as repertory grids have a long running relationship with the theory of social representation as part of a market-ing application (Stewart & Lacassagne, 2005)where attempts are made to elicit what people believe they know about social situations or objectsso too they have application for the sociocultural context (Penz, 2006).

Methodology
In-depth interviews with five festival leaders formed the basis of this exploratory research, which aims to elicit and identify the key factors that they as festival leaders perceive as characterizing sustainable festivals. The interviews were of an hour's duration. This was followed by a separate exercise in which the interviewees had to score the repertory grid that had been drawn up as a result of the initial interview. This research is essentially exploratory and explanatory, and Marshall and Rossman (1995) contend that for this type of study elite in-depth interviewing is suitable. Sarantakos states that this type of interview "involves elites, that is, wellknown personalities, prominent and influential people, as respondents. It therefore aims to collect information that is exclusive and unique to these informants" (Sarantakos, 1998, p. 255).
Sarantakos goes on to contend that for this very reason elite interviewing is a very useful tool for data collection. Consistent with this view, the group of respondents was chosen to provide insights from a number of perspectives based on operational leadership roles within a festival. These were at a strategic administrative and political level alongside a city, national, and international perspective. All the respondents had in-depth knowledge of a range of festivals, which extended beyond their own individual festivals, Edinburgh or the UK, to around the world. The respondents included individuals currently the directors of two large festivals that occur at different times of the year in Edinburgh. Both of these leaders had a well established leadership history in their respective festivals as well as previous experience in other festivals. Two of the other interviewees had responsibility for the strategic development of a range of art festivals and events at a UK national level as well as at a strategic level within the City of Edinburgh. The final interviewee was a former festival director with experience of leading an event outside the UK, as well as, having an international reputation in the 320 ENSOR, ROBERTSON, AND ALI-KNIGHT area of event policy, operations, and management. This individual also had in-depth knowledge of festival leadership in Edinburgh and the UK.
There were two stages to this study. Stage 1 involved eliciting the elements of the Rep Grid and Stage 2 involved eliciting the constructs.

Eliciting Elements
According to Easterby-Smith et al. (1996) there are several ways of generating the elements of a Rep Grid. They can by created by the interviewer supplying them, the interviewer supplying situations or descriptions, a pool of elements is created by the interviewee, or they are elicited through discussion. In this research study the interviewees were asked, prior to the interview, to identify five festivals of which they had detailed knowledge. At the beginning of the interview both parties discussed potential festivals and through this discussion a list of five specific elements was drawn up. Each of the interviewees, therefore, had a specific and unique set of elements. These were used as the elements of the repertory grid.

Eliciting Constructs
In generating constructs the "triading" method was used (Coshall, 2000;Pike, 2003;Selby, 2004). Each element was given a number. Then using five numbered cards, groups of three elements were drawn at random. The interviewee was then asked to consider ways in which two elements were alike but at the same time opposite or different to the third element. The researcher sought to clarify and probe the meaning of each construct elicited by the interviewee by posing questions such as 'What do you mean by?'. Also the interviewee was asked to compare constructs with something else that the researcher was fairly sure the interviewee didn't mean. Laddering was also applied where the researcher enquires as to why a particular construct was important. The researcher tried to ensure that the recorded constructs were not impermeable or vague, as according to Easterby-Smith et al. (1996) constructs with these features can cause problems in analysis and interpretation.
Each interview took at least an hour. A follow up session then took place, where each respondent rated the constructs on the repertory grid that the researcher had drawn up as a result of the initial interview. Throughout the interviews the researcher conscientiously followed the advice advocated by Easterby-Smith et al., (1996) when they state that, "The interview itself can be seen as a conversation in which both parties are seeking to explore the interviewees, not researchers, understanding. . . . The researcher first and foremost must listen/ then probe, clarify and enquire" (p .9).

Method of Analysis
V. Stewart and Stewart (1981) state that there are five approaches to analyzing the data provided through a Rep Grid interview. These are: frequency counts, content analysis, visual focusing, cluster analysis, and principal components analysis. In more recent literature the emphasis is heavily on sophisticated statistical approaches using computer packages to analyze the results of the Rep Grid. Yet there are criticisms of highly statistical analysis. Although Burr and Butt (1992) claim that both ranking and ratings of constructs against elements require respondents to make very fine classifications. They go on to stress that "Subtle nuances of meaning can be extracted from a matrix of figures that can be justified mathematically but not really psychologically. The delicacy and sophistication of the analysis can blind us to the relative coarseness of the data" (p. 123).
More recently there has been support for employing a more qualitative approach to the analysis. Gammack and Stephens (1994) claim that a repertory grid analysis "may take either or both of two forms: a statistical analysis involving the mathematical properties of the grid itself; and/or a more interpretative analysis involving the constructs and their labels and how the constructs abstract each element" (p. 77).
Given that this initial piece of exploratory research involved only five respondents, it would seem appropriate that an interpretive rather than a highly statistical analysis was undertaken on the interviews. On this basis a content analysis of the interviews would be the technique to employ. Jankowicz (2003) states that the problem with many approaches to content analysis is that while they allow the researcher to say something about the sample as a whole, the technique can lose the specific meanings of individual interviewees. The particular ratings individuals applied on their repertory grid become redundant. However, the technique developed by Honey (1979) combine different constructs across a sample while at the same time allowing the researcher to make use of individual meanings expressed through the rating given on any single repertory grid. Jankowicz (2003) characterizes this form of content analysis as being a technique that "assumes that what we're interested in is each individual's personal understanding of the topic in question, and treats each construct offered by the individual as more closely related, or less closely related to the overall issues he has in mind when thinking about the topic" (p. 170).
Although respondents chose different elements (festivals), they were asked to rate how sustainable/ unsustainable each event was from their perspective. All the other constructs they rated on their repertory grid could then be analyzed in terms of how close or removed it was from the rating they gave this construct, which Jankowicz (2003) calls the "overall summary" construct. In order to do this, each construct is labeled with two indices that have been calculated. The first index is the percentage similarity score. A score of 100% indicates that the rating on that particular construct match exactly the ratings on the overall summary construct. A rating of 50% would indicate that the ratings were substantially different. Recognizing that similarity scores are relative to Honey's procedure then requires the researcher to note whether the similarity score on a construct is in the individual's highest scoring third of constructs, the inter-mediate third or the lowest third. Thus, high, medium, or low (H-M-L) values are allocated to each construct creating a second index. Accordingly, each construct by this stage then has a percentage similarity (SIM) by this stage and an H-M-L value as well as its own reference code.
Constructs are then allocated to categories as in a generic approach to content analysis. Each construct is compared with the other. Constructs that are the same in some way are placed together under a single category. Constructs that are different from existing categories are placed separately and start the formation of a new category. This process continues until all constructs have been classified.
The Honey (1979) procedure allows constructs to be placed within a category in order with those closest to the summary "M" construct at the top of the list. The H-M-L scores also allow the researcher to establish exactly how important individuals rated the constructs within a particular category. In carrying out this procedure, the percentage similarity score has to be calculated for the construct in both its positive and reversed relationship with the summary construct. The percentage similarity score that is closest, whether positive or reversed, is the score that is used when attaching the index score and the H-M-L value. This methodology was appropriate to the needs of this research.

Results and Analysis
The findings from the research are summarized in Table 1. Each category is shown in order of rank. Construct categories that were ambiguous or contradictory have not been included in the findings. The percentage total for each construct category (from all constructs) is shown in the fourth column. This adds to 42% of all constructs. Each of these construct categories is then discussed (Table 1).
One major feature that emerges from the repertory grid analysis is that the respondents appear to have interpreted sustainability in terms of a festivals ability to survive and not in terms of a festivals ability to manage its environmental impact. The environment formed only 3% of the total constructs.

Focus
Seventeen percent of all the constructs identified through the repertory grid interviews fell into this category. Of those constructs, 17% rated high on the individual's similarity score with a further 19% rated as intermediate. There was no fixed image on 322 ENSOR, ROBERTSON, AND ALI-KNIGHT the type of festival that was sustainable, constructs ranged across a broad spectrum on the potential focus of a festival including those whose focus was on ideas, special interest, cutting edge, lifestyle, a particular industry, or tourism. One key aspect that emerged, however, was the importance of having a focus on popular culture or arts. This was also referred to in terms of the festival being entertainment driven (Table 2).

Leadership
Thirteen percent of all the constructs identified through the repertory grid interviews fell into this category. Of those constructs, 36% rated high on the individual's similarity score with a further 41% rated as intermediate. In several areas these con-structs were contradictory. Two themes that did emerge were that sustainable festivals were often led by second or third generation leadership. However, it is difficult to know whether second generation leadership is a result of being a sustainable festival or a contribution to the festival being sustainable. The second theme identified was that of the importance to sustainable events of a festival having a leadership that had expert standing in the sector due to previous knowledge, experience, and reputation (Table 3).

Funding
Eight percent of all the constructs identified through the repertory grid interviews fell into this category. Of those constructs, 54% rated high on the individual's similarity score with a further 31% rated as intermediate. The clear perspective that emerged was that sustainable festivals did not rely wholly on public funding but instead found funding from mixed sources. Festivals that were highly ticket driven and that had commercial funding through sponsorship appeared to be more sustainable than those which relied solely on public sector backing. There was also a subsidiary comment, however, suggesting that sustainable festivals should not be overly reliant on commercial funding which implies that a mixed funding regime is beneficial and offers greater longevity (Table 4).

Culture
Four percent of all the constructs identified through the repertory grid interviews fell into this category. Of those constructs, 29% rated high on the individual's similarity score with a further 42% rated as intermediate. The clear theme that emerged was that sustainable festivals had an establishment culture based on a strong institutional memory and a paternalistic orientation. Obviously, only a festival with a strong history and legacy sustained over a long period of time can fall back on a strong institutional memory. However, it is unclear as to whether an establishment culture and paternalistic orientation are integral to festival longevity or emerge as a result of that longevity (Table 5).

Discussion and Conclusions
In responding to the gap in knowledge and understanding of festival leaders and their relation to notions of sustainability in their professional area, four key categories emerged from the constructs identified in the research. These are focus, funding, leadership, and culture. As stated, the elicitations indicate leaders conceive sustainability as a matter of festival survival. In reviewing the constructs, and their scaling, in the category of focus it is revealing to note that while there is some evidence to suggest sustainability is broadly perceived as relating to festival visitor type and event purpose, and while these may indeed affect the capacity of festival leaders to impose their authority on the event (Getz, 2010), the range of constructs also demonstrate the great number of influences on this. The second largest construct, leadership, does highlight the esteem and significance that is attributed to the individual (and individuality of) the festival leader and his or her knowledge, and the far lower significance attributed to shared or group leadership. In attempting to engage festival leaders in networks constructions, this perception is an important one.
It is only in the construct relating to funding that recognition of the multiple or networked nature of operations is given priority. The exploratory nature of the research exercise here does not however, allow evaluation of the extent to which this interaction is viewed as collaborative and not simply as  financially necessary. Similarly, the high values ascribed to history and legacy in the construct emerging from culture cannot be equivocally viewed as indicative of an aversion to innovation and creativity. It is more likely that this emerges from a dislike of financial risk. Additional construct categories: location, audience, community relationship, origin, organic growth, local agencies, decision making, environment, and performers were not substantial enough to be read as significant. The limited priority of the physical environment as a factor of sustainable events has been noted as significant. This finding supports the result of  whose survey of 60 directors in the UK showed that only 18 of the 60 perceived the impact of their festival on the environment as significant. In their findings these respondents (representing 30% of the total number surveyed) consisting of the leaders and leaders of mixed art events, and also those managing festivals with a history of 17 years or more, gave this impact the lowest rating. At one level the view of the leaders can be seen as an act of pragmatic judgment. However, the results of a related analysis by  showed this dissociation with the physical environment by festival leaders to be at odds with the perception of their audiences. The impacts of festivals on crime, the natural environment, and traffic/parking all emerged as strong factors in the perception of the 423 festival visitors surveyed. So it may be as Getz and Andersson (2008) suggested in their analysis of music events in Sweden, that the sustainability of a festival may often resolve less around the issues relating to the physical environment and more to those of a management "competence, adaptability, and success in assuring continuing political support and resources" (p. 1). In their research of livemusic festivals in Sweden, they conclude that sustainability can be formed by the growth of a festival in the minds and hearts of the audience. However, the authors would suggest that adaptability will also have to engage in the minds and sentiments of the audience as political agents. Accordingly, the significance of the "green issue" should resonate in the mind of leaders whose role has been noted as being charismatic, and for whom communication and support is crucial, and as being also a conduit of network knowledge and adaptability.

Limitations of the Research
The research is exploratory and the social representation paradigm has been used in its application. The results do not test hypothesis and do not posit models of investigation. The research is also limited by geographical scope and does not reveal the cultural dissonances in the international festival and event field. In terms of the sample of festival leaders there is also limitation in terms of gender and political influence with most of the leaders originating from western festivals that receive strong public support and subsidies.

Future Lines of Investigation and Application
From this investigatory analysis it is proposed that repertory grids are significant instruments in understanding stakeholder dialogue, i.e. as an important tool to build consensus between stakeholders. Accordingly, repertory grid methodology has future application in identifying the salient visions of festival stakeholders. This in turn may aid the establishment of clusters of agreement. They have potential, thereafter, to facilitate the different individual and group needs as suggested in the Hede's (2007) model of the triple bottom line concept (Fredline, Raybould, Jago, & Deery, 2005).
Future research should focus on a wider number of festival leaders to expand the international, political, and gender dimension of the work. A broader spectrum of festivals and events should also be examined, across different genres. The research could also be extended to other key stakeholder groups such as government funding bodies and community consultation groups. This would enable a more holistic and balanced view of the implications of sustainability to those working within the festival and event arena.
However, the work presented here offers a valuable contribution to the understanding, and limited research available, of the key factors that lead to sustainable events. Through the use of the repertory grid technique the key constructs identified by the festival leaders (focus, leadership, funding, and culture) offer a pragmatic value to the festival community and government policy makers when devising and planning for future sustainable events. There is a need for more research to influence and ensure effective and appropriate event policy. In the Australian context, Whitford (2009) worries that event policy developed by governments is often ad hoc or vague. Accordingly she emphasizes the need for continued research in this area. Similarly, Getz (2009) highlights the need to assess how the prevailing form and structure of policy is in itself legitimizing its lawfulness, and to query the process of institutionalization which has affected this legitimization. This, he states, is required in a new paradigm of sustainability.