The role of prior work experience in top-management teams of nascent
ventures in the creative industries.
Master Thesis – Steffen Wijdemans (10672540)
Supervisor: Prof. Dr. Nachoem Wijnberg Second assessor: Dr. Frederik Situmeang
Table of Contents
Abstract ... 3
1.
Introduction ... 4
2.
Theoretical framework ... 8
1.
Work Experience ... 8
2.
Top Management Team Composition of Nascent Ventures ... 13
3.
The Creative Industries ... 18
3.
Methodology ... 23
1.
Dependent variable ... 25
2.
Independent variables ... 25
3.
Moderating variable ... 26
4.
Control variables ... 26
4.
Results ... 33
1.
Descriptive analysis ... 33
2.
Statistical analysis ... 36
5.
Discussion ... 40
1.
Managerial Implications ... 43
2.
Limitations and future research ... 44
6.
Conclusion ... 46
References ... 48
Abstract
The creative industries have proven its increasing political and economical value during the last decades. Compared to other industries, the creative industries are characterized by innovation but also uncertainty. These characteristics demand tailor fit research and poses generalization issues for managerial studies. The aim of this research is to
contribute to the search for optimal management team composition of nascent ventures in this dynamic environment of the creative industries. It does so by investigating the role of prior related work experience of members of the top management team. In this research we test the effects of work experience and the variation of work experience among members of the top management team. Drawing on a sample of 1511 video game releases, the results of this study show negative effects of both the prior related work experience as the diversity of prior related work experience in top management teams.
1. Introduction
The consumption of creative products has risen considerably in the last decade. As a result, the creative industries have acquired a more substantial share of the market in the world of business. Various explanations have been given for the increase in
consumption. One view is that this is due to the growing importance of intangible and symbolic goods in the global economy (Wong, Millar, & Choi, 2006) . Others argue that the increase is related to the successful marketization of culture (Camelo-Ordaz,
Fernández-Alles, Ruiz-Navarro, & Sousa-Ginel, 2012) . The debate on the reasons for the increase appears to be slightly speculative and lacks evidence. What essentially matters for our research is that the creative industries have acquired a certain level of economic and political significance and are still growing in size. The UK Department of Culture, Media & Sports reports a job growth for the creative economy in the United Kingdom of 8.8 % between 2011 and 2013 (Department for Culture, Media & Sport, 2014) . The total number of jobs in the creative economy of the UK was therefore estimated at 2.6 million in 2013. Policy makers have acknowledged the significance of the creative industries during the recent years and it has since been promoted as “good for the economy” (Banks & O’Connor, 2009)
The economical and political relevance of the creative industries is evident. Less straightforward is what actually belongs to the creative industries. According to
UNESCO, the creative industries include a wide variety of goods and services “including those produced by cultural industries and those that depend on innovation, including many types of research and software development” (UNESCO, 2013). The Video games industry, part of the creative industries and used for the sample of this research, belongs according to different definitions to either the core or wider cultural industries. Video game producers are fairly new participants in the creative industries compared to most other professions (e.g. actors or musicians). The video game industry is however the fastest
evolving media sector of the past decade and is still faster than ever increasing in size (Cadin, Guérin, & DeFillippi, 2006; Christopherson, 2004; Hotho & Champion, 2011) . Video games currently represent a global significant economic share of the
entertainment business (PricewaterhouseCoopers, 2008) . Recent reports show an astonishing fifty percent revenue growth between the years 2006 and 2011. The total European gaming population has more than doubled from 40 million in 2008 to 95 million in 2010 (CCIA, 2013). These reports confirm the economic relevance of the video game industry in the creative industries.
The video gaming industry has consequently caught the attention of the academic literature (Zackariasson & Wilson, 2010; Tschang, 2007; Hau and Kim, 2011). In
particular the pshycological effects of video games received an abudant amount of attention in academic literature in the beginning (for a review, see Bensley & Van Eenwyk, 2001). Video game development firms receive more attention from literature with the focus on for example innovation management and organizational architecture (Johns, 2006) . Tschang (2007) argues that video game development requires for
management research an interesting mix of technology and arts. This mix forces firms to express artistic values and creative contents to meet with the challenging demand and economics of mass entertainments (Cohendet and Simon, 2007). For this reason, Burger-Helmchen & Conhedet (2011) argue that “the video game industry is at the forefront of activities that challenge several commonly-held ideas about the way firms should manage their intellectual capital, property rights, production and organizational
structure”. In other words, the video game industry is not merely interesting because of its economic value. Firms in the industry also operate in an interesting environment, which is dynamic, creative and challenging.
The aim of this research is to contribute to the search for optimal management team composition of nascent ventures in the interesting environment of the creative
industries. Firms in the creative industry may need to emphasize on different factors for optimal team composition compared to firms in other industries due to its dynamic and challenging nature. While there is literature on the effects of management team
composition of nascent ventures (Steffens, Terjesen, & Davidsson, 2012) , it is context specific and lacks focus on the creative industries. Research stresses the importance of the context because it is believed to be the cause of mixed results in studies (Webber & Donahue, 2001). We are interested in nascent ventures because they face start-up specific challenges and have interesting characteristics. One of these startup-up specific challenges that founders face is that they need to select a team of professionals for their management team. The challenge lies in choosing the right individuals who will either combined or individually will contribute to the demands of the business environment. Founders may come across countless different profiles of candidates for their team or may question themselves whether they have teamed up with the right persons or not. The focus on the team is important because the management team members are in most cases the most important assets of a firm in the start-up phase (Delmar & Shane, 2006). Team-founded new firms are proven to outperform individually founded firms (Cooper, Gimeno-Gascon, & Woo, 1994) and over 50 percent of all nascent ventures are managed by two or more team members (Reynolds, Carter, Gartner, & Greene, 2004) .
Management teams of nascent ventures are therefore considered fundamentally important for the venture’s success.
This research on optimal management team composition of nascent ventures in the creative industry is focuses on work experience. It is commonly assumed that experienced employees are more beneficial to firms than less experienced employees. Firms pay a higher premium for more experienced employees (Greeno, Moore, & Smith, 1993) . There is however also evidence that increased experience of individual
(Steffens et al., 2012). This would suggest that experienced management team members are only beneficial for an organization when they work together with equally experienced team members. So what does this mean in practice for management teams of new firms? A two-way approach is used in order to answer this question. Firstly, the effect of work experience of individual top management members in general is tested in order to measure the effect of increased work experience in a team. This approach attempts to shed light on the discussion whether founders should focus on searching experienced management team members. Secondly, the effect of diversity in work experience of team members is tested in order to see whether management diversity affects the success of new firms. Should founders be aware of conflicts due to diversity in the management team or are there possible benefits that origin from team diversity?
The foundation of this paper is consequently structured around three subjects. Namely: 1) work experience, 2) management team composition of nascent firms and 3) the creative industries. The chapter on work experience will discuss previous findings on individual work experience and explains the relevance of industry or occupational work experience of managers. It is aimed to find a theory on the role of work experience and to extent this to the firm’s performance. Theory on work experience is furthermore
extended to management team composition in the analysis. The chapter on management team composition presents frameworks and results of published literature and aims to find an overview of the effects of team composition of management teams. The last chapter of the theoretical framework points out the possible effects of the creative industries specific characteristics on work experience and management team
2. Theoretical framework
1. Work Experience
Career paths of employees have changed over the last decades. To illustrate, Kolb (1984) estimated individuals to have around 7 employers during their entire career in the late 70’s. By 2005, it was estimated that the average employee in the U.S. would work for 10.5 employers before turning 40 (Bureau of Labor Statistics, 2006). This change in the employees’ careers consequently means that the experience an individual gains is less bounded to a single organization. Experience gained in multiple organizations has therefore become increasingly relevant for the total individual work experience in the last decades. The majority of the research is however done on employees in the current organization (e.g. organization tenure) or on job-specific (e.g. CEO tenure) experience. This ignores experience gained in other organizations or experience gained related jobs. Previous research (Greeno et al., 1993; Neal, 1995) stresses the importance of occupation and industry specific work experience. This paper focuses on prior related work
experience and therefore attempts to remain occupational and industry specific while studying various organizations. Prior related work experience (PRWE) will be used in the continuation of this paper to refer specifically to job or industry specific work experience.
An increasing amount of literature is published on PRWE, in particular on job performance of employees. Some studies are limited to single-organizations while others have researched the role of PRWE on multiple organizations. Single-organizational studies present negative (Dokko, Wilk, & Rothbard, 2009a) , insignificant (Castilla, 2005) and positive (Hunter & Thatcher, 2007) results of PRWE on individual job
performance. Multi-organizational studies on the other hand in majority found positive results of PRWE on individual job performance (McDaniel, Schmidt, & Hunter, 1988;
Quińones, Ford, & Teachout, 1995; Uppal, Mishra, & Vohra, 2014) . There is no scientific consensus in literature on why these results differ. The generalizability of much published research on PRWE is therefore problematic. It is argued that the mixed outcomes are a result of the many different factors influencing PRWE (Castilla, 2005; Dokko et al., 2009) . The results of scientific research on PRWE should therefore be regarded with caution. It is wise to focus on the specific effects of PRWE on job performance and the various components of work experience.
The majority of the published literature focuses on specific effects of PRWE on one individual’s knowledge and skills. Neal (1995) and Rynes, Orlitzky & Betz (1997) for example found evidence for a positive relationship between PRWE on knowledge and skills of managers and employees. More recently, Dokko, Wilk & Rothbard (2009) confirmed this positive effect of PWRE on knowledge & skills. The effect works in two ways according to the authors. First, PRWE has a positive indirect effect on job
performance through its effect on knowledge and skills. Indirect, because preceding work experience gives an individual the opportunity to improve one’s knowledge and skills (Borman, Hanson, Oppler, Pulakos, & White, 1993) . The strength of the gained knowledge or skills is however according to Dokko et al. (2009) depending on the relatedness of the previous experience to the current task. “The consensus is that
transfer is more likely to be attempted when the subject perceives similarities of tasks or contexts” (Dokko, Wilk, & Rothbard, 2009) . Moreover, experiences of previous jobs become decreasing important compared to experiences in a current job because recent experiences provide a better fit to the current job (Greeno et al., 1993). This argument is supported by the willingness of firms to pay a premium for management team members with more relevant experience (Greeno et al., 1993). The effect of PRWE on performance of managers and employees through increased knowledge and skills is strongest when the PRWE is task or context specific.
The second way the effect works is negatively. The downside of PRWE is that individuals set expectations of the way to work and what is appropriate behavior in their working environment. Beyer & Hannah (2002) argue this may be problematic for the current job as it increases chance of potential conflicts within teams. The general argument for this is that experienced managers or employees are less flexible in changing their behavior. Many industries or organizations have formal and informal norms of how activities should be handled. Working in another industry or organization may therefore be very different. Moreover, repetitive experiences lead to cognitive schemas or scripts (Woltz, Gardner, & Bell, 2000) . Employees or managers tend to fall back to these schemas or scripts when they are exposed to uncertainty or stress in their working environment. Woltz et al. (2000) argue that this is particularly problematic when activities show similarities, yet are structurally different. In other words, employees or managers become less willing to adept in new situations and fall back to old habits and thoughts when they are confronted with stress or uncertainty. Dokko et al. (2009) note that this results in a negative effect of PRWE on performance through team conflict. The same authors also argue that the positive effect of PRWE through increased knowledge and skills outweighs the negative effect. An overview of the proposed effects of the previously mentioned relationships is illustrated in figure 1.
Figure 1: PWRE on performance
The mixed results of the studies of PRWE on performance ask for a closer look at the different components of PRWE. Tesluk & Jacobs (1998) identify three components for the individual in their model of PRWE. According to the writers, experience is build by quantitative, interaction and qualitative components. Quantitative aspects are measured by the time spent on jobs and the rate of repetitiveness for a task. Effectively measuring the qualitative component for experience is domain and context specific (Tesluk & Jacobs, 1998b) . The essence of measuring quality of experience lies in the strength of the quantifiable component (Uppal et al., 2014). Training, opportunities for supervising and challenge are examples of factors of the qualitative component. Interaction refers to the interaction between the quantitative and qualitative component and it measures the intensity of the experience. High-intensity experiences are likely to have a strong effect on an individual’s career (Uppal et al., 2014). Knowing how much time an individual has spent on a job might therefore not be as relevant without being aware of the qualitative component and the interaction between the quantitative and qualitative component.
To sum up, careers of individuals have changed over the past decade. This
demands an approach without single-organizational boundaries on work experience. The generalizability of studies on PRWE is problematic as they present mixed results on
relationship between PRWE and performance. Positive effects by PRWE occur by increased knowledge and skills of individuals. Knowledge and skills are considered to have a positive relationship to the performance of an individual. The increased
performance is however determined by the relatedness of the experience to the current task or context of the individual. A direct negative effect of PRWE occurs because managers or employees becoming less willing to adept to a new context. Moreover, it is expected that managers or employees will refer back to old habits and thoughts when uncertainty or stress increases. Both effects are considered to increase conflict within teams in general. Studies argue that the perceived indirect positive effect
2. Top Management Team Composition of Nascent Ventures
A large and growing body of literature has investigated the role of top management teams (TMT) on the performance of nascent ventures. Many of these studies base their findings on the TMT upper echelon theory introduced by Hambrick & Mason (1984). It states that the key attributes of managers and their limitations determine the
performance of the firm. The upper echelon theory opposes the population ecology perspective by Baum (1996). This perspective argues that the performance of a firm is not accountable to key attributes of managers. Instead, the responses to environmental circumstance are accountable for the performance of a firm according to this perspective. Evident is that the success of a firm is hardly ever the results of one solely operating individual. New firms are in most cases managed by two or more individuals and team-founded ventures are significantly more successful compared to single individual founded firms (Cooper, Gimeno-Gascon, & Woo, 1991) . Team members are moreover more successful in assessing complementary resources (Vesper, 1990). The TMT of new firms are considered to be crucial for their success as they are the key resources of the organization and their characteristics determine the strategy of the firm (Delmar & Shane, 2006) .
The importance of TMT to nascent ventures is certain. TMT composition research focuses on the attributes of TMT members and the impact of the combination of these attributes on the performance of an organization. The effect of the TMT composition on performance of nascent ventures is measured by two dominant approaches in prior studies. These approaches for indexing team composition and the effect on outcome are the mean values and the diversity approach (Kozlowski & Ilgen, 2006) . The mean values approach valuates the mean attributes of team members combined. This pooled value is tested for the influence on the team’s performance. The distribution of the variable is therefore not taken into account by the mean values approach. The majority of research
is done through the mean values approach on TMT. It has been proven to be a reliable predictor of performance (Stewart, 2006). For example, Devine & Philips (2001) concluded in their meta-analyis of 19 studies ‘that the functional amount of cognitive ability in teams does indeed predict team performance across a broad variety of team contexts”. The mean values approach is used in this research in order to measure the effect of the average level of PRWE within a management team. Theory presented in the previous chapter serves as a fundament for the effect of the mean PRWE value. It was argued that individual PRWE works both positively as negatively on performance.
The team diversity approach considers the influence of management team
member’s heterogeneity on mediators and outcomes. Previous studies have reported that the function of diversity is complex in the explanation of a management team
effectiveness model (Kozlowski & Ilgen, 2006; Marks, Mathieu, & Zaccaro, 2001a) . Researchers make a distinction between demographic and functional backgrounds when analyzing the background of the team members. Demographic background typically include factors such as age, race/ethnicity, gender and type of education. For example, individuals who share demographic characteristics are believed to have increased levels of commutations (Marks, Mathieu, & Zaccaro, 2001b) . This study ignores demographic backgrounds and focuses on functional backgrounds only. Functional background reflects the human capital whereas human capital broadly stands for education, knowledge, skills and how these factors generate returns (Shrader & Siegel, 2007) . PRWE links to the functional background as it influences knowledge and skills.
Linking diversity of teams to performance has been the focus of a number of studies with different outcome in the TMT literature (Webber & Donahue, 2001) . Lester, Certo, Dalton & Canella (2006) did for example find evidence for a positive effect of team heterogeneity on financial performance while Ancona & Caldewell (1992) did not find any evidence. A possible explanation for the inconsistent findings is that types of
diversity strongly fluctuate in results. Steffens et al. (2012) argue that the reason of the inconsistent findings is because the literature fails to differentiate between the negative and the positive effect of team heterogeneity. The effect of management team diversity is argued to work through two mechanisms while most studies focus on one.
Emotional conflict and task conflict function as the two mechanisms for the effect of heterogeneity in TMT composition. Previous research indicates a positive relationship between task conflict and team heterogeneity (Pelled, Eisenhardt, & Xin, 1999) . Task as include for example generating ideas or plans, problem solving and decision-making. Steffens et al. (2012) confirms this and argues that task conflict in return improves outcome of the TMT. There are several reasons why increased task conflict leads to improved performance. “Task conflict leads to multiple perspectives in teams, the avoidance of hazardous situations and it promotes innovative thinking” (Cosier & Dalton, 1990) . Watson, Kumar & Michaelsen (1993) add to this that task conflict avoids group thinking. This results in a wider variety of ideas and solutions within groups. Amason, Schrader & Tompson (2006) confirm the positive view in their study of 46 TMTs by finding evidence for improved decision-making by TMT heterogeneity. It is important to mention that task conflict is considered especially important for non-routine tasks and it improves team performance in highly uncertain environments (Ensley, Pearson, & Amason, 2002). In general, task conflict can therefore be seen as beneficial, particularly when a team performs non-routine tasks and operates in an uncertain environment.
The other mechanism is that of emotional conflict by TMT heterogeneity. TMT heterogeneity is argued to decrease harmony in a team (Steffens et al., 2012). This is because similarities in demographic and functional background improve communications within a team (Greeno et al., 1993). TMT heterogeneity is additionally argued to
(Steffens et al., 2012) and has a negative effect on group cohesiveness (Jehn &
Bezrukova, 2004) . Amason et al. (2006) confirm this by finding evidence for inadequate communications and imperfect relationship by TMT heterogeneity.
Steffens et al. (2012) furthermore argue that the effect of emotional and task conflict is influenced over time. Research indicates that group cohesiveness improves over time and therefore diminishes the negative effects of team heterogeneity (Watson, Kumar, & Michaelsen, 1993) . The two mechanisms of emotional conflict and task conflict are illustrated in figure 2.
Figure 2: TMT composition on performance
To sum up, a large amount of literature has been published on TMT heterogeneity and the results are mixed. Most of the literature is based on the upper echelon theory. The upper echelon theory is especially interesting for nascent ventures as it argues that TMT individual characteristics influence the organizational results and the TMT is argued to be of great importance at a nascent venture. There are several ways to index the effects of team composition. The mean value and the team diversity approach are most commonly used to index the effect of team composition. The effects of
heterogeneous backgrounds are mixed. It is however argued that this is due to a dual mechanism from emotional and task conflict. Heterogeneous teams are argued to perform better in complex tasks, are less vulnerable for group thinking and experience improved strategic decision-making with increased creative thinking. Homogenous teams are however more harmonious, resulting in better communication, decreased emotional conflict and improved social interactions. Time is argued to decrease both task and emotional conflict.
3. The Creative Industries
One could debate on the difference of the creative industries (CI) among industries. Many argue that the creative industries are indeed different and pose specific challenges due to the specific characteristics of the industry (Caves, 2000; Townley, Beech, &
McKinlay, 2009). Products in the creative industries are different than in other industries. “Creative works are symbolic, experiential goods of non-utilitarian value” (Townley et al., 2009). Creative goods are not consumed as regular goods and the consumers base their valuation on the translation in value. Caves (2002) argues that this is why these goods raise an inherent uncertainty. The CI is also different in the importance of intermediaries in the role of gatekeepers. Gatekeepers are considered to be valuable for the success of creative products. Also, individuals drive the CI as significantly as teams and organizations (Caves, 2000).
As with any industry, firms in the CI develop more market orientation when they mature in age. Creative products such as video games and films become increasingly expensive and complicated to produce. Tschang (2007) argues that this leads to a rationalization of production, facing a trade-off to creativity. Moreover, developments of vertical integration in the industry are leading to a limited amount of powerful players, possibly constraining innovation (Mezias & Mezias, 2000) . Tschang (2007) claims that the entertainment products in the CI can be characterized in three ways. 1) A hits-oriented nature, 2) by products with short life cycles on the market and 3) difficulties in estimating product acceptance. These characteristics drive the CI sometimes to a slightly conservative nature with minor ambition to big changes. Lampel, Lant & Shamsie (2002) however state that the search for innovation thrives the creative industries. According to the authors, consumers seek novelties as long as it is accessible to them.
An uncertain market producing complex goods consequently demands the development of products that are adapted to these conditions. When we link this to the
theory on PRWE and TMT composition, we need to make a prediction on the appropriate TMT composition of nascent firms in the creative industries. The theory on TMT
composition states that task conflict improves the capability of the TMT to handle complex and uncertain tasks. TMT heterogeneity is considered to have a positive effect on task conflict. Task conflict enables TMT to gain new insights, prevent risky situation and develop innovative ideas. These benefits are all considered to be especially
important for TMT operating in an uncertain environment performing non-routine tasks. These characteristics are in accordance with the characteristics of nascent ventures in the creative industries. We therefore expect based on the team diversity approach that TMT heterogeneity in any form is beneficial for the performance of nascent firms in the creative industries. PRWE heterogeneity in TMTs of nascent
ventures is also expected to have a positive effect on the performance of nascent ventures in the creative industries.
Hypothesis 1a: Increases in top management team PRWE heterogeneity will increase performance of nascent firms in the creative industries.
Furthermore, TMTs of nascent ventures are considered to have a greater influence on a firm’s performance compared to mature firms (Delmar & Shane, 2006) . Moreover, the effect of TMT heterogeneity through task conflict decreases over time (Steffens et al., 2012). We therefore expect that the positive effect of PRWE TMT heterogeneity on firms in the creative industries is new firm specific and diminishes when TMT collaboration increases over time. This suggests that time moderates the effect of top management team PRWE heterogeneity on firms’ performance. This is supported by the theory on teamwork which states that the interaction effect is a important determinant of the result of teamwork (Tesluk & Jacobs, 1998a) .
Hypothesis 1b: Time has a moderating effect on the relationship between top management team PRWE heterogeneity and the performance of firms in the creative industries
As stated before, working in the creative industries is different from any other industries. New products are constantly developed requiring re-form in value chains of creative firms. Various stages are completed in this process including idea creation, research and development, production, and distributions. The basis of value creation however lies in the intellectual capital according to (Chen & Wangz, 2009). This results in short-term employment contracts or freelance agreements (Creigh-Tyte, 2005). Arthur & Rousseau (1996) argue that careers are “boundaryless” in as they offer hardly any employment stability or hierarchical organization progression and are assigned individually. Bridgestock (2011) adds that formal educational credentials of
professionals are much less likely to be of importance compared to informal contacts and PRWE.
Intellectual capital (IC) is by many argued to be essential for creative work (Bridgstock, 2011; Camelo-Ordaz et al., 2012; Hotho & Champion, 2011) . Intellectual capital was even first used to explain the difference between market and book value of a firm (Chen & Wangz, 2009) . The intangible assets are argued to be more important than the tangible assets of a firm in the creative industries according to these authors. It is a composition of knowledge and abilities anchored in organizational members. Subramaniam & Youndt (2005) place IC capital into three categories, respectively: (1) human capital, (2) organizational capital and (3) social capital. Human capital is
referred to as the employee’s knowledge, skills and experiences. Organizational capital is institutionalized in the organization processes and culture while social capital
incorporates the interaction between employees and their internal and external
relations. Human capital is argued to be the core of creative occupations (Throsby, 2001) and is therefore positioned as a strategic assets used to gain competitive advantage.
Human capital is undoubtedly important for the success of a firm in the CI and is considered to be the core of creative occupations. Careers in the CI are highly flexible and previous work experience is considered to be much more important compared to for example educational credits. It is therefore expected that PRWE of TMT members has a positive effect on the performance of nascent firms in the creative industries based on the mean values approach. The theory of PRWE on TMT members states that PRWE positively influences knowledge and skills of a member. Research on PRWE however also indicates that managers become less flexible when PRWE is high. Due to the importance of human capital in the creative industry we expect that the positive effect outweighs the negative effect of PRWE. This is coherent to the theory on PRWE. This states that the benefits on increased knowledge and skills outweigh the negative effects of PRWE that may possibly cause team conflict due to for example the inflexibility of managers.
Hypothesis 2a: Increases in top management team PRWE will increase performance of nascent firms in the creative industries.
Similar, to hypotheses 1a and 1b we expect this effect to differ at firms in the CI. This is because the effect of the TMT on nascent ventures is considered to be stronger than on other firms as mentioned before. However, it was also stated that human capital is essential for all firms in the CI and all the other arguments for hypothesis 2a are applicable to all firms in the CI. We therefore also expect a positive effect of top
management team PRWE on the performance of firms in the creative industries because we expect that the positive effect of PRWE is not new firm specific.
Hypothesis 2b: Increases in top management team PRWE will increase performance of firms in the creative industries.
3. Methodology
The sample collected for this research consists of three independently databases combined in one. All three are selected in the videogame industry with the purpose to test the proposed hypotheses. This sector of the CI has received little attention of research compared to other creative sectors, such as for example the music and movie industry. Notably the managerial aspect and the sale’s factors of the video game
industry lack sufficient research (Tschang, 2007). Surprisingly one might argue, because the video game industry market has become increasingly important in economic value over the last decade. Video game sales oversized movie ticket sales back in 2004 and music album sales in 2005 (Cadin & Guérin, 2006). The volume of the video game industry market was estimated for 2013 at 65.7 billion USD in total Statista (2014). Based on its economic value, the industry serves therefore as a relevant research sample.
The first database, collected from Metacritic (Metacritic, 2013), contains information on videogames with details about game release details and quality
evaluation. Metacritic provides a publicly accessible database, compiled by Metascores of critics on movies, games, tv shows and music. The scores are generated by distilling opinions of respected reviewers into Metascores by Metacritic. Also, users are able to review games themselves. Metacritic is therefore not only showing the experts’ opinion on the subject but also the users’ opinion. The exact selection procedure of selected experts remains however unknown. The dataset also includes extensive information on the release of a total of 5089 game titles (e.g. date, title & genre).
Publicly accessible information of Mobygames (Mobygames, 2013) is used to compile the second database of 8519 game titles. Mobygames is an online database used by professionals in the industry to retrieve information on games. The information provided by Mobygames provides opportunities to analyze career paths for employees at
game development studios. Mobygames grants access to game credits according to the function of an employee in relation to the released game. The Mobygames database thus adds crucial information to the research as this information enables to obtain values of individual experience of employees in the industry.
The last database contains information on total sales provided by publicly available data on VGchartz. All sales are estimates of VGChartz and compiled by using various methods. Polling end users on consumption, polling retail partners, statistical trend fitting, studying resell prices and consulting publishers are actions taken by VGCharts to establish sales estimates (VGChartz, 2013).
The number of credits for game titles reflect the PWRE of individuals involved in the production of the game. All video games where an employee previously worked on between the years 2000 and 2010 are taken into account for each employee at the current production. Out of this a selection is made of PRWE of team members
responsible for the production of the video games in the sample. Consequently, credits of released games in the database are selected on function title. Main roles in game
production typically involve the roles of a designer, programmer and producer. The designer is responsible for the creative aspect of the game while the programmer works on the technical aspect. The producer typically carries the responsibility of the progress in the production of the game, the financing and the publishing. Most game productions teams in the sample consist of many more individuals than three. Therefore the key employees in charge of these roles are selected, reflecting the top management team. A dummy variable is assigned to game studios first release in order to filter for nascent ventures. The combined databases match a number of 1511 unique video game releases of games released between 2000 and 2010. Out of these 680 have been assigned with the dummy variable of first production.
1. Dependent variable
Sales Performance
Total sales are used as dependent variable and reflect performance. The sales data are estimates for worldwide sales in units of sold games. Video games are possibly released on various platforms. Total sales are therefore a combination of the worldwide sales on all platforms of a game title. The logarithm of Sales performance was used in order to improve the normality of the variable.
2. Independent variables
Team PRWE Diversity
The diversity in PRWE between TMT members is measured by the coefficient of the variation (standard deviation divided by the mean). This provides a direct and a scale-invariant method to measure dispersion (José, Aragón-‐Correa, & Ferrón-‐Vílchez, 2011; Naranjo-‐Gil, Hartmann, & Maas, 2008), in line with the suggested operationalization for team diversity (Kozlowski & Ilgen, 2006). Team PWRE is considered high when the coefficient of the variation is high. The metric value of PRWE individual experience of TMT members is used in order to calculate the standard deviation between teams, consistent with the upper echelon tradition (Hambrick & Mason, 1984) . It is important to note that the value of experience of these top management team members stretches further than their individual experiences as TMT member. All prior work related
experience is counted, based on the game credits of the sample within the timespan of 10 years. TMT members are filtered out of the sample once all individual PRWE are
calculated. This final selection is used for measuring the standard deviation of PRWE for games between the TMT members.
Team PWRE
In addition to team PRWE diversity, we moreover focus on the average level of
experience of TMT members within a team. Theory suggests that a high level of average experience is positively related to product performance (Dreu & Weingart, 2003). This independent variable thus enables the testing of hypothesis 2a and 2b. The team PWRE is calculated by counting the number of release credits of the TMT members, divided by the number of TMT members. This is therefore inline with the suggested
operationalization of the mean value approach by (Stewart, 2006). Similar to Team PRWE Diversity, the metric value of individual PRWE is used to calculate the mean values of team PWRE.
3. Moderating variable
Nth release
Hypothesis 1b suggests a moderating effect of time on the positive effect of Team PRWE Heterogeneity on sales performance. Every release is therefore assigned a variable depicting the release numbers of the game development firm prior to the release in order to measure this effect. This variable is name Nth Release and reflects the amount of time that is dedicated to producing games by the firm. The sample is sorted on game developer and release date in order to assign a metric value to every single release. We expect a moderating effect between the relationship of Team PRWE Heterogeneity and Sales performance.
4. Control variables
Publisher
Game developers are responsible for a large part of the value chain of the game.
Publishers however play a crucial role in the marketing and advertisements of a game. Large publishers have the potential of releasing a game with a large marketing budget
and have extensive sales networks. Broekhuizen, Lampel & Rietveld (2013) found evidence for higher financial performance of video games released by a publisher compared to self-publishing, outweighing the costs for publishing. The difference in performance is due to the specialized assets of the publisher, required for successful commercialization. The publishers furthermore have the ability to build on an established reputation while many game development companies are small and
unknown. We therefore expect that the reputation of the publisher has an effect on the sales performance of a game. Moreover, a publisher posses a distribution network which enables them to reach a larger target group (Broekhuizen, Lampel, & Rietveld, 2013) .
A comparable variable is used in the movie industry as the distributer variable (Gemser, Van Oostrum, & Leenders, 2007) . It is explained that major film distributors generate significant higher revenues because they possess the resources of excellent distribution networks and marketing practices. In Table 1 a selection of the ten largest video game publishers, based on total revenue (Statista, 2014), is made in order to find the most powerful publishers. Games released by one of these publishers have been assigned the value “1” while games independently released or released by a smaller publisher have been assigned a “0”.
Table 1. Revenue of the largest computer and video game publishers worldwide in 2013 (in billion euros) Publisher Revenue 1. Microsoft 7.65 2. Tencent 5.51
3. Sony Computer Entertainment 5.45
4. Nintendo Company Ltd. 4.41
5. Activision Blizzard Inc. 3.45
6. Electronic Arts 2.69
7. Namco Bandai Games 2.15
8. Take-Two Interactinve, Inc 1.77
9. King Digital Entertainment plc. 1.43
10. Gunho Online Entertainment 1.26
Genre
Several genres dominated the sales of video game industry in 2010 (The ESA, 2011). We expect that popular genres significantly influence sales performance of game releases. It is expected that two important reasons cause the effect of genres on sales performance:
1) Some genres are favored more over other genres by consumers, and 2) Video games are placed in a genre to reduce uncertainty for the consumer, however reducing at the same time the target group for consumers.
Similar research has been done on the role of genre spanning in movies. Films that target multiple or greater niches are expected to attract a larger audience (Hsu, 2006). However, while the film attracts a larger audience, the film creates less appeal and reward in this audience. Six genres account for 80.3% of video games sold in 2010 (See Table 2). Respectively: Action, Sports, Shooters, Family entertainment, Role-playing and Adventure. These genres are considered popular and are coded with a value of “1” while the other genres are coded “0”.
Table 2. Best-Selling Video Game Genre By Units Sold (2010) Genre Percentage (%) 1. Action 21.7 2. Sports 16.3 3. Shooters 15.9 4. Family Entertainment 9.1 5. Role-playing 7.7 6. Adventure 7.5 7. Other 19.7 ESRB Rating
The Entertainment Software Rating Board (ESRB) assigns ratings to software for consumers and minors’ parents to facilitate informed choices. As a non-profit organization it categorizes games in six categories on basis of age-linked
recommendations. These categories serve as ratings for age suitability and, depending on the country served and for example legislation or tradition, indicate a possible age barrier. Literature on age restrictions for movies by MPAA (Movie Picture Association of America) ratings indicates an effect on box office sales (Terry, Butler, & De’Armond, 2005) . According to the authors, producers are even willing to replace or remove scenes in order to get a favorable rating, excluding as little audience as possible. Conforming to the literature on MPAA ratings, we also expect to find an effect of ESRB ratings on games’ sales performance due to age restrictions. Table 3 shows the corresponding age-linked categories.
Table 3. ESRB Ratings
Rating Description
Early
Childhood (EC)
Content is intended for young children.
Everyone
(E)
Content is generally suitable for all ages. May contain minimal cartoon, fantasy or mild violence and/or infrequent use of mild language.
Everyone 10+
(E10+)
Content is generally suitable for ages 10 and up. May contain more cartoon, fantasy or mild violence, mild language and/or minimal suggestive themes.
Teen
(T)
Content is generally suitable for ages 13 and up. May contain violence, suggestive themes, crude humor, minimal blood, simulated gambling and/or infrequent use of strong language.
Mature
(M)
Content is generally suitable for ages 17 and up. May contain intense violence, blood and gore, sexual content and/or strong language.
Adult only
(AO)
Content suitable only for adults ages 18 and up. May include prolonged scenes of intense violence, graphic sexual content and/or gambling with real currency.
Expert & User Ratings
Expert and user ratings are expected to have a positive relationship to sales and are therefore controlled for. Expert and consumer ratings are publicly available for many games and are great in numbers. Zhu & Zhang (2010) found evidence for the effect of consumer reviews on video games’ product sales. Less popular (niche) and internet-based games are believed to be more influential to the effect of consumer reviews.
Furthermore, games are even argued to be more vulnerable to consumer reviews than movies due to a higher price per unit sold and a larger and more diversified supply. Experience of a game may differ on one platform to another. Ratings may therefore differ
platforms diminishes this effect. Literature on expert reviews of movies suggest that expert reviews are used by consumers to lower uncertainty, when consumer reviews are not yet available (Basuroy, Chatterjee, & Ravid, 2003) . In line with the literature of expert and consumer reviews, we expect expert & user ratings to influence video games’ sales performance.
Number of User Ratings
In addition to the user and expert data scores, the sample also provides information on the number of users that rated the game. We expect this variable to be an indication of the popularity of a specific video game and therefore expect to find high levels of positive correlation between number of user ratings and sales performance.
Team size
The effect of team size is controlled for due to several reasons. First, formalization and institutionalization increase when the size of a firm expands. There is evidence for a decreasing effect of TMT on performance when formalization and institutionalization increases (Thomas & Ramaswamy, 1996) . The same logic is expected to apply for this sample. Furthermore, the team size of a production is indicating some indirect factors. For example, the production budget is most likely high when the team size of a
production is high. A higher production budget is also likely to provide for a larger marketing budget of the production. A higher team size also indicates that more effort has been put in a production. This is expected to lead to an increased quality of the production. Table 4 summarizes the discussed variables.
Tabel 4. Overview of the presented variables
Variable Description Source
Sales Performance Worldwide sales of the game. VGChartz
Team PRWE Diversity The coefficient of the variation between individual PRWE of core team members.
Mobygames
Team PWRE The mean value of PRWE of come team members. Mobygames
Nth release Numerical variable indicating the number of previous
video game releases including this release by a game studio.
Metacritic
Publisher Variable indicating whether the game is released by the
top 10 most powerful publishers or not.
Statistica
Genre Dummy variable indicating if a release belongs to the top
6 most popular genres.
ESRB Rating The rating classification by ESRB ranging from EC to
AO.
Metacritic
Expert Rating Compilation of expert scores of reviews on video game
release.
Metacritic
User Rating Average score of a video game’s user scores. Metacritic
Number of User Ratings
Metric value, showing the amount of reviews that are submitted by users for the video game.
Metacritic
Team size The amount of employees used for producing the video
game release.
4. Results
1. Descriptive analysis
This section will provide insights on the analyses of the models and the results given by the analyses. The basis descriptive will be provided first before going into depth in the regression analyses. Four hypotheses were formed on the fundament of the theoretical framework. These hypotheses require different samples and therefore it is necessary to analyze several separate models. While hypotheses 1a and 2a focus on nascent ventures, hypotheses 1b and 2b broaden the scope to all firms in the sample. Consequently, the sample of all firms (sample 1) is used to test hypothesis 1b and 2b and the sample of nascent firms (sample 2) is used to test hypothesis 1a and 2a.
The data have been checked for missing values before conducting any analysis. No missing values were found. Outliers were found and removed in order to ensure normality of the variables in the sample. The mean values, standard deviations and bivariate correlations for the variables are given in table 5. PP & QQ plots present normal patterns after removal of outliers in the sample. Moreover, the variables were checked on the occurrence of multicollinearity and heteroscedasticity in all models. None of the variables showed a higher VIF value than 1.476. Scatterplots show no indication of heteroscedasticity in the data. This meets the conditions for performing a multiple linear regression.
The presented standard deviations and means are consistent with the expectations. The logarithm of Sales Performance was accounted for in the model in order to ensure normality. The average releases of the game developing firms included 16.62 releases with a standard deviation of 42.57. This indicates high fluctuations among game
developing firms in the number of produced games. On average, employees in the sample worked on 4.64 game releases with a standard deviation of 3.27, compared to a lower
average of 4.19 games at nascent firms. This indicates that nascent firms are not totally consisting of inexperienced employees, still they do have less Team PRWE. The variable Nth release is not included in the second sample as this model aims to investigate the role of PRWE in nascent ventures. The second sample therefore solely uses first released games by game developing firms in the sample. In the first sample, 29,6% of the games was released by one of the 10 selected most powerful publishers compared to 26,7% of the games of nascent firms. The difference between these numbers could be explained by that nascent firms may:
1) prefer to self-publish,
2) use a smaller publisher or simply
3) are not able to make a deal with one of the big publishers.
It should be noted however that the differences between these percentages is still small. The larger majority, respectively 82,8% and 86,5% of the all firms and nascent firms focused on the games genres marked as Popular. Moreover, both samples indicate that game developers tend to address their games to the larger public as the mean values lean towards “accessible to anyone” instead of adult only ESRB ratings. The descriptive statistics show a mean value of 2,22 and 2,11 on a scale of 1 to 7 ranging from “accessible to everyone” to “adult only”. Ratings of experts show a mean value of 5,99 while users on average rate the games slightly higher with a mean value of 6,35. Surprisingly, the ratings of experts and users on nascent firms games are rated higher, respectively 6,30 and 6,40. Possibly, there is a negative effect of high expectation by the users and experts when a game studio has released video games before. Moreover, team sizes of nascent firms are as expected smaller than team sizes of other ventures.
Table 5 and 6 furthermore provide information on the correlations between the presented variables. The Sales Performance variable correlates with all variables except for Release Number, implicating that the number of releases of a firm has no
relationship with Sales Performance. Team PRWE (r= 0.133, p<0.01) and Team PRWE Diversity (r= 0,114. p<0.01) both have a negative relationships with Sales Performance. Furthermore, the publisher dummy (r= 0.191, p<0.01) correlates to Sales Performance. This is possibly demonstrating the positive influence of these top 10 publishers on sales. Genre, Expert Rating, User Rating, Number of User Ratings and Team Size are all positively correlated to Sales performance (r= 0.096, p<0.05; r= 0.210, p<0.01; r= 0.146, p<0.01; r=0.299, p <0.01; r= 0.433, p<0.01). The ESRB Rating has the only negative correlation with Sales Performance (r= -0.143, p<0.01) of these variables. Similar results are found in the correlation matrix of sample 2 and are shown in table 6.
Table 5. Correlation Matrix Sample 1 (all firms)
Mean S.D. 1 2 3 4 5 6 7 8 9 10 1. Sales Performance (log) .051 .050 2.Team PRWE Diversity .801 .373 -.133** 3. Team PWRE 4.674 3.266 -.114** .517** 4. Nth Release 16.618 42.556 -.064 .004 .047 5. Publisher .296 .457 .191** .072 .075 .060 6. Genre .828 .378 .096* .049 -.006 -.247** -.008 7. ESRB Rating 2.216 .750 -.143** -.103** -.040 .286** .089* -.329** 8. Expert Rating 5.959 3.218 .210** .126** .068 -.146** .050 .124** -.251** 9. User Rating 6.335 1.792 .146** .017 .000 -.064 -.004 -.014 -.159** .491** 10. Number of User Ratings 39.76 163.006 .299** .095** .131** -.065* .085** .031 -.182** .209** .174** 11. Team size 201.031 194.007 .443** .304** .321** -.091* .208** .107** -.244** .150** .164** .062 ** P< 0.01 (two-tailed). * P< 0.05 (two-tailed)
Table 6. Correlation Matrix Sample 2 (nascent ventures) Mean S.D. 1 2 3 4 5 6. 7 8 9 1. Sales Performance (log) .045 .045 2.Team PRWE Diversity .786 .396 -.041** 3. Team PWRE 4.193 3.150 -.008** .550** 4. Publisher .267 .443 .108* .079 -.041 5. Genre .865 .343 .201** .094 .127* -.022 6. ESRB Rating 2.109 .737 -.151** -.034 .005 -.353** -.012 7. Expert Rating 6.303 3.069 .170** -.036 .043 .098* .152** -.249** 8. User Rating 6.398 1.881 .207** .052 .063 .158** .132** -.295** .549** 9. Number of User Ratings 58.38 296.828 .272** .104 .168** -.049 .058 -.065 .178** .111* 10. Team size 183.510 176.138 .420** .285** .330** .115* .255** -.278** .155** .125* .389** ** P< 0.01 (two-tailed). * P< 0.05 (two-tailed) 2. Statistical analysis
It was mentioned in the beginning of this chapter that two different samples have been analyzed in order to test the hypotheses. Before, we focused on the correlation among the different variables in the sample. Now, we will go into depth on the type of relationships between these variables with the hierarchical regression models. Multiple regression models were used in order to test the hypotheses. Five models test sample one and four models test sample two. This is due to the interest in the moderation effect of the Nth Release.
1. Sample 1: All firms
A hierarchical regression analysis was performed in order to thoroughly study the relationship between Team PRWE and Sales Performance. Another hierarchical regression analysis was conducted to study the relationship between Team PRWE Heterogeneity and Sales Performance, plus the proposed possible moderating effect of the number of releases a game-developing firm has released. The results of linear regression of the independent variables on Sales Performance are shown in table 6. According to these results, no evidence was found of an interaction effect of Nth Releases
between the relationship Team PRWE Diversity and Sales Performance. Hypothesis 1b is therefore not supported on the basis of this analysis. There is however evidence found (β= -0.071, p<0.05) for a negative effect of Team PRWE Diversity on Sales Performance of firms in the creative industries. Moreover, Team PRWE shows a significant negative effect on Sales Performance (β= 0.074, p<0.05). This is contrary to the positive effect suggested in hypothesis 2b. Hypothesis 2b is therefore also rejected on the basis of results by the analysis.
Other significant effects were also found in the analysis. A significant positive effect was found by Publisher on Sales performance (β= 0.271, p<0.01). A major
publisher therefore influence Sales Performance positively. There is also evidence for a positive effect on Sales Performance by Genre (β= 0.271, p<0.01). User ratings and the Number of User Ratings both significantly have a positive effect on Sales performance (β= 0.149, p<0.01; β=0.146, p<0.01). Consequently, higher levels and numbers of user ratings result in higher levels of sales. Moreover, evidence is found for a significant positive effect of Team size on Sales Performance (β= 0.300, p<0.01). This means that an increase in employees lead to an increase of Sales Performance at game-developing firms.
Table 7. Results of the hierarchical regression analysis (sample 2)
model 1β model 2β model 3β model 4β model 5β
(Adj. R square = 0.262) (Adj. R square = 0.266) (Adj. R square = 0.271 ) (Adj. R square = 0.276) (Adj. R square = 0.279) (Constant)
Team PRWE Diversity -.071* -.076*
Team PRWE -.074* Nth Release -.046 -.042 -.046 -.041 -.046 Publisher .271** .271** .271** .274** .271** Genre .163** .167** .163** .164** .162** ESRB Rating -.076 -.072 -.076 -.077 -.074 Expert Rating -.018 -.024 -.018 -.028 -.030 User Rating .142** .149** .142** .153** .152** Number of User Ratings .144** .146** .144** .145** .147** Team Size .274** .300** .274** .292** .293**
Team PRWE Diversity
x Release Number .054
** p < 0.01
* p < 0.05
2. Sample 2: Nascent Ventures
The role of Team PRWE and Team PRWE Diversity on Sales Performance at nascent ventures is also tested by a hierarchical regression analysis. The results of the analysis are shown in table 8. Both Team PRWE and Team PRWE Diversity show a significant negative effect on Sales Performance in the different models. Contrary to the hypotheses, these results show a negative effect of Team PRWE (β= -0.176, p<0.01) and a negative effect of Team PRWE Diversity (β= -0.185, p<0.01) on Sales Performance at nascent ventures. On the basis of these results, hypotheses 1a and 2a are rejected. Evidence was found for the opposite effect suggested in the hypotheses. Significant positive results are also shown by User Rating (β= 0.139, p<0.05), Number of User Ratings (β= 0.134,
p<0.05) and Team Size (β= 0.384, p<0.01) on Sales Performance. Publisher and Genre are both not as relevant for nascent ventures as for all firms.
Table 8. Results of the hierarchical regression analysis (sample 2)
model 1β model 2β model 3β model 4β
(Adj. R square = 0.204) (Adj. R square = 0.230) (Adj. R square = 0.223 ) (Adj. R square = 0.254) (Constant)
Team PRWE Diversity -.185**
Team PRWE -.176** Publisher .094 .100 .094 .099 Genre .061 .052 .061 .074 ESRB Rating .010 .027 .010 .022 Expert Rating .009 .004 .009 -.015 User Rating .128* .139* .128* .145* Number of User Ratings .128* .134* .128* .129* Team Size .325** .384** .325** .379**
** p < 0.01
* p < 0.05
Table 9 shows a summary of the researched hypotheses, together with the corresponding conclusions. The contrary effects are also included in the table. All hypotheses are
rejected.
Table 9. Conclusions for the hypotheses
Hypotheses Conclusion
1a. Increases in top management team PRWE heterogeneity will increase
performance of nascent firms in the creative industries.
Rejected
(negative effect found)
1b.
Time has a moderating effect on the relationship between top management team PRWE heterogeneity and the performance of firms in the creative industries.
Rejected
(no effect found)
2a. Increases in top management PRWE will increase performance of nascent
firms in the creative industries.
Rejected
(negative effect found)
2b. Increases in top management team PRWE will increase performance of
firms in the creative industries.
Rejected
5. Discussion
This paper was founded on the upper-echelon framework by (Hambrick & Mason, 1984), frameworks on team-composition (Kozlowski & Ilgen, 2006; Marks et al., 2001) and models on PRWE (Neal, 1995). A considerable amount of the literature on these frameworks stressed the complexity of these models. Many of these studies found evidence for mixed results in specific industries or specializations. There is a consensus in the literature on the positive effects of PRWE, mainly on the knowledge and skills of the individual (Neal, 1995).
Human capital is definitely a crucial factor for firms in the creative industry (Camelo-Ordaz et al., 2012). Human capital was expected to benefit from increased knowledge and skills by PRWE. On the basis of upper-echelon theory on TMT it was expected that mixed results were to be found for team diversity on performance (Steffens et al., 2012). Team heterogeneity in terms of functional background is expected to
improve complex task and creative thinking (Steffens et al., 2012). These results were also expected to be crucial in the creative industries. Furthermore, the need for team diversity was expected to decrease when firms matured over time in the creative industry because rationalization limited innovation (Tschang, 2007).
The first researched hypothesis focused on the effects of Team PRWE Heterogeneity on the performance of nascent firms in the creative industries.
Heterogeneous teams in terms of functional backgrounds are expected to perform better in creative and complex tasks. We did not find evidence for a positive effect on Team PRWE Heterogeneity. Instead, a negative effective of Team PRWE Heterogeneity was found on the performance of nascent firms in the creative industries. The same negative effect was found for all firms in the sample and no evidence was found of a moderation effect by time. This indicates an overall negative effect of Team PRWE Heterogeneity on firms in the creative industries, contrary to the expectations set in the theoretical
framework. However, the theory also suggests that mixed results are found by TMT diversity on performance due to the complex mechanisms associated with TMT diversity. Steffens et al (2012) give an explanation of mixed results due to the two concepts of emotional conflict and task conflict. Task conflict and emotional conflict are both enforced by team diversity. Evidence however shows that task conflict affects
performance by improved decision-making, creative thinking and multiple perspectives (Pelled et al., 1999; Watson et al., 1993).
The explanation for the negative effect according to the theoretical framework is twofold. First, either the video games industry or the creative industry as a whole does not comply with the causal characteristics for the benefits of task conflict. Particularly the sectors or industries that are characterized by high uncertainty and non-routine tasks benefit from task conflict (Amason, Shrader, & Tompson, 2006) . Also, the video game industry has grown to a mature industry in the last decade as before mentioned. Tschang (2007) mentioned that firms in the CI become more market oriented when they mature. In addition, creative products become increasingly expensive to produce,
especially for video games and films. This results in a rationalization of the production. Firms are therefore possibly forced to reduce creative thinking and adept task routines. Moreover, Mezias & Mezias (2000) mentioned the increasing importance of powerful players in the market, constraining innovation. We found evidence for a significant effect of the powerful publishers on sales in this study. This corresponds with the argument of Mezias & Mezias (2000). The large publishers arguably create an environment in which is less room for creativity and innovation according to this argument.
The second explanation for the negative effect of Team PRWE Heterogeneity originates in the distinction between task conflict and emotional conflict. While task conflict is argued to have a positive effect on performance, emotional conflict is argued to negatively influence performance. Increased homogeneity between functional and
demographic backgrounds in TMT increases harmony and stimulates good
communications (Ensley et al., 2002). Moreover, it reduces emotional or relationship conflict and results in more social interaction (Steffens et al., 2012). According to this argument, it is possible that there are positive effects of Team PRWE Diversity on performance. However, the negative effects of emotional conflict could possibly outbalance these effects. It is important to mention that this effect arguably does not work the other way around. While task conflict is considered to have clear benefits, emotional conflict always results in a negative effect on performance (Steffens et al., 2012).
Focusing on the case of nascent ventures, emotional conflict becomes of increased importance due to two reasons. First, teams of nascent ventures consist of less team members. Evidence for smaller TMTs in nascent ventures is found in this study. Therefore, team members are more dependent of each other and are also exposed to increased chances of emotional conflict. Theory on TMT of nascent ventures emphasized the importance of founders and owners during the start-up face as these founding teams control most activities from the beginning (Delmar & Shane, 2006) . Founding teams therefore carry more responsibility and possibly face higher pressure. Second, human capital is considered to be the most valuable asset of the firm at nascent ventures (Delmar & Shane, 2006) . The effect of emotional conflict on performance is therefore possibly of higher importance with nascent firms as there are less other resources to level out this effect.
The other two hypotheses researched the effect of Team PRWE on performance by following the mean value approach. Contrary to the proposed effect in the hypotheses on Team PRWE, we found a negative effect of Team PRWE on performance for all firms. Findings in previous researches presented mixed results of PRWE on performance. Recent studies however predominately focus on the positive results of PRWE. (Dokko et
