The Effects of Networks in Film Production – An Analysis of How Roles and Performance Criteria Affects
Utilization of Networks and Performance
Winter DRUID 2008 Kristina Vaarst Andersen
Copenhagen Business School Department of Innovation and Organizational Economics
Abstract The paper explores whether network positions lead to different network benefits depending on the role of the position holder and the performance criteria they aspire to. This perspective reduces social network benefits to a resource depending on the subjective valuation of the position holder. The influence of roles and performance criteria on the benefits of network positions is analyzed using novel data for the social networks and performance of the Danish film industry in the period 1995‐2005. The analyses show that roles affect performance criteria which in turn influence the utilization of networks. Roles and the corresponding focus on different performance criteria have greater impact on the probability of good performance on these criteria than the specific structure of the ego network. Different performance criteria cause different utilizations of network positions and hence the findings of this paper highlights that analyses and theories of social networks should pay more attention to the motivation of the people, firms etc. inhabiting the analyzed networks.
Acknowledgements: The data used in this paper is collected jointly by Professor Jesper Strandgaard, Associate Professor Mark Lorenzen, Lecturer PhD Lars Frederiksen, and PhD Fellow Kristina Vaarst Andersen. Thanks to Associate Professor Trine Bille, Professor Jesper Strandgaard and Associate Professor Mark Lorenzen for valuable feedback on an early presentation of this paper as well as to Associate Professor Toke Reichstein and CTO Morten Damsgaard for aiding the data handling process. I wish to thank the
Creative Encounters project at Copenhagen Business School for financial support.
1
1. Introduction Scholars exploring the power of embeddedness have shown in numerous ways how social networks influence both norms, options, and performance. A network structure is created through the ties between nodes which position them in the overall network of relations. The network can be seen as a system of positions, which are well defined but nevertheless subject to the enactment of the units occupying them. Through their utilization of network position, ties, and resources expectations are created and maintained. Baker and Faulkner (1991) define this as a role. Through roles individuals can choose to adapt to the expected capabilities and have the possibility of earning legitimization and acknowledgement through imitation of successful role models (Baker and Faulkner 1991). Through networks new information diffuse (Coleman et al. 1966), opportunities become available (Granovetter 1973), and new ideas are born (Burt 1992). While some studies emphasise the benefit of social networks as providing access to more or better resources through trust within a community connected through strong ties (for example Coleman 1988) others highlight the value of social networks as providing access to information and opportunities in many different social settings through weak ties (for example Burt 1992 and Granovetter 1973). Trust is mostly associated with strong ties in closed network settings dominated by common norms (Coleman 1988). Such trust is valuable for transferring and sharing resources (Jack 2005), exchanging sensitive information (Storper & Venables 2004), or relying on friendly co‐operation of partners extending contractual terms (Uzzi 1997). Weak ties on the other hand do not imply the same level of trust and commitment to a group, but are excellent channels for information distribution and endorsement (Granovetter 1973). The value of weak ties lies in the fact that they connect people and settings but due to the weakness they do not conform neither people nor settings to common standards and perspectives (Burt 1992). The research on social networks has so far been focused on how different network positions are linked to different types of network ties (strong versus weak) and how this is beneficial to individuals or companies (depending on the unit of analysis) with strong network positions on either dimension. However, the literature on social networks has so far neglected to contrast the value of network position with the nature of the analyzed performance criteria and the role a person partake in to reach this performance. This paper argues that the benefit of different network positions and the types of ties associated with them depends on role (the task to be performed) and performance criteria. Within a network different network positions can be the source of different benefits depending on the role of the position holder and the performance criteria they aspire to. The paper thus explore whether roles affect perception of performance criteria and how the form of participation and nature of accomplished task influence which performance criteria are valued the most. Further the paper explores whether roles and performance criteria influence the utilization of networks. One consequence of this would be that social network analysis is currently applying objective measures on social recourses of highly subjective value.
2
The analysis in this paper draws upon data from the Danish film industry. The data consists of information on all Danish films from the period 1995 to 2005. Each film production consists of a team of specialists utilizing their skills to fill different clearly defined roles. Role is in this paper defined as the primary task a person performs in the production process. The analysis focus on the seven most prominent roles in the production of a film: the director, the producer, the screenwriter, the five leading actors, the cinematographer, the editor and the composer. Networks can be studies at different levels of analysis: the system properties, groups of network nodes or individual nodes. In this paper the unit of analysis is the individual node which is a person and his/her role. Focus is on the persons involved in the production of movies and the roles they undertake1. The remaining paper is structures thus: section two introduces the film industry and its specific characteristics. Section three summarizes the criteria of performance in the film industry. Section four reviews the literature on social networks and relates this perspective to the creative industries and the film industry specifically. Section five briefly recaptures the focus and hypotheses of the analysis. Section six presents the data and variables. Section seven contains the statistical analysis and section eight concludes the paper with a brief summary and discussion of the findings. 2. The field The feature film2 industry is a setting of great uncertainty. Nobody knows whether a film will be a success or not until it is finished and is out there for the public to assess (Caves 2002). Furthermore, the product life‐cycle for films is extremely short. Films enter and exit the theatrical market continuously and compete against a shifting cast in the few weeks they have to capture the attention of the audience (DeVany & Walls 1999). To capture the attention of the audience a film needs the quality of novelty. Though the concept of a feature film is seldom challenged, all films need variety to distinguish them among the field of competitors (Lorenzen & Frederiksen 2007). Therefore all films are essentially innovations and the constant demand for new films is thus a constant demand for episodic product innovation (Delmestri et al. 2005, Lorenzen and Frederiksen 2007). The uncertainty of demand and the pressure for innovation calls for firm external organization – that way the best available, suitable, and affordable talent can be assembled for each film project (Uzzi & Spiro 2005). However, the quality of talent is hard to estimate beforehand and continuous commitment is almost impossible to ensure through contractual arrangement (Caves 2002), and thus talent is not suited to be traded on the market like standard commodities. Because of the combination of need for the heterogeneity offered on the market and trust ensures by hierarchy, film 1 Difficulties arise when persons take on other roles than their primary role – for example when a director writes the screenplay for a production he or she does not direct. These situations are solved by creating a new entry for these rare situations. The solution is chosen because of the focus on roles. This is a quite rare situation and as such does not present a problem for the analysis. 2 Hereafter the feature film industry is simply labelled the film industry and feature films labelled films.
3
production is neither market based nor integrated within organizational hierarchies but organized as a hybrid form (Williamson 1985, 1991). The production of films is typically organized in project networks (Faulkner & Anderson 1987). When formatting such networks, access to information and knowledge of fellow agents can to some extent replace the lacking knowledge of input quality previous to the exchange itself because reputation and previous relations can increase trust in delivery (Sorenson & Waguespack 2003). Therefore, project networks are not made up coincidently, but derive from the ties created through previous projects (Uzzi 1996). In these project networks agents with different skills come together to create an innovative expression – in this case a new film. Art and commerce is combined and the creative forces must accept the restraints of profit making (Caves 2002, 2003). The production of each film creates ties between the project team of industry agents working on that film and through these subsequent projects an industry network is created, which in turn affect the formation and configuration of new project networks (Grabher 2002, Wasserman & Faust 1994/1997, Uzzi 1996), and the formation and configuration of these project networks affects the probability of good performance (Sorenson & Waguespack 2003, Delmestri et. al 2005). As new agents enter the industry they become embedded in this industry network, and must do their outmost if they wish to make a reputation for them selves and achieve future success (Delmestri et. al 2005). The industry network defines the potentially available possibilities (Uzzi 1996) and has great effect on the innovative potential of an industry (Uzzi & Spiro 2005). The producer holds the role with the most bargaining power in Hollywood. In most European film industries the power is either divided between the producer and director or the most powerful role is occupied by the director (Delmestri et. al. 2005). Such variations are results of different institutional contexts, and this specific one is due to different approaches to the film industry: In USA the film industry is regarded a highly profitable exporting industry, whereas it is seen as a heavily subsidized cultural industry in most European countries (Bakker 2005, Henning & Alpar 2005). 3. Criteria of performance in the film industry This section deals with one important issue resulting from project based organization rather than hierarchy: discrepancy of performance criteria. In this section three crucial dimensions for varying performance criteria are explained: art versus business, focus on national versus international success, and focus on the whole narrative versus on specific tasks. Art versus business Though the basic assumption should be that everybody involved in the creation of a film wish to produce the best artistic product and generate the highest possible revenue with the available resources, the priorities of the persons entrusted with each of the seven roles might differ. The main aspect is that of art versus business. Delmestri et al. (2005) showed how the differentiation of task between director and
4
producer creates differences in network benefits. While producers benefit from vertical ties and network cohesion, horizontal ties are of greater importance to directors and strong ties have been shown to hinder creativity and innovation (Delmestri et al. 2005). The financial dealings of a film project rely on predictability while the creative part of a film production is unpredictable and dependent on the good ideas created through brokerage. The difference in perception of what constitutes good performance results in different utilization of means – which are in this paper operationalized as network relations. Focus on the artistic aspects versus on the business aspects of the film production depend on role and will affect the utilization of networks. National versus international focus For most European film industries international success is a rare thing in the present market for films (Bakker 2005, Henning & Alpar 2005). This is also the case for the Danish film industry though Danish film have experienced increasing international recognition for the past 10‐15 years (imagine.. 2005). However, some genres and roles have a better opportunity of achieving that international success. Comedies are often suited to the local taste and therefore they travel poorly (Delmestri et al. 2005). Participation in English speaking films is more or less a precondition for international success in the film industry. Hence, Danish actors face a substantial language barrier for an international career even if they participate in one of those rare Danish films which achieve international recognition. Even for the best European actors a national or at best partly international career is all they should aspire to. This is not the case for all of the roles. Directors for instance do not face the same massive language barriers. Therefore, focus on international success and an international career will vary between the different roles in a film production. Focus on the narrative versus on individual tasks In their seminal paper Baker and Faulkner(1991) emphasise the difference between on the one side the roles of the director, producer, and screenwriter, and on the other the remaining crew3. Their argument is that these three roles hold the coordinating functions and thus focus on the film as a unity whereas the other members of the production team will tend to focus on their specific tasks. This difference in task orientation might screw the perception of performance. And it must be assumed that the more specific the task, the less focus a role will have on the narrative elements of the film. 4. Social Network Theory Network ties signify prior interaction of some sort and network ties are therefore by definition a latent resource (Graber 2002) which shape the available opportunities (Uzzi 1996) – there is no guarantee of them ever being user. We can compare a system of network ties to a system of roads which increase the probability for future traffic (Graber 2002). Network ties thus increase the probability of future interaction.
3 Delmestri et al. (2005) also focus on the roles of director and producer while Elsbach and Kramer (2003) gives primacy to the initiating role of the screenwriter.
5
Theory on social networks is strongly linked to theory on transaction costs. The embeddedness of networks creates economic opportunities which are difficult to replicate via markets, contracts or vertical integration (Uzzi 1997). Networks can be described as a hybrid organizational form integrating aspects of both hierarchy and market (Williamson 1985, Coase 1937) by embedding economic transactions in social structures (Granovetter 1973, Uzzi 1996, Uzzi 1997). In the film industry trust, reputation and continuity of network relations diminish the otherwise extremely high transaction costs associated with assembling a project team and executing a project as complex and unpredictable as the making of a film (Caves 2003). On industry level the value of networks is the governance structure this repeated interaction create (Scott 2002, Uzzi and Spiro 2005). In the post studio‐system world of movie production such a structure reinforce a common code of conduct and thus trust and reliability. Network ties function as the social glue which enforce contracts and allow for contractual omissions of details, unexpected events, and level of commitment ‐ which are complicated aspects to specify and control. Two main lines of argumentation have evolved within social network analysis: the first one emphasising the value of centrality and reach ability (eg. Barabási 2003, Colemann et al 1957, Colemann 1988), and the second one stressing the value of network brokerage (eg. Burt 1992 and Granovetter 1973). The tension between the benefits of social proximity and social distance derives form the types of ties associated with these network positions. Whereas social proximity is correlated with strong ties and network closure, social distance is often created through weak ties spanning structural holds in the network structure. Both types of ties and network positions have been shown to improve the performance of individuals (Granovetter 1973, Burt 1992 and Coleman 1988), companies (Walker et al. 1997, Gulati 1999, Podolny 1994) and clusters (Storper & Venables 2004, Bathelt et al. 2002, Maskell et al. 2005, Andersen and Lorenzen 2007). In network settings dominated by social proximity all (or most) individuals are linked to each other in clusters through strong ties – the most illustrative example is that of the nuclear family. Strong ties are closely related to network closure – people connected through strong ties spend time together frequently and engage in interaction of some intensity (Granovetter 1973) and therefore often know each others social circles (the other ‘alters’ of each ego). Such network closure aids creation and reinforcement of common norms (Coleman 1988, Uzzi and Spiro 2005) and common ‘language’ (Storper & Venables 2004), and together with the trust embedded in this type of network the tightly knit web of linkages facilitate an easy flow of information (Coleman 1988) and act as a governance structure (Uzzi 1997). Sorenson and Waguespack (2005) show how network ties ensure better treatment and thus higher probability for good performance in the Hollywood film industry. The work of scholars such as Granovetter (1973) and Burt (1992) emphasise that tight knit clusters and cohesion are not the only network characteristics worth analysing, but that brokerage of social distance opens up new possibilities. Such ties spanning great social distances are most often weak ties – that is ties used infrequently, for
6
short durations and without much intensity of interaction. In his seminal paper on the strength of weak ties Granovetter (1973) shows how weak network ties add an individual’s opportunities for finding information about and actually getting new jobs both by the information flow through weak ties and the power of recommendation. This form of network ties does no enforce trust and common norms in the same way as strong ties. Their value lies in the social gap they span (Burt 1992). By bringing together different perspectives and perceptions they facilitate information spread and innovation (Burt 1992, Burt 2004). In the movie industry capabilities are gathered for each project and they are thus shared between the members of the industry. Hence, the crucial point of competitive advantage is not to develop core resources, but to know how to source them (Lampel and Shamsie 2003). This is due to the fact that resources – human capital – are highly mobile and increase the importance of the coordinating roles (Baker & Faulkner 1991). Thus network brokerage is in high demand but due to the problems of writing and enforcing ‘perfect’ contract (Caves 2003) so is a common code‐of‐conduct. Hence, the creation of each film project utilizes both weak and strong ties. Resources – network ties and roles Networks can be studied on different levels – the system, different groups of nodes, or individual nodes can be the unit of analysis. This paper focus on individual nodes which are in this case persons participating in projects teams making films. For individual nodes (persons, groups, companies and other units) social networks can be both a resource in itself and function as distribution channels for resources. Baker and Faulkner (1991) show how roles function as resources in two distinct ways: They provide membership and acceptance and the provide access to social, human and material capital. By adopting and complying with the acknowledged roles of the industry and the behaviour expected from the holders of such roles members enjoy the benefits of swift trust (Meyerson et al. 1996, Grabher 2002) and are therefore able to access resources within the network faster than outsiders. Roles relieve us from relating to each and every person as an individual – instead of taking the time and trouble to get to know them we can deal with them as holders of roles we know and have routines for dealing with (Meyerson et al. 1996). Roles will affect the benefit a person can gain form a certain network position. This is very evident in the film industry where the different role holders join a film project in different phases. The director, producer and screenwriter develop the film project after which a crew of actors, editors, composers, and cinematographers is assembled. One example of what this means for the difference in network value is the difference between a director and an actor. The network position and ties of a director and an actor is hardly utilized in the same way due to their (normally) very different point of entry into a movie project: while the director is part of the conceptualizing phase the actor is most often not casted until the project is structured. Thus the director depends upon his or her network in the active creation of the film project while the actor depends upon his or her network to spread a reputation in order to get invited onboard on attractive projects. Producers, directors, screenwriters, actors, composers, editors, and cinematographers work differently, value different aspects of the
7
production process and thus utilize their network differently. Plenty of research shows how past performance and connections from previous projects affect future possibilities (Faulkner and Anderson 1987, Sorenson and Waguespack 2005, DeFillippi and Arthur 1998, Scott 2002, DeVany and Walls 1999) however, these effects are very likely to depend on which roles participants undertake in the industry. 5. Focus of the analysis Based on the presented literature on the film industry and social network theory two hypotheses arise: Hypothesis 1: The benefit of a network position depends on the performance criteria. For example an ego network structure dominated by cohesion and network closure might benefit the producers need for smooth running business, while a network structure of brokerage might be more beneficial to a director Hypothesis 2: The benefit of a network position depends on the role held by ego. For example actors might benefit from different network positions than screenwriters due to their different utilization of their network ties – actors need invitations to the most promising projects, while screenwriters need assistance to make their ideas into projects. 6. Data The database4 consists of all Danish feature films produced between 1995 and 2005 – in total 208 films5. The database includes information on the financial and human capital involved in the production of each movie as well as different measures on performance. Information on the five most prominent actors, producers, directors, cinematographers, editors, screenwriters, and composers is used to create the social network of the film industry and calculate social network measures. A network tie is defined as a working relation. Because of the project structure of the film industry the industry network is thus made up of a multitude of perfectly interconnected clusters (the 208 projects) connected to each other through those agents who have participated in more than one project. Furthermore information on performance of the film projects measured on five dimensions is linked to each person. For the analysis in this paper the social network position of each individual prior to entering projects is compared to the performance of those projects. The eleven year period (1995‐2005) is divided into two sub period. Because industry networks
4 The database partly funded through the project “The social organization of film production – A comparative study of institutions, network structures, and performance in the Danish, US and Indian film industries” under the KINO-program “Creative Encounters”, supervised by Associate Professor Mark Lorenzen. 5 3 collections of short movies for children (Karla Kanin bio I, II and III) has been deleted due to lack of information and comparability.
8
develop over time the first five years (1995‐1999) are used to construct a basis network. The analysis of the relation between network position and performance is only conducted for the subsequent period. For each year the network position measures calculated based on the summarized network ties since 1995 are compared with the performance the following year for each person. This means that persons who only participate in one project are omitted since they one have values for one year. However, persons who are inactive for several years and the come back into the film industry are simply assigned the social network measures which their previous performance merits in the summarized network structure. Roles are taken into consideration. For persons who have participated in more than one project in a year the average performance of those projects is used. Variables Performance of movie projects can be estimated using several measures on commercial and artistic success. The absolute revenue or the number of tickets sold are often used to measure commercial success while award nominations and won awards as well as reviews by critics and audience are often used as measures for artistic success (e.g. Faulkner & Anderson 1987, Delmestri et. al. 2005, DeVany & Walls 1999, Sorenson & Waguespack 2003). The performance measures used in the analysis are:
• Admission within the EU: The number of tickets sold in the EU (25) excluding Denmark. This indicates the size of the revenue and is thus an indicator of economic performance.
• Admission in Denmark: the number of tickets sold in Denmark. This indicates the size of the revenue and is thus an indicator of economic performance.
• Danish awards: the number of national awards in the categories best actor, best actress and best director won in the two most prestigious national nominations (Bodil and Robert). One is controlled by the industry and the other by its critics. Awards are seen as acknowledgement of the artistic quality of the film and thus an indicator of artistic performance.
• Assessment by critics: The assessment by critics reported on in the International movie data base (imdb) on a scale from 0 to 100.
• Assessment by viewers: The assessment by viewers reported on in the International movie data base (imdb) on a scale from 0 to 100.
Social network measures are calculated for each node for each period (on the network created from 1995 and until that year). The social network measures used in the analysis are chosen based on their association with the two theoretical perspectives on the benefits of social networks presented in section 3 of this paper. The analysis includes measures on6:
6 Unless any other source is specifically named all measures are calculated by ucinet version 6 for Windows, and all explanations on measures are consistent with Borgatti, S.P., Everett, M.G. and Freeman, L.C. 2002. Ucinet 6 for Windows. Harvard: Analytic Technologies.
9
• Centrality: A measure of the degree centralization of each node based on the number of nodes adjacent to the node in a symmetric graph.
• Closeness: The farness of a node is calculated as the sum of the lengths of all the geodesics paths to every other node, the closeness centrality measure is the reciprocal of farness.
• Betweenness: Betweenness is a measure of the number of times a node occurs on a geodesic path. The normalized betweenness centrality is the betweenness divided by the maximum possible betweenness expressed as a percentage.
• Eigenvector: The eigenvector measure calculates the centrality of each node based on the centrality of the nodes they are connected to.
• Structural Holes: A measures based on redundancy and constraint. • Effective size: This measure is developed by Ronald Burt to analyse the
effective size of egoʹs network. For each ego this is calculated as the number of alters minus the average degree of alters within the ego network, not counting ties to ego.
• Efficiency: The above mentioned measure of effective size divided by the number of alters in egoʹs network to calculate the efficiency of the network – the outreach per tie.
• Constraint: This measure is also developed by Ronald Burt to analyse the constraints in ego’s network measured as the extent to which ego is invested in people who are invested in other of egoʹs alters.
• Hierarchy: An adjusted measure of constraint which is also developed by Ronald Burt. This measure analyse the extent to which constraint on ego is concentrated in a single alter.
• Ego Betweenness: This measures capture the share of ego’s alters which must pass through ego in order to connect to each other7
The measures centrality, closeness, betweenness, and eigenvector all reflect the theoretical perspective which emphasise the benefits of cohesion and the sociometric star. While the first three measures reflect the value of ego’s own position, the eigenvector measure tells us whether ego is directly connected to alters with powerful positions. The measures on structural holes, ego betweenness, effective size, and efficiency reflect the other theoretical positions within the literature on social networks: that brokerage is beneficial. While the first two measures are direct measures on the degree of network brokerage, the measures on effective size and efficiency tell us to which degree the network of ego is redundant. This is still consistent with Burt’s (1992) idea of network brokerage, but reflects a slightly different aspect. The measure on constraint and hierarchy are included in the analysis to show the possibilities for power restraint and pressure in the networks of the different roles.
7 From Everett and Borgatti (2005)
10
7. Analysis In this analysis the different measures on performance are correlated with the social network measures for each of the seven roles. The significant correlations are shown in figure 1. Figure 1: correlation of measures on performance and social network measures for each of the seven roles Admission EU Admission DK Award DK Assessment
critics Assessment viewers
Centrality Degree
Actors (‐,248**) Directors (+,479**)
Actors (+,236**) Actors (+,312**) Comp (+,357*) Editors (+,415*) Produce (+,355*)
Actors (+,261**) Cinematographers (‐,416*)
Betweenness Directors (+,684**) Actors (+,217*) Actors (+,335**) Produce (+,358*)
Actors (+,179*) Cinematographers (‐,416*)
Eigenvector Actors (‐,233**) Actors (+,202*) Actors (+,297**) Comp (+,446*) Editors (+,489**) Screenwriters (+,335*) Produce (+,347*)
Screenwriters (+,314*) Directors (+,360*)
Actors (+,246**)
Structural Hole Degree
Actors (‐,250**) Directors (+,474**)
Actors (+,236**) Actors (+,312**) Comp (+,358*) Editors (+,414*) Produce (+,353*)
Screenwriters (+,302*)
Actors (+,263**) Cinematographers (‐,415*)
Effective size
Actors (‐,222*) Directors (+,467**)
Actors (+,217*) Actors (+,306**) Editors (+,423*) Produce (+366*)
Actors (+,243**) Cinematographers (‐,420*)
Efficiency Actors (‐,310**) Actors (+193*) Cinematographers (‐,459*)
Hierarchy Directors (+,544**) Actors (+242*) Cinematographers (‐,481*)
Ego Betweenness
Directors (+,668**) Actors (+,199*) Actors (+,310**) Produce (+,418*)
Actors (+,192*)
Correlations are based on data from the film industry network from 1995 to 2004 and performance measures from 2005. The analysis shows a surprising lack of variation between the effects of the different types of network measures. However, the correlations show great differences in how the network measures of the different roles are related to success measured as the displayed performance criteria. These findings indicate that:
1. Persons with a strong network position measured by one network measure tend to have relatively strong network positions on other network measures. Within the Danish film industry some people thus have extensive networks while others have generally weak network positions.
11
2. Roles and the corresponding focus on different performance criteria have greater impact on the probability of good performance on these criteria than the specific structure of the ego network.
Each of the roles shows a different pattern of correlation between strong network positions and performance. For directors the assessment by critics is positively correlated with the eigenvector network measure. Performance measured as admissions in the EU is positively correlated with both the centrality, betweenness, structural hole degree, effective size, efficiency, hierarchical limitations of the directors network, and ego betweenness. The positive relation between the director’s network position and admissions in EU is partly due to the tendency for the Danish film industry to only promote big and nationally successful productions outside the national context. Such big productions often include a reputable name in the role as director. However, it is also an indication that the reputation of a director travels better than the reputation of other roles, and that the awards won by a film in international film festivals are awarded based on the artistic merits of the production and thus often ascribed to the director – the role in charge of coordinating the artistic side of a film production ‐ and affects his future reputation (Delmestri et al. 2005). A director with a strong network position can thus utilize this position to pull the movie project in the direction he or she desires – a direction focused on artistic values with an international scope. For producers network measure on both centrality, betweenness, eigenvector, the degree of structural holes, the effective size, and ego betweenness are positively correlated with performance. However the correlation is only for performance measured as received national awards for the projects coordinated by the producer. Furthermore, whenever the role of the producer is correlated with performance so is other central roles such as actors. This indicates that a strong network position enables a producer to assemble a strong team of roles and follow through projects acknowledged for their outstanding artistic performance. Furthermore artistic performance measured by received awards seems to be the result of this group dynamic. This is in line with the general assumption that no film production team is stronger than the weakest link (Caves 2002). Different forms of ego networks seem equally useful and hence both brokerage and network closure can be utilized to bring together a strong team and produce award winning films. An interesting aspect of the correlation between the network measures of the producer and performance measured as National awards is that the producer is not directly involved in the creative processes. Hence, the value of networks for the coordinating role of a producer is an effect of his or her ability to utilize networks to assemble a strong team. For screenwriters a positive correlation exists between the eigenvector network measure and the probability of the projects they participate in achieving national awards. Thus screenwriters are not dependent on having a strong network position themselves but benefit from being connected to the central players in the film
12
industry a far as getting their material turned into award winning films. The network position of the screenwriter is also positively correlated with performance measured as the assessment by critics. The assessment by critics is positively correlated with the network measure on eigenvector and degree of structural holes. Thus screenwriters benefit more from knowing central persons in the industry network and having ties to different parts of the industry network than any other network properties. The role on the screenwriter emphasises the narrative as well as the artistic aspects of the film and so do the critics. Hence, it is a logic consequence that a strong network position benefits screenwriters in getting invited to rewrite the best material (books etc.) and promoting their best work among the most accomplished persons in the film industry. The role as screenwriter thus utilize the network ties both for getting invited to participate in interesting projects by centrally placed people (the eigenvector effect) and for developing innovative ideas through bringing together persons from different parts of the industry network (the structural hole effect). The network positions of actors affect the probability of success on all the performance measure of the analysis except that of the assessment by critics. For actors the network measures on centrality, eigenvector, degree of structural holes, effective size as well as efficiency of the network is negatively correlated with the number of admissions in the EU. At the same time the network measures on centrality, betweenness, eigenvector, degree of structural holes, effective size of the network, and ego betweenness are positively correlated with admission in Denmark. The network measures on centrality, betweenness, eigenvector, degree of structural holes, effective size, and ego betweenness are also positively correlated with the probability of receiving an award. There are no correlations between the network positions of actors and performance measured as assessment by critics. The network measures of centrality, betweenness, eigenvector, degree of structural holes, effective size and ego betweenness are positively correlated with the assessment of viewers. The negative effect of actor network on admissions in EU is cause by the fact that mostly big, partly international or very successful national productions get promoted outside the national context, and these productions often involve non‐national actors who are consequently not centrally placed in the national film industry network – these people are in the periphery of the Danish national industry network though they belong to the core of other and bigger networks. An example is the singer/actress Björk who participated in the film ‘Dancer in the Dark’ as a star though she must be said to be very peripheral to the Danish film industry network. Actor network measures have a positive correlation with both national admissions and national awards – actors with strong network positions get offered the best jobs due to effective spread of their reputation from previous projects (which gave them the strong network position). The actor network position also has a positive correlation with assessment by viewers, thus, viewers tend to see productions with actors with good social network positions as better than other movies. Actors with high network measures attract the crowd – nationally. Actors thus utilize their networks for both information search on interesting projects and for spreading their reputation and increasing their brand value.
13
For editors three network measures are positively correlated with national awards. An editor’s position on measures on centrality, eigenvector and degree of structural holes in the industry network all have a positive effect on performance measured as received national awards. This correlation indicates that editors with a central position or links to others in the industry network with a central position as well as editor with ties to different parts of the industry network have more job offers and better opportunities for picking the best projects. However, the correlations between network position of the editor and performance only exists for network measures where several other roles are also correlated with performance. This highlight the role of the editor as editing the input from others – without a good team delivering good input the editor cannot perform well. Hence the editors with the strongest network positions will seek out projects with a strong cast populating the other roles in the film production. They thus utilize their networks for information search and for spreading their reputation in order to get invited to participate in strong projects. The network positions of cinematographers measured as centrality, betweenness, degree of structural holes, effective size, efficiency and hierarchy are negatively correlated with the assessment of viewers. Otherwise there are no other correlations between network measures and performance criteria. The negative effect of strong network position on some of the measures might be due to the technical focus of the cinematographers work. A possible conclusion is that the role of the cinematographer is more associated with the technical aspects of the film production process than the narrative and hence their criteria for attractive opportunities and good performance fall outside the scope of this analysis. The role of cinematographer may share more characteristics with technicians on a film production than with the creative staff and thus technical challenges and not artistic or business aspects influence their utilization and benefit of networks. To sum up the level of admissions in EU is only positively correlated with the network measures of directors while negatively correlated with the network measures of actors. This indicates that director reputation travels better than actor reputation. Thus Danish directors have the option of utilizing their network positions to create films with international potential and one way of doing so is to attract international star actors with their reputation as reputable directors (This line of argument follows the findings of Delmestri et al. 2005 on path dependency in utilization of reputation in the film industry and similar findings of Walker, Kogut and Shan (1997) from the biotech industry). The level of national admissions is only positively correlated with the network measures of actors. This indicates that actors with strong network positions have a good chance of their reputation being distributed and consequently of being invited to participate in strong projects. National awards seem to be the result of an all‐round strong team. Through the awards are given to either the director or leading actor/actress the network positions of the whole team except the director influence the probability of such artistic recognition. This indicates that the probability of achieving a high level of artistic performance depends on a strong network position and being invited to join projects composed by other roles with strong network positions for both actors, composers, editors, producers and screenwriters. Thus it is not surprising that one of the role for
14
which this correlation exists is the role of the coordinating producer. The assessment by critics is positively correlated with the network measures of screenwriters. This indicates that because the narrative elements of films are strongly valued by both screenwriters and critics screenwriters utilize their connections to powerful people and different parts of the industry network to promote material of high narrative value or to get in on projects with copy rights to exciting raw material. The assessment by viewers is positively correlated with the network measures of actors. This is probably an effect of the audience valuation of already well known actors. However the negative correlation between the network measures of cinematographers and assessment by viewers can be caused by a conflict of performance criteria – the general public might not value the technical details admired by technical connoisseurs. 8. Concluding remarks Though different roles must be assumed to utilize their networks differently due to the variance in entry point and focus of the role holders the analysis shows no systematic variance in which network positions are valuable to the different role holders. However, the analyses show that roles affect the valuation of performance criteria. Persons holding coordinating roles tend to utilize their network differently than task focused persons – this is best illustrated with the network position of role of the producer being positively correlated to performance measured as received national awards. Though the role as producer is not a creative function the coordination abilities of a producer with a strong network position is nevertheless crucial for artistic performance. The form of participation in film projects and nature of tasks must thus influence the perception of performance criteria for the different role holders. The time of entry into the projects and the focus the narrative versus specific tasks shape these differences. Roles thus affect both perception of performance and utilization of networks. Roles and the corresponding focus on different performance criteria have greater impact on the probability of good performance on these criteria than the specific structure of the ego network. Hence analyses and theories on social networks should pay more attention to the motivation of the people, firms etc. inhabiting the networks. One consequence of this would be that social network analysis is currently applying objective measures on social recourses of highly subjective value.
15
9. List of references Andersen, Kristina Vaarst & Mark Lorenzen (2007): The Stretching of Weak Ties. Paper presented at DRUID winter conference 2007. Baker, Wayne, E. & Robert R. Faulkner (1991): Role as Resource in the Hollywood Film Industry. The American Journal of Sociology. Vol. 97, No. 2, pp. 279‐309. Bakker, Gerben (2005): The Decline and Fall of the European Film Industry: Sunk Costs, Market Size, and Market Structure, 1890‐1927. Economic History Review. Vol. LVIII, No. 2, pp. 310‐351. Barabási, Albert‐László (2003): Linked: How Everything is Connected to Everything Else and What It Means for Business, Science, and Everyday Life. PLUME, New York. Bathelt, Harald, Anders Malmberg & Peter Maskell (2002): Clusters and Knowledge: Local Buzz, Global Pipelines and The Process of Knowledge Creation. DRUID Working Paper No. 02‐12. Borgatti, S.P., M.G. Everett and L.C. Freeman (2002): Ucinet 6 for Windows. Harvard: Analytic Technologies. Burt, Ronald S (1992): Structural Holes. Harvard University Press, Cambridge. Burt, Ronald S. (2004): Structural Holes and Good Ideas. The American Journal of Sociology. Vol. 110, No. 2, pp.349‐399. Caves Richard E. (2002): Creative Industries – Contracts Between Art and Commerce. Harvard University Press. Cambridge, Massachusetts, and London, England. Caves Richard E. (2003): Contracts Between Art and Commerce. Journal of Economic Perspectives. Vol. 17, No. 2, pp. 73‐83. Coase, Ronald (1937): The Nature of the Firm. Economica, Vol. 4, No. 16, pp. 386–405 Colemann, James, Elihu Katz &Herbert Menzel ( 1957): The Diffusion of an Invention Among Physicians. Sociometry. Vol. 20, No. 4, pp. 253‐270. Coleman, James S. (1988): Social Capital in the Creation of Human Capital. The American Journal of Sociology, Vol. 94, Supplement: Organizations and Institutions, pp. S95‐S120. DeFillippi, Robert J. & Michael B. Arthur (1998): Paradox in Project‐Based Enterprise: The Case of Film Making. California Management Review. Vol. 40, No. 2, pp.125‐139. Delmestri, Giuseppe, Fabrizio Montanari & Alessandro Usai ( 2005): Reputation and Strength of Ties in Predicting Commercial Success and Artistic Merit of Independents in the Italian Film Industry. Journal of Management Studies. Vol. 42, No. 5, pp. 975‐1002. DeVany, Arthur & W. David Walls (1999): Uncertainty in the Movie Industry: Does Star Power Reduce the Terror of the Box Office? Journal of Cultural Economics. Vol. 23, pp. 285‐318. Elsbach, Kimberly D. & Roderick M. Kramer (2003): Assessing Creativity in Hollywood Pitch Meetings: Evidence for a Dual‐Process Model of Creativity Judgements. Academy of Management Journal. Vol 48, No. 3, pp. 283‐301 Everett, Martin & Stephen P. Borgatti (2005): Ego Network Betweenness. Social Networks, Vol. 27, No. 1,pp. 31‐38 Faulkner, Robert R. & Andy B. Anderson (1987): Short‐Term Projects and Emergent Careers: Evidence from Hollywood. The American Journal of Sociology, Vol. 92, No.4, pp. 879‐909.
16
Grabher, Gernot (2002): Coll Projects, Boring Institutions: Temporary Collaboration in Social Context. Regional Studies, Vol. 36, No. 3, pp. 205‐214. Granovetter, Mark (1973): The Strength of Weak Ties. The American Journal of Sociology, Vol. 78, No. 6, pp. 1360‐1380. Gulati, Ranjay (1999): Network Location and Learning: The influence of Network Resources and Firm Capabilities on Alliance Formation. Strategic Management Journal. Vol. 20, No. 5, pp.397‐420. Henning, Victor &Andre Alpar (2005): Public Aid Mechanisms in Feature Film production: The EU MEDIA Plus Programme. Media, Culture, and Society. Vol. 27, No. 2, pp.229‐250. Imagine.., (2005): The Danish Film Industry – Annual Mapping 2005. Copenhagen Business School, Denmark. Jack, Sarah L.( 2005): The Role: Use and Activation of Strong and Weak Network Ties: A Qualitative Analysis. Journal of Management Studies. Vol. 42, No. 6, pp. 1233‐1259. Lampel, Joseph & Jamal Shamsie (2003): Capabilities in Motion: New Organizational Forms and the Reshaping of the Hollywood Movie Industry. Journal of Management Studies. Vol. 40. No. 8, pp. 2189‐2210. Lorenzen, Mark & Lars Frederiksen(2007 forthcoming, 2.nd draft February 12.th.): Why do Cultural Industries Cluster? Localization, Urbanization, Products and Projects. Lorenzen, Mark & Florian Taeube (2007forthcoming): Breakout from Bollywood: Globalization, Institutions, and Organizational Transformation in Indian Film Industry. Maskell, Peter, Harald Bathelt & Anders Malmberg (2005): Building Global Knowledge Pipelines: The Role of Temporary Clusters. DRUID Working Paper No. 05‐20. Meyerson, Debra, Karl E. Weick & Roderick M. Kramer (1996): Swift Trust and Temporary Groups. In Kramer, Roderick M. & Tom R. Tyler (eds.): Trust in Organizations: Frontiers of Theory and Research. SAGE. Thousand Oaks, California. Podolny, Joel M. (1994): Market Uncertainty and the Social Character of Economic Exchange. Administrative Science Quarterly. Vol. 39, No. 3, pp. 458‐483. Scott, Allen (2002): A New Map of Hollywood: The Production and Distribution of American Motion Pictures. Regional Studies. Vol. 36, No. 9, pp. 957‐975. Sorenson, Olav & David Waguespack (2006): Social Structure and Exchange: Self‐confirming Dynamics in Hollywood. Administrative Science Quarterly. Vol. 51, No. 4, pp. 560‐589. Storper, Michael & Anthony Venables (2004): Buzz: face‐to‐face contact and the urban economy. Journal of Economic Geography. Vol. 4, pp. 351‐370. Usai, Alessandro, Guiseppe Delmestri & Fabrizio Montanari (2001): Human Capital, Social Capital and performance: An Empirical Test from an Entrepreneurial Project‐Based Industry. SDA Bocconi, Research Division Working Paper, No. 01‐44. Uzzi, Brian & Jarrett Spiro (2005): Collabaration and Creativity: The Small World Problem. The American Journal of Sociology.Vol. 111, No. 2, pp. 447‐504. Uzzi, Brian ( 1997): Social Structure and Competition in Interfirm Networks: The Paradox of Embeddedness. Administrative Science Quarterly, Vol. 42, No. 1, pp. 35‐67.
17
Uzzi, Brian (1996): The Sources and Consequences of Embeddedness for the Economic Performance of Organizations: The Network Effect. American Sociological Review, Vol. 61, No. 4, pp.674‐698. Walker, Gordon, Bruce Kogut & Waijan Shan (1997): Social Capital, Structural Holes and the Formation of an Industry Network. Organization Science, Vol. 8, No. 2, pp. 109‐125. Wasserman, Stanley & Katherine Faust (1994/1997): Social Network Analysis – Methods and Applications. Cambridge University Press. New York. Williamson, Oliver E.( 1981): The Economics of Organization: The Transaction Cost Approach. The American Journal of Sociology. Vol. 87, No.3, pp. 548‐577. Williamson, Oliver E.( 1985): The economic institutions of capitalism: firms, markets, relational contracting. In Williamson, Oliver E. & S. E. Masten (1999): The Economics of Transaction costs. Edward Elgar Publishing. Massachusetts. Williamson, Oliver E. (1991): The Institutions of Governance. The American Economic Review, Vol. 88, No. 2, pp. 75‐79
18