Knowledge Sharing and Power in the Event Workforce

The aim is twofold: First, we draw on organizational behavior, strategic management, and events literature, to conceptualize the event workforce as highly functional and contract-based heterogeneous, the event organization as temporary and project-based structure with high fluctuation of paid employees and volunteers. Second, we contextualize these unique characteristics to understand knowledge sharing behavior of the event workforce. We analyze interview data from volunteers and paid employees at tourism events to make two theoretical contributions: 1). We advance theory in events literature, showing that event organizations are unique in terms of their processes and team composition. 2). We contribute with novel insights of how knowledge is used in heterogeneous event teams, explaining how knowledge serves as a source of power for both volunteers and paid event managers alike.


Introduction
Despite the growing interest in event studies, little is understood about the unique organizational characteristics of events (Liu, 2018). The lack of research on organizational aspects and the unique managerial context of events is surprising despite the growing awareness of the economic importance of events, the rising overall impacts of the events sector (Mair, 2009) and the increased necessity to design and implement legacy governance structures (Sharp & Finkel, 2018). A deeper understanding of the event organizational context is also needed as there is growing recognition of event management as a profession. Event organizations are advancing towards higher levels of professionalization (Stadler, Fullagar, & Reid, 2014) and event researchers are calling for more studies to contribute to an understanding of the unique organizational context and human resources management of event organizations (Junek, Lockstone, & Mair, 2009). The problem is, however, that there remains an evident and continuing research gap in the strategic and organizational aspects of event management research (Getz, 2008;Liu, 2018;Mair & Whitford, 2013) and as Lockstone-Binney (2018, p. 1049) recently notes, in event management research some "research gaps remain persistently stubborn, while popular topics of study continue to attract focus".
One specific area that has been neglected is an understanding of how event organizers manage knowledge within their workforce (Liu, 2018). We argue that understanding knowledge transfer processes and related workplace behaviors is essential for event organizers, particularly as "the life-blood of most organizations is knowledge" (Mabey & Zaho, 2017, p. 39). The extant literature in the domains of organizational behavior and strategic management has exhaustively confirmed that effective knowledge sharing increases employees' positive work attitudes, and organizational performance (Bock, Zmud, Kim, & Lee, 2005;Lee, Gillespie, Mann, & Wearing, 2010;Szulanski, 2000). Additionally, "over the past three decades, scholars have increasingly come to view knowledge as one of the most important resources necessary for successful organizations in the contemporary socioeconomic landscape" (Barley, Treem, & Kuhn, 2018, p.278). But, it is known that stimulating employees to share their knowledge has proven to be difficult due to a range of individual factors; e.g., for employees, power and status influence the willingness to share knowledge (Elias, 2008;French & Raven, 1959).
Institutional factors and organizational culture and climate also influence motivations to share knowledge, e.g. volunteers have been found to be particularly motivated to share their Electronic copy available at: https://ssrn.com/abstract=3487337 knowledge when they perceive themselves to be in an autonomy-supportive environment (Allen & Bartle, 2014).
The problem with viewing knowledge as an asset of power within organizations is that if an individual perceives knowledge to be a source of power and influence, and decides to hold on to this asset, it might lead to competing interests between employees and the organization.
Another problem for organizations is that too much formal effort and planned knowledgesharing activities can also lead to ineffective outcomes (Mabey & Zhao, 2017). Further, the specific event context, e.g. the heterogeneous team composition, and temporary, and 'pulsating' nature of events, influences the way employees and volunteers share knowledge. Hence, we argue that existing frameworks and assumptions of knowledge sharing need to be contextualized to the event context.
The aim of this study is to explore the unique processes and team composition of events and examine how these influence the internal knowledge sharing behavior of event workforces. To do this, we integrate organizational behavior, strategic management, and events literature and unpack the views of volunteers, casual and paid employees. Our findings make two major contributions to the tourism and events literature. First, we show that event organizations are distinct in their processes, and team composition, and by doing this, we are addressing in part the significant shortfall in strategic and organizational event literature (Getz & Page, 2016;Liu, 2018). Second, we advance understanding of how knowledge is used in event organizations. In showing that knowledge is used as a means of power by both volunteers and paid permanent event managers, we highlight theoretically and practically that management in events requires a detailed contextual understanding of each event's characteristics.

Event workforce and organizational structures
Events have a number of unique characteristics that define the profile of the event workforce.
Heterogeneity arises from the diverse nature of contract forms, e.g. long-term permanent employees, short-term and long-term volunteers, as well as casual employees. Further, external contractors such as security and catering employees form part of the extended event workforce (Mair, 2009). Larger event organizations may be able to maintain a core full-time team, but smaller and one-off events may have to outsource many of these functions to external contractors (Mair, 2009).
Additionally, volunteers are critical for the success of the event, and have different needs than paid permanent employees. Aisbett & Hoye (2014) show that volunteers offer time and expertise to engage in the event. However, research has shown that event organizers are often less considerate when it comes to understanding the needs and offering organizational support for volunteers, resulting in weaker organization-employee relationships. This lack in understanding and managing volunteers can lead to high turnover rates and less commitment when compared with permanent employees (Aisbett & Hoye, 2014). Bang, Won, and Kim (2009, p. 69) noted that "to sustain a volunteer workforce, building a sense of motivation, commitment, and intention to continue volunteering is one of the most important tasks of event organizations and managers". This complexity gives rise to particular challenges for human resource management, and for knowledge sharing in event teams.
Second, the functional and contract-based heterogeneity is a key characteristic of an event's team and is likely to impact on event management. Permanent employees will have rather longterm career goals, and might find it important to demonstrate organizational commitment to proceed in their careers, but casual employees might prefer to be short-term oriented and flexible. In contrast, volunteers might seek a purely recreational experience (Mojza et al., 2010). Volunteers show higher levels of altruism then permanent employees (Cnaan & Goldberg-Glenn, 1990). Hence, permanent employees might be more competitive, and less Electronic copy available at: https://ssrn.com/abstract=3487337 willing to share their knowledge. Yet, volunteers are not a homogenous group and they vary in terms of their motivation to participate in events (Holmes et al., 2018;Treuren, 2014). Edwards (2005) proposes that volunteer motives might vary in according to their needs of selfexpression; social needs; and personal interest amongst others. Bang et al. (2009) point out that extrinsic rewards and community involvement are key motivators for volunteers to participate in events. All of these tensions and differences make it likely that knowledge sharing in heterogeneous event teams will present unique challenges.
Third, events usually operate within a temporary, project-based structure where high speed and clear goals are key performance indicators (Bartsch et al., 2013;Hobday, 2000). Further, temporary, project-based organizations are usually more effective in managing complexity, innovation, and dealing with ambiguity-but less effective in managing routine tasks, and in generating economies of scale (Hobday, 2000). Importantly, the need to share knowledge for project-based firms is high. Each project is unique, and new information needs to be acquired permanently (Keegan and Turner, 2002). However, the problem is that they keep knowledge inside the team and rarely use knowledge strategically (Bartsch et al., 2013), perhaps as such temporary organizations often exist for one single project and are less long-term oriented.
Fourth, another key characteristic of events is the high fluctuation of employees. The notion of the 'pulsating organization' suggests the change between expansion and contraction of employees' numbers (Toffler, 1990), ranging from a small team of permanent employees before the event to a full complement of employees during the event. In the immediate run-up to, and during the event operations, there is a significant increase of employees, followed by a sudden decline after completion of the event operations, returning to the original core employees. According to Hanlon and Jago (2004), some event organizations transform their structure almost overnight, increasing their employees by up to 1000% for the event period, then contracting back to their original size within a week. Event organizations, therefore, share the characteristics of such pulsating organizations (Hanlon & Cuskelly, 2002). Event organizations are thus quite different to permanent organizations, which arguably have a more stable workforce, long-term orientation, and well-established relationships between employees and management.
In addition to this unique workforce and organizational structure profile, working at an event provides a number of unique challenges that employees, volunteers and their managers have to tackle. For example, on an operational level the high pace of events, ongoing recruitment, difficulty around employee retention and team cohesion due to the constant fluctuations of employees are major challenges; on a strategic level, storage of knowledge, particularly for one-off or single pulse events, adds to the complexities (Holmes et al., 2015;Stadler et al., 2014). In spite of this, knowledge is often kept inside the team and not passed on for future use nor evaluated after the experience of its use (Bartsch et al., 2013;Parent & MacIntosh, 2013).
Research has confirmed that integrating knowledge management into short-term oriented firms can be problematic (Thiry & Deguire, 2007). Presumably, problems in knowledge management arise because of a lack of time, when employees work intensively together, yet only for a limited duration, merely focusing on operations (e.g. Parent & MacIntosh, 2013).
Thus, we conclude that event organizations are complex organizational structures that operate at high speed in usually temporary and project-based structures, with a high functional and contract-based heterogeneous and highly fluctuating workforce. Yet, as a consequence of the strong emphasis on short-term operations, event organizations are less strategic and thus are less likely to be long-term oriented -and subsequently less strategic in terms of their knowledge management.

Knowledge sharing
Knowledge sharing is the key to the success of any knowledge management activities (Szulanski, 1996;Wang, & Noe, 2010). Effective knowledge sharing increases employees' positive attitudes and organizational performance (Bock, Zmud, Kim, & Lee, 2005;Lee, Gillespie, Mann, & Wearing, 2010;Szulanski, 2000). Nevertheless, it is known that stimulating employees to share their knowledge has proven to be difficult due to a range of individual and institutional barriers, especially when employees perceive that the costs of sharing knowledge outweigh the organization's rewards (Bock et al. 2005;Cabrera, & Cabrera, 2002;Szulanski, 1996). In this paper, we use the term knowledge sharing but acknowledge that some authors use knowledge sharing and knowledge transfer interchangeably (e.g. Wang & Noe, 2010). We align with Tangaraja et al.'s (2016) view that knowledge sharing is an entirely behavioral concept, constructed through individual actions; in contrast, knowledge transfer also includes non-behavioral, processual steps.
Electronic copy available at: https://ssrn.com/abstract=3487337 There is an ongoing debate on why some organizations manage knowledge sharing well, while others fail (Røvik, 2016). Bock et al. (2005, 82) note that "individuals' knowledge does not transform easily into organizational knowledge even with the implementation of knowledge repositories". It is clear that employees have specific individual roles which either foster sharing or prevent them from sharing their knowledge. Yet, it remains difficult to predict what drives the sharing of knowledge on an individual level, as there are both individual factors and organizational-contextual causes influencing individuals' willingness to share or not to share their knowledge (Cabrera, & Cabrera, 2002;Wang, & Noe, 2010).
A theoretical lens of power can help to elucidate the difficulties of knowledge sharing, as knowledge and power are closely interlinked (Elias, 2008;Foucault, 1980;French and Raven, 1959). Knowledge can be regarded as a source of power (French & Raven, 1959). Along with the employees' education, expertise and status (Davenport & Prusak, 1998), knowledge is the employee's asset -and means to influence -within the organization. Power theory can explain some of the problems that occur with knowledge sharing, where some employees perceive knowledge as a form of 'asset' that is of great value and needs to be controlled and monitored (Heizmann & Olsson, 2015).
Hence, the question for the individual arises: Should this source of power be shared? As a consequence of transferring knowledge, the individual's perceived source of power within the organization -and to an extent their individual competitiveness, might be lost (Bock et al., 2005). Subsequently, there are competing interests between maintaining personal knowledge as a source of power on the one hand, and the need for the organization to increase their knowledge on the other. Within the heterogeneous event workforce, permanent employees, casual employees and volunteers are all likely to have different views on making use of knowledge as a source of power; especially as the three groups can be considered to have different resources and levels of power available (e.g. due to their positions and contract forms). Allen and Bartle (2014), for example, found that volunteers, similar to permanent employees, feel motivated to engage, share knowledge and provide feedback when they perceived positive leadership and being stimulated by an autonomy supportive environment.
From an organizational perspective, Baskerville & Dulipovici (2006, p. 2) even point to a potential ethical issue that can arise if organizations force their employees to share their Electronic copy available at: https://ssrn.com/abstract=3487337 personal knowledge: "As an individual's private attribute, organizational programs that aim to forcibly develop knowledge-sharing cultures could violate individual privacy rights". As a result of this discrepancy there often is little willingness to share knowledge, resulting in 'knowledge hoarding', where "individuals may feel they may be disadvantaged for having obtained knowledge that, if shared, could threaten their employment, i.e. whistle blowing" (Heizmann & Olson, 2015, p. 835). Surprisingly, and despite the omnipresence of power in knowledge processes (Foucault, 1980), there is little research providing a deeper understanding between the co-occurrence of links between knowledge sharing and power in organizations (Heizmann & Olson, 2015).
In summary, drawing on organizational behavior, strategic management, and events literature, the aim of this paper is to unpack how knowledge sharing behavior unfolds in event teams.
Bearing in mind the unique characteristics of event organizations and the issue of power relations, we explore motives to share (or not to share) knowledge of volunteers, casual and permanent employees.

Research Approach
The study takes a qualitative approach, to explore how these unique processes and team composition influence knowledge sharing in event organizations. We adopt an underlying social-constructionist ontology acknowledging the multiple meanings that are constructed as people engage and form relationships with the world around them (Crotty, 1998). Data were collected through nine in-depth semi-structured interviews, enabling participants to express their judgment in a guided and interactive manner, and allowing the researcher to identify and probe further into responses and participants to clarify and explain themselves (Zikmund, 2003). The interview guide was developed based on the literature review which had previously been conducted. Questions were designed to uncover how, why and what knowledge participants shared, along with a discussion of challenges or barriers to knowledge sharing in event teams. In order to ensure the quality of the data to be collected, the interview guide was pre-tested for ease of comprehension and for content, and minor adjustments were made.
Electronic copy available at: https://ssrn.com/abstract=3487337 Snowball sampling was used for this study. This type of sampling relies upon social networks (Biernacki & Waldorf, 1981), and therefore, participants were able to recommend others with similar interests and roles who were then approached to take part in the research. In order to obtain data from multiple perspectives, our sampling strategy included both event leadermanagers and event team members such as volunteers. One volunteer and one paid employee were approached to be the initial participant 'seeds' after which the snowball sampling approach was used. Care was taken to ensure a range of participants with different levels of experiences, age groups and levels of responsibilities were included in the research. Interviews were carried out with respondents in Canberra and Brisbane (Australia). Participants were mostly event managers, casual employees, and volunteers from not-for-profit event organizations facilitating small to medium sized tourism events, although larger scale for-profit tourism events were also represented. Participant details are provided in Table 1.

Table 1. Participant details
Interviews had an average length of 60 minutes and were conducted in person, or via Skype video call. The interviews were undertaken until there was a level of saturation of material and when no new information was being obtained (Guest et al., 2006). Saturation was achieved when both quality and depth (e.g. rich, multilayered, and detailed) as well as 'thick' data (e.g. Experience levels: low: 1-5 events or less than 1 year experience; medium: multiple events or 1-5 years of event experience; high: continuous or more than 5 years of event experience.

Interview
No.

Role relevant for this research
Other professional roles Level of experience 1 Gender Type of Event Electronic copy available at: https://ssrn.com/abstract=3487337 the right amount, from a suitable sample from multiple events, with a structured interview guide) was obtained (Fusch & Ness, 2015). The empirical data were then analyzed using an inductive approach and thematic analysis. We adopted Miles and Huberman's (1994) three stages of coding -descriptive, interpretive and pattern codes.
Initially, both researchers coded the data individually and manually, using an inductive approach and descriptive coding as a data reduction technique to identify basic concepts within the text. These initial codes were then further refined using interpretive codes as a way to develop meaningful categories in the data. Finally, pattern coding was used, whereby connections between the categories were noted and the data re-categorized based on this new understanding (Jennings, 2010). Patterns can be characterized by similarities, differences, frequencies, sequences, correspondences or causation (Saldana, 2009). The researchers then met to discuss and agree on the final categories or themes emerging from the data and the agreed-upon themes are those presented in the discussion section.
Before presenting the results of the data analysis, it is important to consider the trustworthiness of the research. A widely accepted basis of assessing trustworthiness within qualitative research has been proposed using four criteria of credibility, transferability, dependability and confirmability (Lincoln & Guba, 1985). Credibility in qualitative research, assesses the ability of the researcher to present findings that best represent the data collected (Lincoln & Guba, 1985;Patton, 1999). For this study, we followed Patton's (1999) suggestions to improve credibility of this study: we used combination of purposeful and snowball-sampling strategy, further we systematically analyzed data opting for an inductive approach, and finally we were careful not to over-analyze results in order to enhance credibility. Transferability, in qualitative terms, is the extent to which the research findings may be applied to a different context (Lincoln & Guba, 1985).
While it was not the intent of this research to make generalized findings, participants were selected to represent a range of different viewpoints and ages, including managerial and general employees, paid employees and volunteers, people with significant event experience and those with little event experience. Dependability is defined as the researcher's level of consistency in interpreting across the range of data sources (Bradley, 1993). An audit trail is one of the ways to enhance dependability -this includes transparently describing the research steps taken throughout the analysis and reporting of findings, as identified in this paper (Korstjens & Moser 2018). Finally, confirmability is defined by Lincoln and Guba (1985) as the process of evaluating whether the researcher has taken reasonable steps to evaluate the data. For this research, intercoder agreement was used to mitigate the subjectivities associated with qualitative research (Lombard et al., 2002). Table 2 outlines and excerpt of intercoder reliability of 86% -thus, with a requirement of percent agreement scores above 80% this result presents an acceptable level for exploratory research (Lombard et al., 2002).

Results
In order to address the aim of this study, we firstly had to conceptualize the unique characteristics of event organizations, and show how these processes and team composition influence knowledge sharing. We then moved on to analyzing knowledge sharing behavior in light of these identified unique characteristics. An overview of our deductive coding process is presented in Appendix 1, displaying representative descriptive codes, which led to interpretive codes and finally were distilled into causation and difference pattern codes. Pattern codes represent our final three themes that will consequently be discussed. Surprisingly, the data analysis revealed that there are more reasons for the event workforce not to share and transfer knowledge then to actually share it.

Difficulties of sharing knowledge in events
A major finding is that participants agreed that knowledge sharing in event organizations is more difficult than in traditional organizations with ongoing operations, due to the large amount of information that needs to be shared in a short timeframe, and the lack of opportunities to actually share information. Participants stated that one difficulty in knowledge sharing lay in the large amount of ad-hoc operational knowledge that required before the event: "lots of organization/coordinative knowledge" (I6), which changed to "technical knowledge such as understanding 'ticketing'" (I5) that was needed during the event. Importantly, knowledge sharing before the event was generally found to be more manageable due to the amount of time available for it, while participants noted that it was more difficult to share knowledge during the event.
Participants also stated that it was problematic to try to process the large amount of knowledge required. The information had to be shared in in a very short timeframe -just in a few hours or days of operations: "People need instant information but they don't know where to go to and whom to ask" (I6). Participants agreed that speed and flexible knowledge, and the ability to adapt quickly to unforeseen situations, were most important: "Knowledge is required instantly and requires a lot of flexibility" (I6). In addition to problems around volume and speed, our data also revealed that knowledge often stayed with the experienced knowledge holder and even when shared, may still be lost: "keeping records is very difficult, because people move on all the time"(I6).
A lack of opportunity for sharing knowledge was seen as another problem. Findings here indicate perceived differences between permanent employees and volunteers. Permanent event managers stated that they felt that there was a lack of opportunity to share knowledge with all employees. Experienced volunteers, on the other hand, complained that their event managers did not make good use of existing opportunities to actively manage knowledge sharing. Experienced volunteers, for example, suggested that post-event meetings and informal gettogethers could serve as an opportunity to bring feedback forward.

Creativity and self-fulfillment vs standardization and career-orientation
Another set of challenges in knowledge sharing for event organizations emerged from different underlying motives to share knowledge. Individual motives varied between permanent employees and volunteers, but also differed between experienced and less experienced members of the event workforce, regardless of contract type. One reason not to share knowledge for volunteers was the motive to remain creative. An experienced volunteer, for example, admitted that sometimes they chose not to share knowledge, saying that "sticking too much to the rules is not helpful" (I2).
Electronic copy available at: https://ssrn.com/abstract=3487337 Similarly, it was suggested that knowledge could actually hinder flexible problem-solving during the event. "Often people are over-informed, which creates a problem" (I2). At the same time, an experienced event coordinator highlighted the difficulty of "trying to standardize behavior" whilst "having access to a lot of creativity" (I4

Knowledge sharing and power
A major theme around motivation to share knowledge emerged around the notion of 'power'.
Knowledge was found to be utilized by experienced event workforce members, regardless of their contract form, as means to influence and exert control over others. Both experienced volunteers and on-going, paid event coordinators, both of whom had been involved in the events for the longest time, were seen as the guardians of most of the knowledge. This knowledge placed them in a powerful position. Event coordinators, however, also recognized that it was often difficult for volunteers to articulate problems -due to issues around position power with regards to paid employees. "I experienced low intention to share knowledge" (I3).

"Knowledge is power, don't lose power; this is what some volunteers think" (I7). "Often it [the knowledge] was left with one person. It is a very vulnerable point, as knowledge is not
recorded, and often sits with only one person. The power of the knowledge holder becomes great and therefore problematic. When this person leaves, the whole knowledge is gone" (I3).
Interestingly, while volunteers perceived an increase in power through not sharing knowledge, an event manager respondent felt that knowledge sharing served her as a means to empower employees and to distribute and facilitate decision-making. She pointed out that failing to share knowledge leads to duplication of effort: "The more informed people are, the more empowered they are (…) I love to share, because I enjoy it, I want people to be empowered and have the parameters for making decisions, and I don't like being asked the same questions over and over again" (I9).

Discussion
Our findings suggest that knowledge sharing in event organizations is indeed more difficult to understand than in traditional organizations with ongoing operations. The key feature of event organizations that contributes to these difficulties is the pulsating nature of events (c.f. Toffler, 1990). Better integration needs to occur between long-term and short-term volunteers; those casual, temporary employees who need to instantly collaborate with the permanent, ongoing and paid workforce; they all need to share, absorb and use large amounts of knowledge very quickly. After the event, most of the workforce moves on, taking their knowledge with them.
This sequence might resonate with Thiry and Deguire's (2007) suggestion that teams that only exist for one single project do not require knowledge to be shared and passed on.
Another interesting finding from this study is that the highly functional and contract-based heterogeneous team composition of events leads to potential tensions between the diverse motivations of the workforce. For example, permanent employees see the long-term benefits of learning, and sharing knowledge in order to progress towards independent and informed decision-making. This may also be reflected in the differences in motivation between the different types of staff, e.g. a volunteer's higher need for community involvement, and extrinsic rewards (Bang et al., 2009); and differences between long-term orientated employees, and those with a more flexible and short-term orientation (Kim & Cuskelly, 2017;Mair, 2009;Mojza et al., 2010).
Volunteers, too, varied in their willingness to share knowledge. Whilst our findings confirm that some volunteers are willing to share their knowledge and learn from others (as previously suggested by Allen & Bartle, 2014), we also found contradictive evidence, that for some volunteers sharing knowledge was an impediment to creativity. This group of volunteers felt that they were potentially being stifled by the necessity of sticking to what they considered to be rigid rules and requirements. This finding regarding unwillingness to share information that may result in perceived limits to volunteer creativity is likely to be linked to the underlying reasons for volunteering in the first place, such as self-expression and personal interest (Holmes et al., 2018). Importantly, our findings indicate that event employees view knowledge as a form of power, and used this power to gain influence in different ways. Whereas event managers shared knowledge as a means to empower employees and increase their efficiency and effectiveness, experienced volunteers appear to have chosen not to share knowledge as a way to maintain their own power position. Again, in contrast, younger, less experienced volunteers were more interested in absorbing information -yet they still seemed to want to attempt to hold on to their small share of power and knowledge.
As a result of this finding, we posit that knowledge is a source of power that is guarded as a resource by experienced volunteers and provided as a resource by on-going event employees, particularly in leading positions. Our findings are in line with previous research that in organizational settings, knowledge can be perceived as an asset and can be utilized as a source of power (Foucault, 1980;French & Raven, 1959

Contribution, Implications and Future Research
This study makes a twofold contribution. First, in showing that event organizations are unique in terms of their processes and team composition, we advance existing literature and add to the significant shortfall in strategic and organizational event literature (Getz & Page, 2016;Liu, 2018). We conclude that tourism events are temporary, pulsating project-based, complex organizational structures with high-speed processes. Moreover, heterogeneous team composition due to contract types, levels of expertise, and diverse motivations make event organizations unique. We argue that acknowledging these unique organizational-contextual factors is necessary as they influence knowledge sharing on an individual level. In identifying this, we advance theory by extending current literature in strategic event management and responding to the existing gaps in organizational perspective and strategic event management (Liu, 2018), since "strategy for event tourism is a relatively new topic for scholars" (Getz and Page, 2016, p. 611).
Second, we advance understanding of how knowledge is used in event organizations.
Interestingly, we found that knowledge is used as a means of power by both volunteers and paid permanent event managers-however, knowledge as a source of power serves different purposes. Findings indicate that a lack of knowledge sharing in events can be explained by the perception of knowledge as a source of power. Whereas permanent employees use knowledge to influence and empower, volunteers retain knowledge to influence and balance their lack of legitimate power. Our examination of knowledge sharing within the context of unpredictable, fast-paced, and highly heterogeneous event organizations contributes towards addressing the gap in the extant literature relating to understanding knowledge sharing in events (Liu, 2018).
Managerial implications of our findings are directed to event and tourism managers, but certainly also resonate with other organizations with heterogeneous workforce composition (e.g. visitor centers or museums) or other temporary or project-based teams. For managerial practice, we propose that processes and behaviors of the heterogeneous workforce composition needs to be understood. Motivation and work values of both volunteers and employees impact managerial practice. Our study, for example, showed that, career orientation, and extrinsic motivation are different for volunteers and paid employees and an awareness of this is needed as these aspects influence teamwork and performance.
Further, event and tourism managers need to make knowledge sharing a managerial priority.
Although fostering knowledge sharing is often difficult, laborious and time consuming (Szulanski, 2000), and event management has traditionally been short-term orientated (Parent & MacIntosh, 2013), we argue that a strategic long-term orientation on knowledge sharing is essential. Specifically, we recommend that managers consider incorporating knowledge sharing practices such as storytelling, digital story-telling, or informal practices to share communication of suggestions, reflections, ideas and suggestions, or concerns about eventrelated issues.
Further managerial implications include the suggestion that event managers need to be aware that knowledge is perceived as a source of power for both paid employees and volunteers -i.e.
that knowledge is used as means of gaining influence. There is a discrepancy between the organization's need to increase knowledge sharing, and the competing interest of the individual to keep personal knowledge as a source of power (Mabey & Zhao, 2017). Hence, to prevent knowledge hoarding (Heizmann & Olson, 2015) by volunteers, the underlying motivations and values of their diverse employees need to be understood. Knowledge sharing practices should be put in place that are fair (Baskerville & Dulipovici, 2006) and create a moral contract (Heizmann & Olson 2015). Event organizations should also make their organizational values in relation to knowledge sharing explicit, to provide the context for employees and volunteers to internalize these values. Further, we suggest that event managers should focus on the facilitators of knowledge sharing: an understanding of challenges of the pulsating workforce such as employee's retention; establishing a team culture; and addressing the lack of an ongoing corporate/organizational body of knowledge, particularly for one-off or single pulse events (Holmes et al., 2015). Importantly, event managers need to find a balance between informal and formal activities that stimulate knowledge sharing, because too much formal effort may lead to adverse outcomes (Mabey & Zhao, 2017).
A limitation of this study may lie in the exploratory qualitative nature of this research. While rigorously undertaken, a cautious approach should be taken towards attempts to generalize beyond the specific events and contexts researched in this study. In particular, the results of this research are not intended to be fully generalized to other contexts beyond event organizations.
However, we suggest that our findings will have relevance to other event organizations, and to other temporary, time-bound or project-based organizations. This research can also be used as a theoretical basis to underpin future studies examining the role of power in knowledge sharing, not only in the events context, but in other organizational contexts.
Future research should continue to add to the evident gaps in event management research and enhance our understanding of the organizational event context. Future studies could extend our understanding of other managerial antecedents of heterogeneity, or the kind of similarities that the tourism and event context creates in terms of motivation, value, identities and experiences for employees and volunteers. We also suggest that event researchers also probe further into the motivation, job experience and job satisfaction of the paid event workforce. As the event management literature largely focuses on understanding volunteers, the paid part of the workforce has rather been neglected. However, in terms of attracting and retaining talent in the increasingly professionalized workforce, we suggest that paid workers should be included in integrative models to understand the unique organizational context and human resources management of event organizations.
Event organizations might also serve as case studies to explore how diversity is managed in these heterogeneous teams -and with new contextual knowledge, implications for other more traditional organizations could possibly be drawn. It will be very interesting to see how creativity and innovation occurs in this fact-paced context. Importantly, we propose that future research should explore how practices and mechanisms that foster knowledge transfer -and how the need for volunteer's creativity and the need to standardize for event managers -might be aligned. Here, studies could draw upon the latest research in knowledge management and explore how storytelling might help to creatively transfer knowledge between both volunteers and paid staff.

Conclusion
The aim of this study was to explore how the unique processes and team composition of event organizations influence knowledge sharing. We drew on organizational behavior, strategic management, and the extant events literature, to conceptualize the event workforce. Our synthesis of the literature showed that the event workforce is highly heterogeneous based on both functions and contracts; event organizations are temporary and project-based structures where staff is difficult to retain. With these unique characteristics, we suggest that knowledge sharing in event organizations is both unique and more complex than in traditional organizations with ongoing operations.
Unique characteristics of knowledge sharing in events include the large amount of ad-hoc operational knowledge that needs to be shared in a short timeframe, the lack of opportunity to share, and individual motives of the members of the heterogeneous workforce to share or not to share their knowledge. Our findings indicate that knowledge is utilized as a means of power.
Experienced event employees, regardless of their contract form, use knowledge to influence and exert control over others. Volunteers perceive that they can gain power through not sharing knowledge. Paid event staff share knowledge as a means to empower employees and to distribute and facilitate decision-making. Sticking too much to the rules is not helpful When over-informed, it creates a problem Paid employees need to be reminded to be respectful towards volunteers Some volunteers say that they don't want to do certain things. They feel they can choose because they are volunteers Motivation [of volunteers] firstly, get free entry, secondly to socialise For young people, it is about gaining life experience. Most important learning was group Either standardise behavior or be creativity at the event Higher experience leads to more complacency to share knowledge When they [volunteers] are young they want to be 'safe' and do exactly what they are told to do. When they become older, they become more relaxed. Relaxedness comes from experience 40' they think they know everything and will not follow any instructions Volunteers: One reason not to share knowledge for volunteers, was the motive to remain creative Event organizers felt that most volunteers did not bring the 'right' motivation and it was difficult to collaborate Different value of knowledge transfer: the need for independent decision making vs the need to learn and progress careers Creativity and selffulfilment vs standardisation and career-orientation (type of pattern: difference and causation pattern)

APPENDIX. Results of coding process
Knowledge is power, don't lose power; this is what some volunteers think I experienced low intention to share knowledge. Often it [the knowledge] was left with one person. It is a very vulnerable point, as knowledge is not recorded, and often sits with only one person. The power of the knowledge holder becomes great and therefore problematic. When this person leaves, the whole knowledge is gone [Event manager]: The more informed people are, the more empowered they are (…) I love to share, because I enjoy it, I want people to be empowered and have the parameters for making decisions, and I don't like being asked the same questions over and over again Volunteers opt to not share knowledge to retain power Permanent employees choose to share knowledge to empower and distribute decision-making Knowledge is a source of power for both volunteers and permanent employees (type of pattern: causation pattern)