PROJECT RESOURCES AND FIRM EXPERIENCE: A STUDY OF BOLLYWOOD FILM PROJECTS B y Da Huo A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Business Administration Strategic Management Doctor of Philosophy 2019 ABSTRACT PROJECT RESOURCES AND FIRM EXPERIENCE: A STUDY OF BOLLYWOOD FILM PROJECTS By Da Huo The Resource - Based View (RBV) has made a major contribution to the field of strategy over the last t hirty years. In spite of its growth in importance, the traditional RBV emphasizes the possession of VRIN resources as a sufficient condition for firms to achieve sustain able competitive advantage. In this dissertation , I show that an increase in the amount of VRIN resources that a firm may deploy could actually lead to diminishing returns. Barney later recognized t hat besides the mere possession of resources the firm might also p lay a role in creating an advantage, modifying VRIN to VRIO. However, this stil l did not clarify the characteristics of the firm that may also enable it to be better organized to draw upon i ts resources. I therefore this may allow it to derive more value from its available resources. I use data from Bollywood, one of the major centers of film production in India . By drawing on this data, I link the box office performance of motion picture projects to project - based resources and to firm - based experience. As such, I investig ate the extent to which benefits are embedded in the resources , as opposed to iii To my parents, whose wisdom, faith, and support always encourage me moving forward . iv ACKNOWLEDGEMENTS I wish to begin by thanking the members of my dissertation committee for all of their guidance and support. Thank you to Jamal Shamsie, my advisor, for his direction and encouragement through the di ssertation process and more importantly , for his guidance that he has provided me as a mentor through my doctoral studies . I am truly grateful for each of my committee members, Bob Wiseman, Gerry McNamara, and Kevin Miceli, who have given me valuable insig hts and supported me i n the development of this dissertation . I also thank the current and former faculty members, the staff members, and graduate students of the Department of Management at Michigan State for their support in my development over the past several years. I am grateful to have had the opportunity to work with them. In addition , I wish to thank Joe Mahoney, Chris Parsons, Janet Bercovitz, and Joey Engelberg , who have inspired and encouraged me to pursue this journey . Finally, I a m deeply grate ful . S he is caring, hard - working, smart , and wonderful in every way. Her love and support has been crucial for me through the Ph . D. program. I look forward to our future together. v TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........................ vii LIST OF FIGURES ................................ ................................ ................................ ..................... viii INTRODUCTION ................................ ................................ ................................ .......................... 1 THEORY ................................ ................................ ................................ ................................ ........ 5 The Role of Resource Access ................................ ................................ ................................ ...... 6 The Role of Prior Experience in Extracting Value from Resources ................................ ......... 11 Using Resources and Leveraging Experience in Project - based Industries ............................... 16 ASSESSING THE CONTRIBUTION OF RESOURCES AND EXPERIENCE ........................ 21 Effect of Resources on Performance of Motion Picture Projects ................................ .............. 21 Production Budget ................................ ................................ ................................ ................. 21 Top Stars ................................ ................................ ................................ ................................ 24 Production Partners ................................ ................................ ................................ ................ 27 Effect of Stud - Defined Experience on Performance of Motion Picture Projects .. 29 Broadly - defined Experience ................................ ................................ ................................ .. 30 Combining Broadly - defined Experience with Production Budget ................................ ........ 32 Combining Broadly - defined Experience with Top Stars ................................ ...................... 33 Combining Broadly - defined Experience with Production Partners ................................ ...... 35 - Defined Experience on Performance of Motion Picture Projects 36 Narrowly - defined Experience ................................ ................................ ................................ 36 Combining N arrowly - defined Experience with Production Budget ................................ ..... 38 Combining Narrowly - defined Experience with Top Stars ................................ .................... 40 Combining Narrowly - define d Experience with Production Partners ................................ .... 41 METHODS ................................ ................................ ................................ ................................ ... 44 Sample ................................ ................................ ................................ ................................ ....... 44 Variables ................................ ................................ ................................ ................................ .... 45 Dependent variable ................................ ................................ ................................ ................ 45 Independent variables ................................ ................................ ................................ ............ 46 Using film genres to st udy studio experience ................................ ................................ ........ 49 Control variables ................................ ................................ ................................ .................... 52 Estimation ................................ ................................ ................................ ................................ .. 52 RESULTS ................................ ................................ ................................ ................................ ..... 55 Descriptive Statistics ................................ ................................ ................................ ................. 55 Primary Analyses ................................ ................................ ................................ ...................... 58 Supplemental Analyses ................................ ................................ ................................ ............. 70 DISCUSSION ................................ ................................ ................................ ............................... 76 vi The Role of Resource Access ................................ ................................ ................................ .... 77 The Role of Prior Experience i n Extracting Value from Resources ................................ ......... 79 Limitations ................................ ................................ ................................ ................................ 82 Future Directions ................................ ................................ ................................ ....................... 84 REFERENCES ................................ ................................ ................................ ............................. 86 vii L IST OF TABLES Table 1 Data Measures ................................ ................................ ................................ ............... 54 Table 2 Correlations and Descriptive Statis tics ................................ ................................ ......... 59 Table 3 GEE Regression Coefficients and Robust Standard Errors Predicting Domestic Box Office ( with a 5 year time window of experience variables) ................................ ..................... 64 Table 4 GEE Regression Coefficients and Robust Standard Errors Predicting Domestic Box Office (with a 10 year time window of expereince variables) ................................ ................... 74 Table 5 O LS Regression Coefficients and Robust Standard Errors Predicting Domestic Box Office ( with a 5 year time window of experience variables) ................................ ..................... 75 viii LIST OF FIGURES Figure 1 Logged production budget and predicted values of logged domestic box office revenue with 95% confidence intervals ................................ ................................ ........................ 61 Figure 2 Number of stars and predicted values of logged dom estic box office revenue with 95% confidence intervals ................................ ................................ ................................ .............. 63 Figure 3 Number of pr oduction partners and predicted values of logged domestic box office revenue with 95% confidence intervals ................................ ................................ ........................ 63 Figure 4 Interaction Effect of Broadly - defined Experience with Number of Stars Predicting Domestic Box Office Revenue (logged) ................................ ................................ ....................... 66 Fig ure 5 Interaction Effect of Narrowly - defined Experience with Production Budget Predicting Domestic Box Office Revenue (logged) ................................ ................................ ....................... 67 Figure 6 Interaction Effect of Narrowly - defined Experience w ith Number of Stars Predicting Domestic Box Office Revenue (logged) ................................ ................................ ....................... 68 1 INTRODUCTION The Resource - Based View (RB V ) has made a major contribution to the evolving field of strategic management during the last thirty years ( e.g. , Wernerfelt , 1984; Barney, 1991; Peteraf, 1993; Mahoney & Pandian, 1992; Peteraf & Barney, 2003). It has even been claimed that RB V has become one of the most influential theoretical perspectives in the history of management theorizing (Kraa ijenbrink, Spender, & Groen, 2010, p. 350). It shift ed the focus of strategy research back from industry characteristics to firm attributes in explaining sources of economic rents and competitive advantage (Ho skisson, Hitt, Wan, & Yiu, 1999). In particular , this perspective claims that the simple possession of heterogeneous resources can enable the generation of Ricardian rents and quasi - rents (Lavie, 2006; Conn e r, 199 1 ; Peteraf, 1993; Mahoney & Pandian, 1992). Barney (1991) parameterized the RB V by arguin g that firms can achieve a sustainable competitive advantage (SCA) when their resources meet the criteria of being valuable, rare, inimitable, and without substitutes (VRIN) . As such, RBV placed considerable e mphasis on the possession of resources that pos sess these attributes as a sufficient condition for firm s to achiev e SCA. This also implied that firms could obtain much better performance by drawing upon more of these resources. Although studies have found some evidence of the contribution of VRIN resou rces, there has been little effort to examine whether the addition of such resources may eventually produce diminishing returns. In other words, firms may not be able to keep benefiting by increasing the amount these resources that they draw upon. In this dissertation, I argue that firms can enhance the returns they can receive from their VRIN resources by how they decide to use them. The RBV has not explicitly considered this issue. However, it is likely that a firm may derive advantage not only from its access to resources 2 but also how it decides to deploy these resources. In fact, firms are likely to differ substantially in their ability to extract value from their resources based upon their prior experience with their use . In other words, the manner in which resources are used by different firms reflects their accumulated exp eriences with prior use of resources. Prior research has of it has focused on the experience of a s (Holcomb, Holmes, & Connelly, 2009; Mannor, Shamsie, and Conlon ; 2016 ; Sonenshein, 2014 ). Furthermore, studies have produced conflicting findings about the influence of managerial or firm experience on the deployment of a t udies have examined the relative impact of the availability of resources and their use by a firm . I n other words, my dissertation explores the extent to which the performance of a firm may be driven by its access to resources or its use of resources. As such , the traditional RBV view that emphasizes the possession of VRIN resources as a sufficient condition for firms to achieve sustainable competitive advantage h as impeded our efforts to gain better insights into the RBV (Kraaijenbrink et al . , 2010). This results in a lack of understanding as to how firms can derive maximum possible value from available resources in creating competitive advantage . Although firms can benefit from the amount of VRIN resources that they can deploy, by themselves the advantage s that a firm can obtain from these resources will eventually decline. In fact, it has even been argued that resources themselves that are inputs into the production process, but only the services that the res ources can re nder , and it is largely in this distinction that we find the sources of the 25) . Specifically , taking the traditional RBV view that m erely having possession of VRIN resources can provide a firm with a n advantage (Barney, 1991; Wernerfelt, 1984; Peteraf, 1993) , 3 the question arises about the role of the firm. In particular , what aspects of the firm may enhance be tter leverage its resources remain unclear . Barney (1995) later recogni zed that the firm may also play a role in creating an advantage, modifying VRIN to V aluable, R are, Inimitable, and Organization (VRIO) . He stated that a firm must be organized to exploit th e full potential of its resources 56) . However, it still did not clarify the characteristics of the firm that may enable it to be better organized to draw upon its resources. ; 1995 ) assumption that if a firm has VRIN resources then it should have some understanding about ho w to use them . Following t his line of reasoning, t he firm will not matter if it has these resources and organize s them properly. As a result, there is s till pervasive criticism that this research overly emphasizes inherent value of resources instead of exp loring what role the firm play s in using its resources to achieve an advantage and what characteristics of the firm drive that ( Kraaijenbrink et al . , 2010 ; Priem & Butler, 2001) . In this dissertation, I aim to contribute to the RBV literature by examining the limitations of simply possessing and deploying VRIN resources. I shift my focus to the role of a experience in obtaining benefits from its resources because this can be important in that it leads to continu ou s development and refinement of rout ines to reflect its learning ( Nelson & Winter, 1982; Winter, 1995; March, 2010). I argue that firms can draw on certain type s of experience to better understand the value of various resources and how to enhance this value by figuring out how to use them. A s the firm is continually bundling and reconfiguring its resource base, it accumulates experiences with managing different re sources and gradually refines its mechanisms of using various resources (Eisenhardt & Martin, 2000; Shamsie , Martin, & Miller, 2009 ). 4 By bringing in allow it to make better use of its VR IN resources while also gaining a better understanding of how resources can provide a firm with an advantage. Specifically, i n expl oring different kinds of prior experience, I can further examine how the benefits to be obtained from VRIN resources can be enhanced by leveraging a firm can draw upon past experience to generate better performance from its resou rces. Moreover , it builds on recent work that increasingly emphasizes extracting value from managing available resources ( Holcomb, Holmes, & Connelly, 2009 ; Sirmon, Hitt & Ireland, 2007; Sirmon et al . , 2011). Con sequently, in this dissertation, I explore the different ways a firm can obtain benefits from its resources. Specifically, I s performance , how its prior experience influence s performance , and how the exp erience of the firm enhance s its ab ility to leverage value from its resources . I empirically test, separately and jointly , the relative contribution of each of these on the performance of projects in the Bollywood movie industry between 1997 and 2016. This dissertation is organized as follows. First, I make distinctions between different ways in which a firm can derive value from its resources. Next, I develop theoretical arguments regarding how the it may be abl e to obtain f rom its resources. As such, I investigate the extent to which benefits are embedded in the resources as opposed to relying on the firm to derive value from using them. Then , I develop hypotheses within the context of the Bollywood movie indust ry . In what f ollows, I outline the methodology used for testing my hypotheses as well as the findings from my analyses. I finish by dis cussing some of the limitations in this study and sharing my concluding thoughts on future directions. 5 THEORY The basis by which a firm pursues and sustains a competitive advantage has been one of the central themes in strategy research. Whereas the Industrial - Organizational (IO) view with Bain (1956, 1964) and Po rter (1980, 1985) focuses on the competitive environment of firms as determinants of performance of firms, the RBV complements this perspective by bringing scholarly attention back to firm attributes in explaining sources of economic rents and competitive advantage ( Kraaijenbrink et al., 2010 ; Hoskisson et al. , 199 9 ; Mahoney & Pandian, 1992; Miller & Shamsie, 1996). Specifically, the RBV emphasizes the heterogeneity of resources that firms have acquired and developed, relative to each other, as the internal sources of different performance outcomes. The central prop osition is that in order to achieve Sustainable Competitive Advantage ( SCA ) , a firm must o wn or control valuable, rare, inimitable, and nonsubstitutable resources and to apply them to a desirable end (Barney, 1991; 2002; Wernerfelt, 1984). In spite of cri ticism ( Priem & Butler, 2001; Foss & Knudsen, 2003; Kraaijenbrink et al . , 2010 ; Arend, 200 6), RBV research has drawn considerable attention over the years. Management scholars have devoted considerable effort in their theoretical work as well as in their e mpirical research on the development of the RBV ( e.g. Barney, 2001b; Peteraf, 1993 ; Petera f & Barney, 2003; Mahoney & Pandian, 1992; Lavie, 2006 ; Croo k, Ketchen, Combs, & Todd, 2008 ; Miller & Shamsie, 1996; Holcomb et al. , 2009; Mehra, 1996; King & Zeitham l, 2001 ). Despite their collective efforts, however, the answers to certain critical questions have remained unclear. To what extent can a firm increase its performance by simply increasing the amount of VRIN resources that it can access and deploy? Furthe rmore, can firms reply upon their previous experience with such resources to extract more value from them? In other words, how much 6 contribution is made by the resources themselves as opposed to the way they are used by a firm? In the following sections, I will brie fly review the literature and then explicate how these issues can be addressed. The Role of Resource Access It has been acknowledged by some researchers ( Kraaijenbrink et al . , 2010 ; Kor & Mahoney, 2000, 2004; Pitelis, 2004; Lavie, 2006) that the modern RBV has its roots in the The Theory of the Growth of the F irm. Conceptualizing firms as heterogeneous entities consist ing of firm - spe cific resource bundles, Penrose (1959) pointed out the basis by w hich firms could better pursue market opportunities and create economic value through effective use of idiosyncratic resources. In doing so , she highlighted how the resources that a firm may draw upon could provide it with a competitive advantage . Building RBV research by further articulating the importance of resource - position barriers and isolating mechanisms in secu ring and sustaining economic rents. The notion of a resource - posi tion barrier is an analogue of an entry barrier and may indicate the potential for high returns (Wernerfelt, 1984). Specifically, the heterogeneous resource bases of firms provide them with access to different market opportunities and growth , and a firm is able to exploit opportunities better and maintain its competitive position over peers by leveraging resource - position barriers and isolating mechanisms. Extending the idea that resource - position barriers and isolating mechanisms enable a firm to sustain economic rents by protecting it from imitation and substitution, Barney (1991) developed this research into a theory of SCA. Barney (1991) proposed four ch aracteristics of 7 resources essential for gaining SCA : valuable, rare, inimitable, and nonsubstitutabl e. Similar to the earlier formulation s of this perspective, Barney (1991) stresses that value is derived from using resources to exploit opportunities that possession of resources enables the firm to conc eive and implement strategies that improve its efficiency and effectiveness ( p. 101). As such placed considerabl e emphasis on the attributes of resources themselves. Collectively, these developments in the RBV ha s been insightful by shedding light on how resources enable the focal firm to exploit various market opportunities (Wernerfelt, 1984; Penrose, 1959) and to continuously reap benefits from these resources to more efficiently satisfy ing B arney, 1991; Peteraf, 1993). To this point, Peteraf and Barney resource - based theory , competitive advantage derives from firm - specific resources that are scarce ( unique resources are more 'efficient' in the sense that they enable a firm to produce more economically (p. 311). Thus, it was argu ed that the possession of heterogeneous resources could enable the generation of Ricardian rents (Peter af, 1993; Conn e r, 199 1; Mahoney & Pandian, 1992 ; Lavie, 2006). Furthermore, the formalization of the RBV has provided strategy scholars with important p arameters for subsequent theorizing and empirical testing. The empirical findings to date mostly suppor t the notion that differences in resources can explain inter - firm performance difference s (e.g. , King & Zeithaml, 2001; Miller & Shamsie, 1996; Mehra, 1996; Newbert, 2008 ; Holcomb et al. 2009; Huesch, 2013; Gruber, Heinemann, Brettel, & Hungeling, 2010 ) . I n particular, Crook and colleagues (2008) reviewed research in this vein and found that resource s are positively 8 related to firm performance and they found more support when resources , more clearly conformed to the RB V framework , are valuable, rare, inimit able and difficult to substitute . , however, emphasized complete control over resource s suggesting that resources must be owned by the firm 1 . This is reflected in the underlying assumption in the RBV research that resource s are immobile. Specifically, the RB V tradition tying resource heterogeneity with SCA builds on the assumptions that resources are lacking mobility (Peteraf, 1993) because of isolating mechanisms (Rumelt, 1984) and resource posit ion barriers (Wernerfelt, 1 984). What if resources are not immob ile or not completely owned by the firm? What would be the critical factors driving performance then? To put it differently, if a firm constantly draws on resources that were only made available for specific use s , can i t still achieve SCA? In this disserta resources as available assets to the firm, independent of the ways in which they were used. This conceptualization emphasizes the availability and accessibility of resources. In other words, resources are not necessarily owned or completely controlled by the firm in order to perform organizational tasks. In contrast, resources can be contracted from external sources for specific use by the focal fir m. In fact , Penrose (1959) noted that it is the services that resources can render that allow s a firm to pursue opportunities. Building on this argument, r ecent research has argued that focus ing only on resources that are owned by the firm undermines the essential contribution of mobile interconnected business environment (Lavie, 2006). For instance, Huesch (2013) finds that many cardiac surgeons work across multiple hospital s in a county , in light of heightened demand for 1 Som ewhat differently, Wernerfelt (1984 , p. 172) defines resources as those tangible and intangible assets, which are tied semi - permanently to the firm . 9 surgical pro grams in recent years. As work on RBV has progressed, scholars have started to lessen the emphasis on ownership of resources such that having access to resources on a semi - permanent basis may sim ply suffice (Helfat & Peteraf, 2003). As such , I follow this extended view of resource s that emphasizes the accessibility of resources that gives the firm right s to utilize and deploy resources across firm boundaries. Furthermore, the traditional RBV view that emphasizes the possession of VRIN resources as a suffici ent condition for firms to achieve sustainable competitive advantage also implies that firms can continu e to improve performance by adding more of these resources. However, it is questionable the extent to which firms can keep benefiting from an increasing amount of VRIN resources that it has access to. Th e previous argument build s on the assumption that these resources can always produce positive returns to the foca l firm by enabling the firm to produce more economically and to more efficiently satisfy cus Barney, 1991). What if adding more of these resources can no longer help to desires in the most efficien t manner ? Could the focal firm still derive benefits from adding more of these resources? In this dissertation, I challenge one of the commonly held assumption s in the RBV literature that having greater access to more VRIN resources can always enable the focal firm to improve performance . A lthough not explicitly specified in their theorizing, ma ny RBV theorists have either acknowledged or implicitly implied that more of these VRIN resources are likely to be better for firm performance. For instance, P enrose (1959) highlights the importance of various resources and suggests that higher level of th ese can help the firm to overcome growth barriers. Similarly, effect ive strategy to improve firm performance based on the RBV logic (p. 175). Furthermore, 10 Barney (19 91) argues that when a firm is able to accumulate more resources than its industry peers this can help it to gain a competitive advantage. Alternatively, as Pe teraf and Barney (2003) reiterated, competitive advantage is the result of having more valuable resources than other firms in an industry where there is a heterogeneous distribution of resources (p. 317). Similarly, Mishina, Pollock, and Porac assumed that more resources are usually better than f ewer resources in promoting firm 1180). Nonetheless , the argument of benefiting from economic efficiency through the use of more VRIN resources seems inherently paradoxical. As a focal firm continue s to allocate more resources to its project , the possibility of the marginal benefits to be deduced from additional resources may decrease due to the economic efficiency of prior resources all ocated to the project. Similarly , c onsumers may become less tuned to additional improvements in the underly ing products as the focal firm continue to escalate resources in refining the products . In other words, the economic efficiency and potential benefits to be derived from such additional resources decline as the amount of these resources increases. As such , I argue that there are limits to the benefits that can be derived from simply adding mo re of these resources. In other words, it is possible that adding more VRIN resources could potentially yield declining returns at some point . Although prior literatur e in the RBV has shown some evidence that VRIN resources could contribute to firm perfor mance, few studies have examined whether the addition of such resources may eventually lead to diminishing returns . Pierce and Aguinis (2013) describe such a phenomen on as the Too - Much - of - a - Good - Thing (TMGT) effect when beneficial antecedents cease to have additional positive impacts on the outcomes after inflection 11 points. Further, reaching beyond inflection points will produce suboptimal results due to waste or worsene d outcomes. Moreover, the asser tion that adding more VRIN resources, by themselves, can enable the focal firm to derive more value from its access to more resources assumes that if a firm has VRIN resources then it should have some understanding about how to use them (Barney, 1991; 1995 ). As such, t he role of the firm in extracting value from its resources is neglected. However, a firm plays a significant role in utilizing resources to obtain an advantage given its access to resources ( Kraaijenbrink et al . , 2010 ). The efficiencies and effectiveness of firm - specific processes through which resources are used to perform various organizational tasks have critical implications for firm performance. Certain firm attributes could set the binding constraint to its growth desp ite an increasingly larger pool of resources that it has access to (Penrose, 1959). Specifically , scholars argue that important firm attribute such as its prior experience can be leveraged upon in utilizing its resources to create value (Winter , 1995; Kor & Mahoney, 2004). In what follows, I elaborate upon the role of prior experience of a firm in extracting value from resources. The Role of Prior Experience in Extracting Value from Resources While resources of the focal firm contribute to the heterogeneity that may have rent - generating potential, the manner in which a firm may have learned to use them can also play a significant role (Collis, 1994; Amit & Schoemaker, 1993; Kor & Mahoney, 2005) . In this regard, scholars argue the prior experienc e of a firm with the effective use of resource s can also be an important source of heterogeneity that may also lead to sustained competitive advantage (Winter, 1995; Huesch, 2013; Kor , Mahoney , & Michael. , 2007). Whereas resources are usually tied to Ricar dian rents due to their value and scarcity, quasi - rents are contingent upon the best possible 12 use of these resources within the firm (Klein, Crawford, & Alchian, 1978; Ethiraj , Kale, Krishnan, & Singh, 20 05; Mahoney & Pandian, 1992 ). Specifically, having m ore prior experience might available resources ( Cuypers, Cuypers, & Martin, 2017 ; Taylor & Greve, 2006 ). This is because learning can arise from the accumulated experience and creates a growing stock of kn owledge that can be ref lected in its ongoing resourc e use . It occurs as the firm encodes experience into routines that guide its future behaviors (Levi tt & March, 1988 ; March, 2010 ). The notion of extracting value from resource s s (1959) original thesis in which she states that resources are closely tied to a bundle of potential services that provides differential value to firms. In other words, even if resources were available to all firms, the extent t o which a firm can extract value from these resources would not be the same. This is because the same sets of resources may be combined in different ways, applied for different - dependent processes of using the se re sources (Penrose, 1959). Nelson and Winter (1982) further reiterated this path - dependency nature of using resources that is tied closely They proposed that a s and outcome s reflect a collection of firm - specif ic routine s that are derived from the past experiences of the firm (Mahoney, 2004 , p. 186) . Because firms can vary in the accumulation of their experience with the use of resources, these can be difficult to imitate and therefore can be a crucial source of performance differences among firms (Barney, 1991; Dosi, Nelson, & Winter, 2001). As such, scholars have shown that differences in the prior experience of firms in making use of various types of resources can lead to differences in their performance (e.g. Huesch, 2013; 13 Kor & Mahoney, 2005; Yeoh & Roth, 1999; Moliterno & Wiersma, 2007; Danneels, 2002; Gruber et al., 2010; Lampel & Shamsie, 2003). Nonetheless, there appears to be a lack of consensus regarding the extent to which resources, by themselves, in fluence firm performance (Armstrong & Shimizu, 2007). For instance, in his review of the emp competitive ad vantage to a far greater extent than do resources themselves and suggests this may be due to different theoretical approaches conceptualizing resources. Similarly, there are different viewpoints regarding the degree to which prior experience of a firm mig ht contribute to its performance. Specifically, whereas some studies suggest that firms may gain an advantage from their accumulated prior experience (e.g., Egelman, Epple, Argote, & Fuchs, 2016; Fong - Boh, Slaughter, & Espinosa, 2007; Kor et al., 2007), ot hers have argued that sometimes experience may backfire and that it hinders creativity and c ould lead to undesirable performance (Taylor & Greve, 2006; Lawrence, 2018; Haas & Hansen, 2005). This reflects some critical issues that have continued to confound strategy scholars. To what extent can the performance of a firm be merely explained by the inherent characteristics of resources themselves? Moreover, when could prior experience of a firm enhance its ability to better leverage its resources? Taking a sli ghtly different tone, Winter (1995) argued that despite the importance of resources, resourc es themselves are insufficient contribut ors to achiev ing competitive advantage if there is a lack of necessary routines based on prior experience of coordinating an d deploying resources. Despite the attributed importance of access to experience with using resource s , there have been few studies that have untangled the different 14 ways that a firm obtain s benefits from its resources . This makes it difficult to ascertain the precise role of either resources or their use on firm performance. It also results in a lack of understanding of how resources can be better leveraged to achieve organizational goals on the xperience. This omission is glaring given that firms accumulated experience is assumed to allow firms to generate the maximum p ossible value from their resources. Furthermore, i obtain value from available resou rces reflects the development of routines that are tied to its accumulated experien ce which determine its idiosyncratic way of using resources ( Nelson & Winter, 1982; Levitt & March, 1988; Winter, 1995; March, 2010 ). Notwithstanding these helpful insights in conceptually distinguishing between access t o resources and its idiosyncr atic use of resources based on prior experience , few empirical studies have examined the relative impact of each of these on performance. Instead, most studies in this vei n have continually lumped the availability of resources with their use ( Newbert, 200 8; 2007 ; Armstrong & Shimizu, 2007; Lavie, 2006 ; P eteraf & Barney, 2003). Indeed, making the distinction between access to resources and the use of resource s represents a c hallenging endeavor given the conceptual interrelatedness of the two constructs. On the one hand, firms draw on their prior knowledge about how to use resources effectively to obtain more value from resources (Kor & Mahoney, 2004) . Firms use resources on t he basis of the emergence of routines that are derived from their past experiences (Nelson & Winter, 1982). On the other hand, the res ource base of a firm co - evolve with its accumulated experience with managing different kinds of resources as the firm cont inually develop s , integrate s , and reconfigure s its resource base (Eisenhardt & Martin, 2000). 15 Empirically, it is even more difficult to separate the contribution of resource access from their use of different types of resources and distinct mechanisms through which resources are used bring additional complexity to this undertaking (Armstrong & Shimizu, 2007). However, as argued earlier, it is critical to explore the different ways that a firm can obtain value from its resources by distinguish ing between its access to resources and its use of resources to achieve a better understanding of their relative roles in explaining economic rents and competitive advantage. In other words, t he value that a f accumulated experience with their use in various activities . Specifically, firms gradually accumulate experience and gain knowledge over time from ongoing tasks about how to make prope r use of availabl e resources (Mahoney, 2004; Nelson & Winter, 1982; Nag & Giola, 2012). Drawing from their accumulated experiences, firms manage and deploy resources to perform various functions (Shamsie & Mannor, 2013; Sirmon et al., 2011; Peteraf & Barne y, 2003). As such , it transforms resources as raw inputs into desirable outcomes in distinct markets and to execute various organizational tasks following the - based processes. As Argote and Miron - Spektor (2011, p . 1124) noted , is what transpir es in the organizations as Furthermore, a firm can rely on its prior experience with the effective use of various resources in managing all av ailable resource stocks . For instance, contracted resources can also be used for generating economic value as long as the focal firm can effectively leverage I apply this distinctio n between the valu e that is derived from resources and the value that stems from the 16 accomplished within a project - based industry where specific resources have been assigned to each particular pr use the resources can be deduced. In other words, we can separately identify and more accurately capture the different roles of distinct project - based resources and firm - based experience in a project setti ng. Using Resources and Leveraging Experience in Project - based Industries In a project - based industry, such as motion picture, video games, television, music production, and architecture, a certain line - up of specific resources are usually assigned explic itly to complete a particular project (Shamsie et al., 2009; Garzon - Vico, Gibbons, McNamara, & Rosier, 2016; Ethiraj et al., 2005). In addition, the resources assigned to any individual project may not necessarily come from within the firm. For example, th e production budget assigned to a giv en project may come from various funding sources. The key productive individuals assigned to a project may come entirely from outside the firm such as the members of a research team that have been recruited specifically to carry out a particular project. In fact, resources that are contracted or arranged from outside sources play more and more important roles across different industries (Huesch, 2013; Lorenzen & Frederiksen, 2005; Miller & Shamsie, 1996; Aime, Johnson, Ridge, & Hill, 2010). Thus, a project - based setting can more fully capture the availability of various resources by building upon the recent development of RBV research that has been recognizing the roles of resources that may not be permanently attached t o the firm (Lavie, 2006; Aim e et al . , 2010 ; Fang, Wade, Delios, & Beamish, 2007; Peteraf & Barney, 2003). Specifically, I focus on three kinds of resources , financial resources, human resources, and partner resources , that have been argued to play critical roles in project - based industries ( Lorenzen & Frederiksen , 2005; Shamsie et al ., 2009 ; Vandaie & Zaheer, 2014 ) . In particular, a 17 firm needs to secure financial resources that determine the amount of funds that can be allocated to its projects. In addition , a firm should also ass ign human resources that include specific types of talents to perform various tasks associated with its projects . Moreover, a firm may develop various partnerships that it can leverage on a given project. Whereas resources assigned to a specific project m ay be drawn from outside of a firm in a project - based setting, firms can leverage their own prior experience on different projects to effect desirable outcomes. This is consistent with the notion that resources from external sources also have the potential to create value if the focal firm can draw on its own experience to deploy and transform these resources to exploit opportunities. In fact, firms gradually gain experience from a range of activities and draw from such accumulated e xperience to improve the ir future actions (Choi & McNamara, 2017; Acquaah, 2012; Shamsie & Mannor, 2013) . Because it is possible to identify the various resources that are assigned to any given project, we can then explore how these different resources in fluence the performance of those projects. We can also examine the role of prior experience of a firm on its project performance. In other words , the manner in which various product markets contrib ute to performance can be assessed relative to each other in project - based industries. It also allows us value o f its resources enhanced by its prior experience. accumulated experience can guide the processes through which different resources are used in various projects to deliver desirable outputs ( Nerkar & Roberts, 2004; Yeoh & Roth, 1999; Mahoney & Pandian, 1992; Penrose, 1959). Firms gradually accumulate knowledge throug h learning by doing in performing various projects and embedding acquired knowledge about how to use different resources in firm - specific , path - dependent 18 routines (Ne lson & Winter, 1982) 2 . Drawing from its accumulated experience, firm - specific routines coe volve over time with revised practices and improved production processes in pursuing various goals set forth in different projects (Clark, Kuppuswamy, & Staats, 2018) . Because firms vary significantly in their accumulation of prior experience, this is an i mportant source of inter - firm heterogeneity that can lead to different levels of performance in various projects (Barney, 1991; Winter, 1995; Haunschild & Sullivan, 20 0 2; Yeoh & Roth, 1999). In other words, a of resources assigned to particular projects (Huesch, 2013). This is because the extent to which resources can generate value is contingent upon effective use of resources based on different kinds of a with transform ing inputs to outputs in different types of projects . Furthermore, scholars have argued that experience is not only firm - specific but also context - specific (Nelson & Winter, 1982; Argote & Miron - Spektor, 2011) . Thus, it is important to explore the different kinds of experienc e in different contexts in order to better understand their relative roles in contributing to firm performance (Argote, McEvily, & Reagans, 2003; Clark et al., 2018; Lawrence, 2018 ). In fact, research has shown that different kinds of firm experience lead to different learning outcomes and may have different impacts on firm performance ( Fong - Boh et al. , 2007; Haunschild & Sullivan, 2003; Eggers, 2012; Ingram & Baum, 1997). Specifically, I distinguish between two kinds of firm - based experience, b roadly - d ef ined e xperience and n arrowly - d efined e xperience, in this setting. In project - based industries, a firm gradually gains broadly - defined e xperience with managing all kinds of projects. For instance, an 2 . In addition to the largely tacit accumulation of knowledge embedded in learning by doing in each cont ext, firms also make deliberate efforts in refining their routines and practices to improve their resource deployment (Zollo & Winter, 2002), such as an IT firm can make persistent investments in infrastructure to improve its ability of managing multiple p rojects (Ethiraj et al., 2005). 19 architecture firm gains broadly - defined experience with d esigning different styles of structures such as residential building, office building, shopping malls, or sports stadiums. Accumulating broadly - defined experience with managing projects of all kinds triggers generalized learning which creates a broader kno wledge base of the firm (Haunschild & Sullivan, 2003 ; Huber, 1991 ) . L everaging such broadly - defined experience with managing different kinds of projects , firms become aware of the potential benefits in each of these resources, learn to better use them unde r different situations , and deploy them in the best possible ways to extract value in different contexts (Mannor, Sh amsie, & Conlon, 2016; Egelman et al., 2016; Ethiraj et al., 2005; Holcomb et al., 2009). This may further contribute to project performance than do the inherent characteristics of those resources. In other words , broadly - defined experience with managing p rojects of different kinds can enhance ability to derive value from resources assigned to its projects. Whereas broadly - define e xperience describes the level of experience of the firm across different domains, narrowly - managing projects of a certain kind. For instance, we can look at Pixar, an Ame rican film studio that f ocuses on producing animation films and accumulates such narrowly - defined experience with managing animation film projects . Furthermore, n arrowly - defined experience typically reflects deliberate choices a firm makes that are involve d with specific purpose s to address opportunities within a certain market (Eggers, 2012) . It gives the firm more opportunities to develop specialized learning and expertise ( N arayanan, Balasubramanian, & Swaminathan, 2009) . However, having more narrowly - defined experience might o nly help to a certain extent because repeating similar projects will limit the opportunity that the firm can learn from performing different tasks (Staats & Gino, 2012) . ty to 20 search for more innovative use of resourc es assigned to its project (Schilling , Vidal, Ployhart, & Marangoni , 2003; Abernathy & Wayne, 1974). Stated differently , narrowly - defined experience with managing project s within a narrow range may undermine the potential of the firm to derive value from r esources assigned to its projects. To summarize, the different ways that a firm can obtain value from its resources are conceptually clear and can be assessed , relative to each other , empirically in project - based industries. The access to VRIN resources o f a firm and its accumulated experience of different kinds can both influence performance. Unpacking the mechanisms through which they create value can allow us to achieve a better understanding of the specific contribution of each of these to the overall performance of each individual project. I propose that acce ss to VRIN resources may be able to contribute to firm performance, but these additional resources may produce diminishing returns. Instead, a firm may be able to draw on its prior experience defin ed broadly or narrowly on the basis of genres of the previo us films to enhance the value that it obtains from available resources . In the next section, I will develop these ideas within the context of the motion picture industry. 21 A SSESSING THE CON TRIBUTION OF RESOURCES AND EXPERIENCE This dissertation looks at motion picture projects and identifies three key resources that are attached to these projects namely production budgets, top stars, and production partners. I begin with an assessment of t he relative contribution of each of th ese resources on the performance of film projects. Next, I - defined experience on performance of its film projects and the role of broadly - defined experience on the relationship between various resources and overall performance of film s . Finally, I explore the role of a - defined experience and its impact on the relationship between project resources and film performance. Effect of Resources on Performance of Moti on Picture Projects Production Budget The size of production budget is the first critical consideration of resources assigned to any particular film project. It has long been argued that the production budget assigned to a film project is one of the most important factors for its box office performance (Litman, 1983; Litman & Ahn, 1998; Elberse & Eliashberg, 2003; Perren & Schatz, 2004). Despite certain evidences suggesting that big budgets are not necessarily correlated with greater financial returns (e.g . Ravid, 1999; John, Ravid, & Sunder, 2002), the average production budget for a Bollywood film has continually been on the rise. Between 2000 and 2014 the average budget of a Bollywood motion picture has risen from a little above 3 crore ( approximately 42 0 , 000 USD) 3 to well over 25 crore ( approximately 3.5 million USD) . In 2014 alone, high budgeted films included Happy New Year costing 150 crore ( approximately 21 million USD) , PK costing 120 crore 3 1 crore = 10, 000, 000 rupee; or approximately 140,000 USD based on the exchange rate as of 4/29/2019 22 ( approximately 17 million USD) and Singham Returns costin g 110 crore ( approximately 15.5 million USD) . Although the production budget of a film can be tied to the use of expensive actors and actresses, Basuroy, Chatterjee, and Ravid (2003) convincingly argued that the influence of production budgets on project performance should be disentangled from the effects of stars because the two have different mechanisms affecting moviegoers 4 (I elaborate upon these two mechanism in sections that follow) . H igher production budgets can allow the studio to expand on the var ious elements that it can incorporate into its films in order to attract more audiences and increase box office performance. Specifically, s cholars of motion picture economics argued that bigger budgets could Ahn, 1998), indicating higher technical quality and thus may lead to greater box office popularity (Litman, 1983). Basuroy and colleagues (2003) further contend that assigning bigger budgets to film projects can serve as an insurance policy by applying sp ecial techniques to attract audiences since studios find it hard to predict the market trend. More specifically, the bigger production budget assigned to a film project also showing exotic on location scenes, magnificent sets and costumes, and dazzling visual effects. However, I argue that such added benefits associated with assigning bigger budgets are likely to yield diminishing marginal returns. As the studio keeps spendi ng more and more in production by building in more actions, shooting at more locations, and adding more sp ecial effects, moviegoers become less able to detect additional improvements in the picture quality or might not even notice any marginal changes. T he RBV suggests that resources could provide a 4 Basuroy and colleagues (2003) found minimal correlation between the production budget and the presence of stars. 23 firm with an advantage as long as they can enable the firm to effectively (Peteraf & Barney, 2003) . If the customers do not even notice the changes in entertainment value, the ben efits that a firm can derive from increase s in the production budget of a film are likely to show diminishing returns . As such, I predict the following: Hypothesis 1 a : Increases in the production budget will lead to diminishing marginal returns in box off ice revenue. Furthermore, I argue that an increase in the size of the production budget assigned to films projects will eventually yield negative returns in terms of box office performance of these films. Specifically, excessive production budgets assigned to film projects may lead to inappropria te use of the funds. In other words , an excessive amount of financial resources might potentially lead to managerial opportunism . Specifically, agency theorists argue that excessive financial resources can lead to s elf - serving and value - destroying behaviors of the managers (Jensen & Meckling, 1976). As such, keep increas ing production budget not only yield s inefficient use of financial resources but could also lead to inappropriate application of financial resources that might actually make the film less appealing to potential audiences. More specifically, assigning too much production budget to a film could reduce the overall consumption experience of moviegoers from watching the film that it confuses viewers from f ollowing the plot and storyline of the film. Specifically, spending on extravagant settings and shooting across too many different locations could distract the attention of viewers. In other words, adding too many features and specia l effects to a film cou ld potentially overwhelm moviegoers. In fact, scholars noted that an excessive amount of information could turn consumers away (Malhotra, 1982; 1984; Lee & Lee, 2004). 24 Finally , excessive spending in film production by the studio in s hooting across differen t locations and featuring more special effects will incur additional complexity of these projects and thus make it more challenging to manage these projects. Specifically, spending on extravagant settings and shooting on - location set s require coordination among various specialist s. I n fact, recent research in project management has shown that greater complexity of projects will lead to more delays and worse performance of project s (Cicmil, Cooke - Davies, Crawford, & Richardson, 2009; T atikonda & Rosenthal, 2 000; Bjorvatn & Wald, 2018). As such, I predict the following: Hypothesis 1b: Increases in the production budget will eventually lead to negative returns in box office revenue such that there is an inverted - U shaped relationship . Top Stars The role of po pular stars on the performance of a film has attracted a great deal of scholarly attention in motion picture research (e.g. Litman & Kohl, 1989; Wallace, Seigerman, & Holbrook, 1993; Desai & Basuroy, 2005). In Bollywood, top stars such as Ashkay Kumar, Sha hrukh Khan and Salman Khan have been dominant for a few decades. Nonetheless, there appear to be conflicting viewpoints about the influence of stars on film performance since scholars have long debated on whether or not studios sho uld hire top stars (Litma n, 1983; Sochay, 1994; Levin, Levin, & Heath, 1997; Elberse, 2007; Liu, Mazumdar, & Li, 2014). On the one hand, some empirical evidences suggest that the presence of stars in a film do es not necessarily guarantee greater financial returns (Litman, 1983; De Vany & Walls, 1999; Litman & Ahn, 1998; Ravid, 1999). On the other hand, other scholars argue that the presence of stars in a film have De 25 Silva, 1998). In particul do (p. 23) . Top stars have widely recognized acting talent that are critical i nputs to filmmaking (Elbe rse, 2007; Sochay, 1994). In addition, they also bring additional appeal to a broader fan base at the box office . Levin and colleagues (1997) showed in an experiment that popular stars provided moviegoers with a decision heuristic for choosing whether or n ot to watch a film before considering other information. Albert (1998) further contended that stars were important not only because they could bring additional appeal to the fans, but also because their presence may signa l potential audiences to categorize the film as having better quality drawing from previous experiences. Although featuring top stars in a film project may enhance the public appeal of the film at the box office, the marginal benefit associated with featur ing additional stars in a film is l ikely to diminish . In fact, s ome of the top stars are likely to have overlap ping bases of fan s . Thus, adding more top stars does not add much more additional box office appeal. In other words , adding more top stars does not produce proportionally more sign als to the film to attract more pote ntial audience. Moviegoers may have collected enough information with respect to top stars featured in the film upon making a decision whether or not to wa tch the film. Gonzalez and Madhavan (2011) showed in an experimen t that individuals are less likely to pay attention to additional stimuli that are similar to what they have observed simultaneously . Therefore, the added benefits are likely to diminish with additional increase in top stars. As such, I predict the followi ng: Hypothesis 2 a : Increases in the number of top stars will lead to diminishing marginal returns in box office revenue. 26 Furthermore, I argue that the increase in the number of top stars in a film may eventually lead to negative return s in box office reve nue. Specifically, having too many stars in a film project creates additional difficulty for the studio to manage the process of filmmaking effectively . For instance, the increase in the number of top stars in a film project lead s to more coordinat ion and communication issue s between the stars and various specialist units such as wardrobe consultant, make - up artist, stunts specialists, on the project . This makes it more difficult for the studio to manage and execute a film project with an increase in top st ars. In addition, each of the top stars may have personal pre ferences in approaching the film such as how they would like to be filmed and portrayed and how they would desire a story - ending based on their own characteristics . Their preferences or personal characteristics, however, m ight not fit the ir roles in the film. Nevertheless , as Defi l lippi and Arthur (1998) observed, top stars coul d have a fragile ego that makes it more difficult for the studio to address various priorities in filmmaking. Therefore , having too many top stars creates more challenges to the studio to weld all the creative inputs that can lead to a final product that will perform more poorly at the box office . Furthermore, unnecessary (and often costly) issues can also arise , as more a nd more top stars assigned to a film, in the creative process of filmmaking in terms of social dynamics among the top stars . Scholars have highlighted the importance of social integration and communication among key individuals for group decision - making (S mith et al., 1994) and collaborative creativity and team innovation (Bissola & Imperatori, 2011 ; Paulus, Dzindolet, & Kohn, 2012 ) . However, interpersonal conflicts and negative emotion s could arise among top stars that stem from status competition due to t he pervasive and self - r einforcing nature of the status of the stars (Magee & Galinsky, 2008). This implies what Bendersky and Hays (2012) describe as the status 27 conflicts in group s that could negative ly affect task group performance. Taken together, I pred ict the following: Hypothesis 2b: Increases in the number of top stars will eventually lead to negative returns in box office revenue such that there is an inverted - U shaped relationship . Production Partners In the motion picture industry, studios often partner up with others to jointly work on a film project. Bollywood production firms such as Rohit Shetty Productions or Sarawati Entertainment have mainly worked with larger studios such as Yash Raj Productions. The collaborative arrangements enable studi os to pursue additional market opportunities that are typically outside of their individual realms. This is because such collaborative arrangements can help studios to gain quicker and less costly access to resources needed to execute a particular film pro ject. The interfirm partnership can serve as an effective means for mobilizing resources that are traditionally immobile (Lavie, 2006). For example, the various types of production resources that are used in making di fferent kinds of films are accumulated over time and difficult to acquire instantaneously. Specifically , collaboration with other studios can allow a studio to do more in a film by leveraging its partners expertise . In particular, the use of partners can a llow a studio to more efficiently manag e the production processes by avoiding costly delays and minimizing takes and re - shoots. Enlisting partner firms can also allow the studio to exploit complementarities when there are mutual benefits to join forces (Ro thaermel, 2001). For example, a studio may decide to draw upon its partner as a resource to draw on its specialized knowledge in making certain types of films . Although e nlisting a partner may enable the focal studi o to add more appealing attributes to the film that it could not offer on its o wn , adding more and more partners does not continue to 28 help to improve box office revenue of the film. This is because adding more production partners does not provid e the studio with proportional ly more opportunities to add more features to increase the e ntertainment value of its film . In addition, adding more and more production partners on a film may lead to problems with ensuring efficient communication and coordination among the multiple partners. This may reduce some of the added benefits a ssociated w ith enlisting more partner s in a film project . Ta ken together, adding more production partner s in a film can enable the s tudio to do more in a film by using partners expertise to enhance the entertainment value of the film. However, such added benefits a re likely to diminish with the addition of more partners. Thus, I predict the following o! H ypothesis 3 a : Increases in the number of production partners will lead to diminishing marginal returns in box office revenue. As argued earlier, the marginal benefits associated with adding more production partner s in a film is likely to diminish with t he increase in th e number of partners. I further contend that the increase in the number of production partners may eventually lead to negative returns in box office revenue. In other words, while enlisting one or two partners in a film may allow a studio to do more in a film by leveraging its partners expertise , having too many partners in a film might actually hurt the performance of the film . Specifically, adding more and more production partners incurs additional costs and entails more coordination and communication w ork among production partners. Effective coordination among partners is critical for the performance of inter - firm collaboration ventures (Gulati, Wohlgezogen, & Zhelyakov, 2012 ; Gulati & Singh, 1998 ). However, a s the film project gets incr easingly more co mplicated with the addition of other partners, coordination issues could 29 further arise and make it even more difficult for the focal firm to navigate the production process of filmmaking. In other words, th e additional cos ts of managing a l arge number of partners could outweigh t he potential benefits to be obtained from adding more partners. Thus, adding too many partner s can lead to weaker performance of the film at the box office. In addition , a s Gulati and colleagues (2012) noted, cultur al differences among multiple partners could further exacerbate coordination failures. Similarly, conflicts that stem from different existing structures and processes of each partner might present additional obstacles for the focal firm to collaborat e with partners. In fact, different partners may approach the underlying project very differently and have distinct procedures in collaboration following their firm - specific, path - dependent routines (Nelson & Winter, 1982; Eth i raj et al., 2005). The clashe s with procedures and routines might hurt the performance of the project even further . Moreover, Das and Teng (2002) further contend that free - riding and opportunistic behaviors are more likely to emerge when there are multiple partners involved in a colla borati ve venture , thus undermines the likelihood of venture success. Taken together, I argue that having too many partners in a film project not only reduce s the potential benefits to be obtained from collaboration, it might even hurt the performance of t he pro ject due to increased coordination costs, cultural differences, conflicted routines , and free - riding issues . As such , I predict the following: Hypothesis 3b: Increases in the number of partners will eventually lead to negative returns in box office revenu e such that there is an inverted U - shaped relationship . Broadly - Defined Experience on Performance of Motion Picture Projects While the characteristics of resources - such as the size of the production budget, the number of to p stars, and the number of production partners - are important for the performance of 30 production process is also a critical factor for the likelihood of success of that f ilm project. Poorly conceived and executed film projects may flop at the box office despite employing top stars or partnering with multiple studios. As compared to the resources that can be contracted to work on a film project, the specific experience of a studio in prior experience of making films cannot be easily imitated by others and p rovide it with a substantial competitive advantage. Furthermore, I differentiate between two kinds of experience of th e studio : broadly - defined experience with making all types of films and narrowly - defined experience with making movies within specific gen res . I then examine the different impacts of each of these experience s on performance of film projects . In the next se ction, I explore the role - defined experience on performance of its film projects. Broadly - defined Experience To begin with, studios rely on their ongoing activities with film projects of all kinds to develop and accumulate broadly - d efined experience of designing and executing these projects. First, studios carefully decide which projects to take on and develop further from a lineup of new ideas, pitches, and story lines. As scholars noted, the gree n - lighting process to approve or dec line a film project is not only idiosyncratic but also very difficult, involving much careful consideration and debate about the financial viability of the project (Eliashberg, Elberse, & Leenders, 2006; Elberse, 2002). When the project enters the product ion stage, a complex set of activities needs to be well planned, effectively coordinated, and smoothly executed. Efficient communication must be ensured at all time s among various specialists including cinematographers, stunts specialists , 31 camera and elect rical crews, costume and wardrobe consultants, among others. Last but not least, broadly - defined t of tweaking special effects, dubbi ng, and editing into a final product that meets artistic and technical standards. The studio relies on its prior experience to perform its different activities following its idiosyncratic and path - dependent production p rocesses. As the studio continually performs these activities in various film projects across market domains , it gradually accumulates broadly - defined experience and develop s a better understanding about how to more effectively deploy its various resources on a ny given film project. In contr ast with resources assigned to a film project that are only lined up for the duration of that project, the studio continues to draw upon s uch broadly - defined experience as it works through various film projects. Synthesizing motion picture economics with research on firm experience, I argue that a - defined experience with managing film projects in distinct markets can allow the studio to achieve a better understanding of film production and become more proficient in managing film projects i n different genres, thus leading to better performance of its projects. Specifically, research has shown that having exposure to a broader range of experience can . Similarly, Nerkar and Roberts (2 004) argue d that a firm can better pick new projects and introduce new products across distinct market categories by leveraging its greater experience in different domains . Furthermore, having greater experience with managing projects of different kinds creates more opportunities for various types of learning and produces a more substantial knowledge base that the firm can leverage upon in different situations ( H aunschild & Sullivan , 2002 ; Huber, 1991) . As such, a studio learns to become more effective in managing a complex set of activities in production, 32 more proficient in d ubbing , tweaking, and final editing to show a higher quality film as it gains more broadly - defined experience . In addition, since this experience is specific to a particular studio, i t tends to be causally ambiguous and difficult to imitate ; therefore, extensive broadly - defined experience, aside from the resources assigned to projects, is an important source of competitive advantage (Barney, 1991; Kor et al., 2007; Ethiraj et al., 2005 ). This will translate to better performances on various film projects that the studio takes on . Taken together, I predict the following: Hypothesis 4: An increase in the broadly - defined experience of a studio will lead to an increase in the box office rev enue of its films. Combining Broadly - d efined Experience with Production Budget As argued earlier, the characteristics of resources that are attached to a film project and broadly - defined experience with managing projects of different kinds can separately contribute to the performance of that project. I further contend th at the assigned production - defined experience interactively influence the performance of that film . Specifically , I argue that the broadly - def ined experience of a studio can enhance its ability to derive more value from the production budget assigned to its film s . First , having greater experience with managing projects of all kinds enables a studio to engage in a broader search for better solut ions to enhance the contribution of the assigned financial resources that can be a pplied to different film projects across all genres. Research has shown that firms are more likely to find creative ways to use available resources when they ha ve more experi ence (Sonenshein, 2014). This is because the greater prior experience of different kinds enable s a studio to explore and compare from a wider range of alternative s to recombin e and reconfigur e the available resources to obtain more value. 33 In addition , a s a studio gradually accumulates more broadly - defined experience with managing projects of different genres, it learns to make better use of financial resources and deploy them in improved ways to achieve the best possible outcomes in different contexts (Ege lman et al., 2016). Specifically, having more experience of different kinds creates more opportunities for various types of learning that can help improve performance in different situations (H aunschild & Sullivan , 2002 ). Drawing upon its greater broadly - d efined experience with managing film projects across different genres , a studio is more likely to be aware of better ways to enhance the value of the assigned production budget in a film. In other words, a higher level of broadly - defined experience will a llow a studio to more fully exploit the potential of working with a lavish production budget by leveraging its extensive broadly - defined experience across different genres to obtain value from the available resources. In contrast, t he lavish financial reso urce s assigned to the project may not produce corresponding results if the studio does not have substantial experience with managing different kinds of projects to make proper use of those resources. Taken together , a studio that has more broadly - defined e xperie nce become more aware of potential ways of using financial resources that it has learn ed from different genres , which give them more options to improve outcomes. Thus , I predict the following : H ypothesis 5 a: The positive effect of production budget o n box office revenue of a film will be enhanced when the firm has a higher level of broadly - defined experience . Combining Broadly - d efined Experience with Top Stars I argued earlier that featuring top stars in a film project can potentially improve the box office performance of that film and that greater broadly - defined experience of the studio can also lead to higher box office revenue. Furthermore, I argue that the benefits that a studio can obtain 34 from a top star in a film project can be enhanced by its broadly - defined experience with managing film projects of different genres. Featuring top stars in a film can bring additional appeal to a broader base of fans (Alb ert, 1998), providing moviegoers with decision heuristics (Levin et al., 1997), and making the film more recognizable and attractive (Desai & Basuroy, 2005) . T hus , they can improv e the entertainment value of films in which they appear (Litman & Kohl, 1989) . Nonetheless, it is the re smoothly executed, and all the creative inputs were welded well together into a final art piece (Litman, 1983) . Broadly - defined experience of a studio determines the extent to which it can apply to the proper use of its creative talents across all types of contexts thus influences its ability to obtain value from the featured stars in its film projects. Specifica lly, g reater broadly - defined experience allows a studio to explore a range of alternatives from different genres in finding more ways to derive value from the featured stars . I ncreasing evidence suggests that more experience of different kinds contributes to creativity by increasing the number of potential paths one can search and the number of potential ways of resources can be used (Amabile, 19 97; Shane, 2000) . Moreover , schol ars ha ve noted that a diverse base of experience can improve the productivity of available human resources (Koch & McGrath, 1996). Recent research also highlights that greater experience of different kinds can help the firm to discover new ways to assign their human resource to perform different tasks (Salvato, 2009) . Furthermore, a higher level of broadly - defined experience also gives a studio more opportunities to learn how to better manage various situations working with a great variety of top s tars and the way s through which the value of stars can be more fully exploited . Specifically, the 35 studio can make more appropriate adjustment s to make stars fit the ir roles and better accommodates the stars preferences in its films . In other words , greater broadly - defined experience creates a broader base of knowledge that can be leveraged upon in enhancing the value of top stars that are assigned to a film to improve the entertainment value of the film . Taken together, I argue that a higher level of broadly - defined experience can enhance the contribution of featuring top stars in a film. Thus, I predict the following: H ypothesis 5b: The positive effect of top stars on box office revenue of a film will be enhanced when t he firm has a higher level of broadly - defined experience . Combining Broadly - d efined Experience with Production Partners Earlier arguments suggested that collaborating with other studios on a film project can improve the likelihood of box office success of that film . T his is because enlisting production partners can allow the focal firm to draw on a broader base of knowledge from its partners (Vandaie & Zah eer, 2014; Lavie, 2006 ). It also enables the studio to add more appealing attributes to its film to improve its ent ertainment value . Furthermore, I argue that such benefits will be diminished when the studio has a higher level of broadly - defined experience . Specifically, a studio accumulates broadly - defined experience from managing film projects across different genre s and relies on such experience to guide its practices on any given project that it takes on throughout the processes from preproduction planning to postprodu ction editing. Such processes highlight the importance of the focal broadly - defined experie nce with turning available resources , on its own , into a finished product that meet s technical standard s and has public appeal (Lampel & Shamsie, 2003; Dannee ls, 2011; Mahoney & Pandian, 1992 ; Shamsie et al., 2009 ) . In other words, a higher - level of broadl y - defined experience enables the studio to leverage more of its own knowledge and expertise in different 36 genres in managing the underlying film project and th erefore reduce s the importance of its partners contribution . Under these conditions , collaborati ng with production partners can create coordination and communication issues with the different partners while performing complex tasks on a project (Gulati et al., 2012 ; Albers, Wohlgezogen, & Zajac, 2016 ) . Moreover, conflicts could arise from different r outines and procedures among partners during collaboration. Consequently, the focal firm can be forced t o make changes in order to accommodate the conflicting demands from a higher number of partners that may be detrimental to the appeal of the film to pot ential audiences. T hese issues can lead to problems with the development of the project resulting in the reduction of the box office potential of the finish film. T aken together , I therefore predict the following: H ypothesis 5c: The positive effect of prod uction partners on box office revenue of a film will be diminished when the firm has a higher level of broadly - defined experience . Narrowly - Defined Experience on Performance of Motion Picture Projects Narrowly - d efined Experience While th e broadly - defined experience of a studio indicates a collective assessment of this narrowly - naging film projects with in a particular genre . As noted earlier, experience is not only firm - specific but also context - specific (Argote & Miron - Spektor, 2011; Haas & Hansen, 2005) and different kinds of firm experience may produce different learning outco mes (Fong - Boh et al., 2007; In gram & Baum, 1997). Specifically, narrowly - to pursue opportunities within a particular market . This leads to specialized learning that may 37 improve s effectivene ss in performing tasks and act ivities related to that product market ( N arayanan et al., 2009; Eggers, 2012). In the motion picture industry, studios are constantly faced with decisions about which products and markets to pursue (Neale, 2000). As Shamsie an d colleagues (2009) point out, a studio chooses to build upon a specific set of experience to address certain product markets through the selection of specific genres. In fact, several theorists and practitioners have elaborated upon the economic and socio - cultural functions that a genre performs (Thompson & Boardwell, 1994; Altman, 1984; Neale, 2000; Buscombe, 1970; Jauss, 1982). A film genre stands for a cinematic story form that contains various generic elements based on certain formul a. Such formula inv olves the patterns of actions, sequences of events, and relationships among characters to be portrayed in ways that link films within the same genre together and distinguish them from other genres (Grant, 1986 ; Sobcha c k, 19 75 ; 1988 ). Thu s, different genres denote distinct product categories which have different demands for plotting, artistic portrayal, and aesthetic appeal (Neale, 2000; Buscombe, 1970). As such, the kinds of skills and talents needed to produce a film project are primaril y determined by the film genre (Shamsie et al . , 2009). In addition, the significance of genres ties closely with industrial and commercial nature of the motion picture industry. In this context, genres facilitate scale economies of production of films, as artistic products a re similar within category while being unique across ranges. To this difference inherent in products but also enable it to manufactur e its products in a cost - effe ctive (p. 218) . Thus, the n arrowly - defined experience that is governed by genres guide s studios to improve upon their practices over time from repeated activities that are tied to each of the film 38 genres. It reflects deliberate effort and commitment to exploit opportunities within a specific genre. Such effort and commitment translate into a better understanding of how to manage film projects within that genre (Rossman & Schilke, 2014; Shamsie et al. , 2009) . As a studio gains more experience within a narrow range, it develops specialized knowledge and expertise and learn s to carry out a selected range of tasks and activities in a more effective manner ( C lark & Huck man, 201 2; N arayanan et al., 2009 ). However, narrowly - defined experience may lose some of its benefits with the increase d accumulation of such experience limited to a certain genre. This is because repeating film projects of the same genre underscore s production p rocesses that aim to meet similar demands for plotting, setting, aesthetic appeal and so forth (Gomery & Pafort - Overduin , 2011; Jauss, 1982). It limits the opportunity that a studio can learn from performing different tasks related to different genres. In other words, t he narrow range of experience produces a limited knowledge domain that the studio can draw upon in executing diffe rent tasks (Taylor & Greve, 2006). As such, a studio becomes less likely to find alternative solutions in developing and executi ng new projects as it keeps stacking up experience within a specific genre . Taken together, having greater narrowly - defined expe rience can only help the studio to a certain extent in terms of the box office revenue of its films. Thus , I predict the followi ng: H ypothesis 6: An increase in the narrowly - defined experience of a studio will lead to diminishing marginal returns in the box office revenue of its films . Combining Narrowly - d efined Experience with Production Budget I theorized earlier that the produc narrowly - defined experience with managing the project of a certain genre could both contribute to the performance of this project. I further contend - defined experi ence 39 and p roduction budget interactively influence the project performance . To be more precise , I argue that a studio can derive more value from the production budget assigned to its film of a certain genre when it has a higher level of narrowly - defined ex perience w ith managing film projects within that genre. Specifically, a studio chooses to build upon a certain set of experience over time which leads to more opportunities to develop specialized learning about how to execute a selected range of tasks and activitie s related to films within a certain genre (Shamsie et al., 2009; Narayanan et al., 2009). As the studio continues to stack up the narrowly - defined experience with film projects within the particular genre, it gradually develops and refines a spec ific set o f routines related to projects of that kind (Ethiraj et al., 2005) . This will allow the studio to deploy financial resources most effectively to achieve the best possible outcomes with film projects within that genre. In other words, a higher lev el of narr owly - defined experience of the studio can enable it to obtain more value from the assigned financial resources in a film project within that genre. Indeed, assigning a greater amount of financial resources to a film may give the studio more opportunities to do more with the film by adding more attributes . Nevertheless, a higher level of narrowly - defined experience of the studio can allow it to better exploit those opportunities by invest ing the financial resources in the most appealing plots and attributes within the genre of that film to enhance entertainment value . Thus, a studio can further enhance the potential contribution of the lavish production budget assigned to its film s when it has a higher level of narrowly - defined experience with fil m projects within that genre. As such, I predict the following : H ypothesis 7a: The positive effect of production budget on box office revenue of a film will be enhanced when the firm has a higher level of narrowly - defined experience . 40 Combining Narrowly - d efined Experience with Top Stars Furthermore, I propose that t he extent to which a studio can benefit from assigning top stars in a film project is also influenced by its narrowly - defined experience with managing film projects within the genre of that film. Specifically , the potential contribution of featuring top st ars to the box office performance of films will be reduced when the studio has a higher level of narrowly - defined experience with film projects within that particular genre . As noted earlier, films of differ ent genres require distinct skills and knowledge to meet different demands concerning normative storytelling and artistic setting associated with those genres (Neale, 2000; Buscombe, 1970; Gomery & Pafort - Overduin, 2011). Studios that have gained rich experience in the given genre apply their knowledge and understanding about how to design and manage a film project in that domain. They rely on their accumulated experience with managing film project s within this genr e to guide the processes through which acting talents are portrayed and managed. In other words, a studio is more likely to draw upon its own knowledge and expertise to make a good film within that genre rather than relying on the use of top stars when it has a higher level of narrowly - defined experience of projects in that genre. Th is could re duce the likelihood that the studio can derive value from the use of top stars. Moreover , studios have to make adjustments to make stars fit their roles and to which may conflict with the ways that studios manage the pr ocesses based on their prior experience. However, a studio that possesses greater narrowly - defined experience is less able to adapt to work with different stars and less likely to find better solutions to use those stars. This is because deep experience wi thin a narrow range can make the studio over - reliant on its routines such that repeating similar projects could reinforce old behavior, drawing from familiar patterns and relying on heuristics when approaching a new 41 project ( Benner & Tushman, 2003; Audia & Goncalo, 2007 ). In other words, a studio that possesses greater narrowly - defined experience is less willing to experiment with new ideas or to search for alternative ways of using the resources (Schilling et al., 2003; Abernathy & Wayne, 1974). Thus, it b ecomes less able to detect improved ways to portray top stars to more full y exploit the potential benefits they could offer . As such, greater narrowly - defined experience could attenuate the potential contribution that can be derived from the use of top sta rs. Taken together, I predict the following: H ypothesis 7b: The positive effect of top stars on box office revenue of a film will be diminished when the firm has a higher level of narrowly - defined experience . Combining Narrowly - defined Experience with Pro duction Partners Similarly, I further propose that the extent t o which a studio can gain from enlisting - defined experience such that greater narrowly - defined experience reduces the potential benefits a studio can gain from collaborating with partners. As a studio gradually accumulates experience with managing film projects within a particular genre, it learns how to better design and execute a film project within that genre. The more of these experience that the studio has pertain ed to the genre , the in a film project within that genre . As such, the potential benefits that adding production part ners could offer is diminished when the studio has a h igher level of narrowly - defined experience. Moreover, through the accumulation of experience within a specific genre a studio becomes more entrenched in its own routines from repeating similar project s. This may hinder its ability to adapt to the routine s of its partners on a project , even if those routines might be superior. Thus, a studio which possesses greater narrowly - defined experience might not be aware 42 of potential ways in which it could combin This could also reduce the benefits that the studio might gain from collaborati ng with partners in a film project within that genre. Taken together , I predict the following: Hypothesis 7c: The positive effect of produc tion partner on box office revenue of a film will be diminished when the firm has a higher level of narrowly - defined experience . I theorized earlier that the potential benefits that a studio can obtain from collaborating with production partners will be d iminished when the studio has a higher level of broadly - defined experience across different genres . Similarly, I argued that having greater narrowly - defined experience with managing film projects within a certain genre also reduces the potential contributi on of collaborating with partner s. Furthermore, I argue that the negative effect is stronger when the firm has a higher level of broadly - defined than when it has a higher level of narrowly - defined experience. Compar e to the studios that possess a higher level of broadly - experience with managing film projects across different genres, a firm with more narrowly - defined experience has fewer kinds of expertise of its own. Even though collaborating with partners may create potential coordination and comm unicat ion issues (Gulati et al., 2012 ; Albers et al., 2016 ) , a studio with more narrowly - defined experience may have to leverage more of its partners resources in the film project. This is because such a studio has a limited knowledge domain of its own from its prior experience tied to a certain genre. expertise and obtain some benefits from collaborating with partners in the underlying project . In contrast , studios with greater broadly - defined experienc e are more likely to leverage their various types of expertise across different film genres and less likely to rely on their partners contribution. Indeed, each film project is unique. However, t he various types of 43 learning associated with the enriched ex perience with managing different kinds of projects create a broader base of knowledge which confers an ability to apply its expertise from other domains to the underlying project. To this point, Garzon - Vico and colleagues (2016) noted that more experience of different kinds leads to a greater range of knowledge that, in part, can be applied to new and different areas. Similarly, Zander and Kogut (1995) argued that enriched experience can facilitate knowledge transformation across contexts. As such, the stud greater broadly - defined experience with managing different kinds of projects further reduces its need to leve rage its partners as resources because it can do more on its own in the project . Therefore, the potential benefit to gain from having producti on partners is further reduced for the studio with a higher level of broadly - defined experience than those have mo re narrowly - defined experience. Thus, I predict the following: Hypothesis 7d: the positive effect of number of partners on box office revenue will be much lower when the firm has a higher level of broadly - defined experience than when the firm has a higher level of narrowly - defined experience . 44 METHODS This dissertation explores how specifi c project - based resources and distinct firm - based experience, separately and jointly, contribute to the performance of film projects within the motion picture industry. To test my theory and hypotheses, I draw on a sample from the Bollywood film industry , focusing on films produced and distributed by majo r Bollywood film studios between 1997 and 2016. Sample Over 1500 films in more than 20 languages are released in India each year, making the Indian film industry the largest in the world in terms of production volume (Deloitte, 2016) . This industry is d ominated by Bollywood, which produces Hindi films, account ing for almost half of aggregated domestic box office revenue in India (more than the next two, Tamil and Telugu combined 5 identity 2 , p. 514). Although the history of Bollywood can be traced back to the first half of the 20 th century, the industry was dominated by the masala (mixed - genre) formula for almost 50 years. Since mid - 19 80s, the major Bollywood studios have started to adopt the genre - based movie format and have become more comparable and competitive to their international counterparts. Furthermore , the Bollywood film industry is relatively less concentrated when compar ed to Hollywood. Whereas Hollywood is dominated by a few studios including Twentieth Century - Fox, Warner Bros . , Paramount, Universal, Sony /Columbia , and Disney (Miller & Shamsie, 1996; Vandaie & Zaheer, 2014; Mannor et al. , 2016), the Bollywood film industr y does not show such oligopolistic market structure. T h ere are numerous studios competing in this market 5 Alt hough Bollywood films only account for 15% of total outputs released in a year, they represent 4 0 % - 45% of total domestic box office revenue in India. 45 contributing to an averaged 10.5% cumulative annual growth in recent years ( Deloitte, 2016 ). To gain a better understanding of how project resources an d experience of studios contribute to their film performance, I focus on major Bollywood studios that have released at least 10 films since their inception. Although the full list contains more than 3300 films released over a twenty - year horizon, the large r share of films were produced by smaller studios and arthouses that were either short - lived or that only released one or two films. Therefore, the final sample of this dissertation consists of approximately 800 films released by 47 major Bollywood studios over the 20 - year period from 1997 to 2016. The list of films was first drawn from www.boxofficeindia.com and was then IMDB, and Wikipedia. This sample represen ts a near - comprehensive list of released featured films produced and distributed by major Bollywood studios within this market. Variables Variables for this study fall into two distinct categories : those related to each film and those related to the fir m that produces each of these films. In order to determine the firm that we assign to each film , I gathered data on the production companies from IMDB. When there are more than one production company listed on the site, I checked on Wikipedia and the studi websites as secondary source s . If this did not provide any more clarification, I assigned the film to the production company listed on IMDB with the most experienc e as the main production company. Dependent variable Domestic box office revenue . The g oal of this dissertation is to explore the contribution of resources that are assigned to a film project as well as the experience of the studio with 46 managing such a project , separately and jointly, to the box office performance of th is film project . Domes tic box office gross has been one of the most important and reliable performance indicators of films in the motion picture industry (Litman, 1983; Wallace et al., 1993; Sochay, 1994; Litman & Ahn, 1998; De Vany & Walls, 1999; for a review, see Eliashb erg e t al., 2006). Moreover, Crook and colleagues (2008) highlighted that RBV researchers can draw more meaningful implications by adopting market - based measure of performance than measures of value appropriation. Thus, I use domestic box office revenue as the measure of film performance. Domestic box office revenues of films were first collected from www.boxofficeindia.com (BOI hereafter) and then cross - checked with other sources, such as IMDB and Wikipedia, to ensu re the ir validity. Because d omestic b ox o ffice r evenues of films is highly skewed, I applied a log transformation of this variable. This is consistent with prior literature that log transforms the domestic box office revenue as the dependent variable to ad dress influence of extreme values in motion picture research (Basuroy et al., 2006; De Vany & Walls, 1999). Independent variables Production budget . The size of production budget indicates the level of financial resources assigned to each film project. I nformation about production budget was collected from BOI and was cross - checked with IMDB and Wikipedia. I then applied a log transformation of this measure to min imize potential influence of extreme values due to the high skewness of the distribution of p roduction budget . A quadratic term of the production budget was then created to examine the hypothesized curvilinear effect of production budget on box office reve nue of film projects. 47 Top stars. T op stars indicate the level of key acting talents assigne d to each film project. Featuring star actors and actresses in a film project can improve the entertainment value of films and appeal to a broader base of moviegoers (Albert, 1998; Levin et al., 1997). I have comp ile d a list of top stars primarily based on information from BOI. BOI maintains a list of top actors and actress in Bollywood of all time based on the total number of films in which they took on lead acting roles throughout their career that were box office hits . Household names such as Akshay Kuma r, Ajay Devgn, Aamir Khan, Govinda, Rani Mukherjee , and Kareena Kapoor were frequently mentioned on this list. However, I needed to determine the specific years that eac h of these all - time stars should either appear or not appear on the annual list of top stars. I tied their appearance s on the list based on the box office performance of their recent films in accordance with prior research on top stars (Albert, 1998; Wallace et al., 1993). In particular , a star remains on the list in the next three 6 years ( and was coded as a top star in films which they took on lead acting roles ) after the year of his or her most recent box office success based on information from BOI. Specifically, BOI describes the performance of a film on the basis of nine categories: A ll Time Blockbuster, Blockbuster, Super - Hit, Hit, Semi - Hit, Average, Below Average, Flop, Disaster . Although BOI does not disclose its propriet ar y formula of these categorizations, a film that is labeled as a Semi - Hit at least is typically one of the top g rossing film s within th e year in which the film was released. As such, I in cluded a star on the list in the next three years after the year in which he or she has assumed lead ing roles in a film that was categorized as either one of the followings: All Tim e Blockbuster, Blockbuster, Super - Hit, Hit , Semi - Hit . 6 To test the robustness of my finding, I also coded stars on the basis of a two - year window following th eir most recent successful films. This alternative measure is discussed in the results section (supplemental analyses). 48 A s a result, a pproximately 10 actresses and 15 actors were identified as top stars in a ny given year between 1997 and 2016. Drawing from this revised list, I first aggregate d the total number of top s tars that take on lead ing roles in each film. I then create d a quadratic term of the number of top stars assigned to each film so that the curvilinear effect of top stars on box office performance of film projects in which they appear can be examined. P ro duction partners. T he number of production partners captures the extent to which the focal studio can draw from a broader base of resources from its partners in jointly developing and executing the film project by mobilizing resources across firm boundarie s. In order to identify the number of production partners in a film project, I need ed to separate the studio that is the principal production company from its collaborating partners. Specifically, I first collected information from IMDB regarding the prod uction companies in each film. A film was then assigned to the studio that was listed as the principal production company in the film. When there is more than one production company listed on IMDB in a film, I then gathered data on the production companies from Wikipedia. In the additional clarification when their websites can be found 7 . In cases when following the above steps still did not provide me with a clear in dication, the film was assigned to the production company listed on IMDB with the most experience as the principal studio. The number of production partners in a film is then the total number of collaborating companies listed for each film. Furthermore , I create d a quadratic term of this variable to examine its curvilinear influence on the box office performance of films. 7 Some were not readily available during the data collection stage of this dissertation. 49 Using f ilm g enre s to study studio e xperience As noted earlier, scholars have elaborated upon the functions that a genre performs in mot ion picture industry ( Thompson & Boardwell, 1994; Buscombe, 1970; Altman, 1984; Jauss, 1982). Originally a French word meaning a film genre depicts a cinematic story form that has certain thematic components, characters, plots , settings, and techniques (Neale, 2000) . It usually contains certain elements from generic formula (Schatz, 1981). Because of the different formula and components tied to the different genres, Neale (2000) fur ther noted that dual characteristics, their own conflicts, and their 215). Therefore , different genres represent distinct product markets that have different demands for plotting, setting, characters, and artistic styling (Buscombe, 1970; Gomery & Pafort - Overduin , 2011). As such , film genres determine the kinds of skills and talents neede d to produce and execute a film project . Through the selection of film genres, s tudios choose to build upon a specific set of exp erience to address different product market opportunities (Shamsie et al., 2009). In spite of some minor differences, scholars have identified a list of genres that can serve as the primary basis for categorization of films in the motion picture industry ( Thompson & Bordwell, 1994; Schatz, 1981; Grant, 1977; Gomery & Pafort - Overduin, 2011; Finler, 2003). It should be noted, however, that the previous categorization of film genres has mostly focused on featured film s in Hollywood such that some film genre s (e.g. W estern 8 ) do not apply to th e context of th is stud y. In order to use film genres to study the experience of studios in Bollywood , I draw from previous studies (e.g. Miller & Shamsie, 2001; Shamsie et al., 2009) and employ a 8 A western film typ ically involves a frontier existence that encounters opportunity, hardship, violence, (Buscombe, 1970) . 50 modified categorization o f 12 genres in this dissertation including : action, adult, animation, comedy, drama, horror, love story, masala, mystery, romantic comedy, science fiction / fantasy, and thriller . I gathered data on the genre of each film from BOI. It should be pointed out t hat BOI only identifies the primary genre of each film while movie sites such as IMDB may associate a film with more than one genre. I chose BOI , instead of IMDB, as the primary source for film ge nre for two reasons. First, while IMDB may associate a film with two or three genres, it does not identify the primary genre of the film. Instead, the film genres are listed alphabetically for a film. Second, scholars have highlighted that it is the primar y genre of a film that determine s the normative structure of a film and the kinds of expertise needed to execute the film project ( Perretti & Negro, 2007; Gomery & Pafort - Overduin, 2011 ). 9 Broadly - defined experience. Broadly - defined experience serves as a proxy of the to manage a complex set of tasks in designing and executing film projects following its firm - specific processes that are accumulated from prior projects. Furthermore, existing research has shown that recen cy of prior experience should be taken into consideration in examining its impact on firm behaviors and outcomes (Argote, Beckman, & Epple, 1990, Eggers, 2012) . Specifically, it was argued that only recent experience is relevant to a , since acquired knowledge may become irrelevant over a long stretch of 9 As a side of caution, I also randomly selected a subsample of films and cross checked the genre of each film with IMDB . I have also compared the synopsis of each film, the primary genre listed on BOI with the definition of film genre provided by Neale (2000). The information was mostly consistent. 51 time (Argote et al., 1990). Therefore, - defined experience is measured by the . 10 Consistent with prior research (Egelman et al., 2016; Staats & Gino, 2012), I adop ted a Herfindahl - Hirschman Index (HHI) type of measure for the broadly - defined experience of a studio. in different film genres within the past five years in which the underlying film was released . Next, I obtained t heir squared values and aggregated the components. Because a larger HHI value is inversely related to the broadly - defined experience, I then subtract the value from 1 . 11 As mentioned before, the experience was measure d for the firm that was assigned as the major production company for the film . Narrowly - defined experience. Narrowly - defined experience reflects the commitment to pursue distinct opportunities in different product markets as categorized by film ge nres. The primary genre of a film defines the basis for its normative storytelling and artistic setting ( Neale, 2000; Perretti & Negro, 2007; Gomery & Pafort - Overduin, 2011). A studio gradually accumulates experience with managing film projects in certain genres reflecting its deliberate choices of selective markets and persistent efforts in pursuing those opportunities in those markets (Shamsie et al., 2009). Therefore, I firs t collected information about the primary genre of each film project that a studi o has carried out previously. I then obtain ed a count of the 10 T o test the robustness of my finding, I also created an alternative of this variable that measures a - defined experience by the number of different genres of films that it has produced in the previous 10 years. 11 I also measured broadly - de fined experience of a studio by obtaining the total number of differen t genres of films that it has produced in the previous five years a s a nother robustness check . 52 number of films that the studio has produced in the previous five 12 years in the given genre that is tied to each f ilm. Moreover, a quadratic term of this variable was created to examine its curvi linear impact on the box performance of films. Control variables Screens. Theatrical distribution and exhibition of a film has major influence on its box office performance ( Albert, 1998; Eliashberg et al., 2006; Shamsie et al., 2009; De Vany & Walls, 199 9). Thus, I control for the number of screens on which each film was released. Sequel. Previous research shows that the use of sequel is likely to contribute to box office revenue of films (Basuroy , Desai, & Taladar, 2006; De Vany & Walls, 1999; Eliashber g et al., 2006; Ravid, 1999 ) . Therefore, I control for sequel in this dissertation. This variable is coded 1 for sequels and 0 otherwise. Year. Motion picture research has shown that the timing of release of a film can influence the box office performanc e of th at film (Sochay, 1994; Litman & Kohl, 1989; Goldberg, 1991) . In this dissertation , I added a dummy for the year in which the film was released to control for the influence of release time and potential influence from changing market conditions. Estimation To analyze the data that I have sampled, I used both g eneralized e stimating e quation (GEE) modeling and ordinary least squares (OLS) analysis . I chose GEE modeling as my primary estimation method given that my data structure can be described as an unbalanced panel in which multiple observations reside within studios. A lthough the unit of analysis in this 12 - defined experience by aggregating the total number of films that it has produced in the previous 10 years in the given genre that is tied to each film. 53 dissertation is the film project, I theorized that the experience of a studio that manages the project also plays a crucial role in extracting value from resources assigned to this project. The experience of a studio is firm - specif ic and accumulates over time. It is important to correct for within - group correlation at the firm level and GEE modeling can effectively address this issue (Hardin & Hi lbe, 2013). As such, GEE modeling appears to be a more appropriate technique . Furthermor e, I standardized each of the variables before creating their multiplicative terms of these variables to test for interaction effects (Cohen, Cohen, Aiken, & West, 2003 ). Moreover, multiple robustness checks were performed to ensure the validity of my findings. Most centrally, I will report the results of the GEE analyses as the primary analysis in the next section and the results of the OLS analyses in the supplemental analyses section. 54 T able 1 D ata M easures Construct Measures Calculation Level Dependent Variable Domestic box office revenue Gross revenue from domestic box office ln(In_Gross) Project Independent Variables Quadratic term of the size of production budget The size of production budget assigned to each film squared [ln(Budget)^2] Project Quadratic term of the number of Top stars The total number of top stars featured in each film squared (The number of Stars^ 2) Project Quadratic term of the number of production part ners The total number of production companies in each film squared (The number of production companies ^2) Project Broadly - defined experience experience with different genres of films it has produced in previous 5 years Studio Quadratic term of the narrowly - defined experience The number of films that a studio has produced in the given genre in previous 5 years squared ( The number of films produced in the given genre ^2) Studio Control Variables Screens The number of screens on which each film was released Project Sequel Dummy code films which are sequels Project Year Year of release Year 55 RESULTS I n this chapter, I present the results of this dissertation in three se ctions . First, I present the descriptive statistics and inter - correlations of the variables in this study. These can be found in Table 2 , and I will briefly discuss some of the interesting findings from this table. Next, I present the results of my primary analyses, in which I utilized GEE modeling to test each of the hypothes ized relationships . Specifically, I selected independence correla tion structure in conducting GEE analyses because QIC tests showed this one best specified the working correlation stru cture in my sample (Cui, 2007; Pan, 2001). 13 These results are presented in Table 3. Following that, I will also present the results of my supplemental analyses in Table 4 and Table 5. Table 4 features the results from my first set of supplemental analyses , in which I used alternative measure s of the moderato rs to test my hypotheses . In addition , I also utilized an other analytical method - OLS regression to analyze my data. I present these results in Table 5 . Descriptive Statistics As mentioned, t he descrip tive statistics and correlations matrix of all the variables in my study can be found in Table 2. As expected, domestic box office revenue has an extremely large variance in my sample with the maximum value of 4 , 95 3 , 0 00 , 000 rupees (approximately $7 1 millio n USD 14 ) for Dangal to a minimum value of 14000 rupees (approximately $200 USD) for Dhara . Similarly, production budget also shows a very large variance with the maximum value of 1 , 800 , 000 , 000 rupees (approximately $2 6 million USD) for Prem Ratan Dhan Payo to a minimum value of 1 , 000 , 000 rupees for Munnibai B.A.B (approximately 14000 USD) . Though I 13 Although I report the results of GEE analyses using an independence correlation structure, the results generally hold for analyses using an exchangeable correlation structure. 14 The cur rency conversions are b ased on the exchange rate between Indian Rupees and US Dollars as of 4/29/2019 56 reported raw values in the descriptive statistics , log transformation of both domestic box office revenue and production budget were applied because t he distribu tion s of these two variables we re very highly skewed . This is c onsistent with prior research in motion picture economics that log transform box office receipts and production budget to minimize the influence of extreme value s due to their skewness ( Basuroy et al., 2006; De Vany & Walls, 1999) . In my sample, t he number of production partners in a film project range s between zero and eight . Approximately 4 8 % of films were produced by a single studio (no partner) while the remaining 5 2 % of films were jointly p roduced by at least two production companies . Two films were produced by eight production companies including Speedy Singhs and Chauranga . In addition, 54% 15 of films featured at least one top star. Among them, 27% of films used one top star and 27% featur ed more than one top star. In addition, three films had five top stars, the highest number of stars assigned to a film in my sample , including Eklavya The Royal Guard, Kabhi Alvida Na Kehna , and Kabhi Khushi Kabhie Gham . Furthermore, a total of 11 genre s were identified in my sample. Drama represents the highest percentage (29 %) while s ci ence - f iction (Sci - Fi) has the least amount (0.6%) 16 . Th is low appearance of Sc i - Fi films in my sample seemed a bit surprising at first given the increasing popularity of Sci - Fi films in Bollywood in recent years (Deloittee, 2016) . Perhaps this is because such information is drawing on the basis of the primary genre of each film. Many times, films are associated with more than one genre s . However, scholars argue that it is the primary genre of a film defines its normative structure and artistic setting (Perretti & Negro, 2007; Neale, 2000 ). 15 This is using a list of top stars on the basis of a three - year window following th eir most recent successful films as described in the previous chapter . Th e alternative measure of top stars which used a two - year window shows that 51% of films featured at least one top star. 16 Due to missing information on certain key variables, two an imation films were dropped. Thus, Sci - Fi films represent the smallest category in this study. 57 For instance , films such as PK and Rudraksh are not primarily Sci - Fi films despite having some Sci - Fi twists . T he broadly - defined experi ence of a studio ranges between zero and one, that Eros International appear s to have the greatest breadth of experience across film genres . Meanwhile, UTV Motion Pictures has the deepest experience in managing drama films that it was able to leverage prio r experience with 18 drama films (or 28 drama films) from the previous five years (or the previous 10 years) in managing the film of Haider . Although the descriptive statistics are generally in line with expectation, there are some strong correlations amo n g the variables in my sample. For example, both production budget and screens are strongly correlated with domestic box office revenue at 0.81 and 0.77 respectively. These correlations are expected to be high as big - budget films are likely to attract more audiences to theaters. The strong correlation between screens and production budget can also be explained by the same reason. The other correlations are mostly as expected. For instance, the production budget has a p ositive relationship with the number of top stars and the number of production partners (r=0.4 7 , r=0. 30 ). In addition, both the number of top stars and the number of production partners are related to domestic box office revenue at 0.4 6 and 0. 20 , respectively. Perhaps an interesting note, the relationship between the size of production budget and narrowly - def ined experience (r=0. 1 1 ) is weaker than the relationship between the size of production budget and broadly - defined experience (r=0. 2 4). In addition , the number of production partner has a moderately positive relationship with narrowly - defined experience ( r=0. 21 ) and with broadly - defined experience (r=0.2 0 ). Moreover , both broadly - defined experience and narrowly - defined experience have negative relationships with the number of top stars (r= - .0 6 , and r= - .0 6 ), suggesting that firms feel less compelled to work with top stars in their films as they become more experienced. Moreover, although only 44 sequels were identified in 58 my sample, the use of sequels appears to have moderate relationships wi th screens, pro duction budget, and the number of top stars (r=0.2 3 , r=0.1 3 , r=0.1 2 , respectively). Primary Analyses Table 3 reports the results of the primary analyses of this study using GEE modeling . Model 1 only includes baseline control variables. Of note, the numbe r of screens on which each film was released is expected to be positively related to domestic box office revenue . The coefficient is positive and statistically significant (p<0.001). Similarly, the use of sequel is expected to contribute to box office reve nue of films. However, this variable is not statistically significant ( b = 0.18; p=0. 1 4 ). Perhaps this can be explained by the low percentage of sequel films identified in my sample (0.6%). Next, in order to examine the curvilinear effect on box office performance of production budget, top stars, production partners, and their interactions with experience variables, the first order standardized terms of these variables were then included. As Model 2 indicates, both the size of production budge t an d the number of top stars are positively related to box office revenue. Their coefficients are statistically significant (p<0.0 01 ; p<0.0 5 ). Contrary to expectation, the number of production partners is negatively related to box office revenue. However, its coefficient is not statistically significant ( b = - 0.0 5 ; p=0. 43 ). Finally, both broadly - defined experience and narrowly - defined experience positively relate to box office revenue although only the narrowly - defined experience was marginal ly significant (b=0 .07, p= 0.09) . 59 Table 2 Correlations and Descriptive Statistics 60 Hypotheses 1 a, 2a, and 3 a predict diminishing marginal return s on domestic box office revenue with the increas ing amount of project resources assigned to fil m s . As indicated by the negative coefficient of the quadratic term of production budget ( b = - 0.15, p< 0 .0 5 ), hypothesis 1 a is supported such that there is a tapering effect on box office revenue associated with the increase in the size of production budget. Similarly, hypothesis 2 a predicts that there is a diminishing marginal return on box office associated with the use of more top stars in a film. The coefficient of the quadratic term of the number of top stars is negative and statistically significant ( b = - 0.08, p< 0 .01). Thus, I find strong support for hypothesis 2 a as well. While t he coefficient of the quadratic term of the number of production partner is negative, it is not statistically significant (b= - 0.02, p=0. 59 .). This result fails to provide suppor t for hypothesis 3 a . Hypotheses 1b, 2b, and 3b predict that increases in the amount of these resources will eventually lead to negative returns on domestic box office revenue such that the relationships are inverted - U shaped . To examine t he curvilinear im pacts on the domestic box office revenue of films by adding m ore of production budget, top stars, and partners, I first plotted these results in Figure 1 through 3. As shown in the se figures, the estimates of production budget are relatively more precise t han those of top stars and production partners . In Figure 1, the marginal benefits associated with assigning greater production budget diminish as the (standardized) size of production budget increase s . In order to examine whether or not adding more produc tion budget may eventually lead to declined performance , I specified several spline functions. Following recommendation, I began with placing multiple equally - spaced knots over the range of produc tion budget. When placing observations in to 5 knots, the initial test showed a negatively significant coefficient of the 5 th quintile following the first four positive knots. While it seems to suggest that there might be a change in slope from positive to 61 negative in the 5 th quintile, further exp loration at a more granular level is needed to examine whether there is a turning point at which adding more production budget will produce declining return. Thus, I divided the observations into 10 groups and 20 gr oups and constructed linear spline functi ons respectively. Figure 1 Logged production budget and p redicted values of logged domestic box office revenue with 95% confidence intervals Drawing from prior studies that use spline function to test curvilinear effect on pe rformance above and below aspiration levels (e.g. Greve, 2003 ; Joseph & Gaba, 2015; Park, 2007 ), a change in the slope from positive to negative at a threshold would indica te an inverted - U relationship such that increasing production budget, by itself, wil l yield declining return beyond this threshold. Although the slope for the 86 - 90 percentile interval was negatively 62 significant 17 , the results show an overall non - significa nt slope above this level after splitting the correlations into two variables at the 86 th percentile. Therefore, an increase in production budget leads to declining marginal returns but this does not turn negative. Thus, hypothesis 1b is not supported. Next , I followed similar steps for the use of additional stars using mkspline in Stata 13 in constructing linear spline functions in measuring change in slopes before and after the inflection point. The results show that adding the fourth top star is unlik ely to make further contribution to box office revenue based on the statistics from spl ine function s . Although Figure 2 seems to suggest that having more than 3 stars in a film might actually have an adverse impact on the box office revenue, I was unable t o find a statistically significant slope for this interval . In other words, this result failed to conclusively demonstrate negative return for the use of additional stars . H ypothesis 2b is not supported. Finally, I also plotted Figure 3 illustrating the relationship between the number of production partners and box office revenue . However, n o spline tests were performed since both c orrelation coefficient s of first order and quadratic term were non - significant . Thus, I was not able to find support for hypothesis 3b. 17 I suspect this is likely influenced by two extreme observations which were very expensive (89 th percentile of production budget) box office failures including Broken Horses and Zanjeer. 63 Figure 2 Number of stars and predicted values of logged domestic box office revenue with 95% confidence intervals Figure 3 Number of production partners and predicted values of logged domestic box office revenue with 95% confidence intervals 64 Table 3 GEE Regression Coefficients and Robust Standard Errors Predicting Domest ic Box Office ( with a 5 year time window of experience variables) Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Domestic Box Office Revenue (logged) Screens 0.002 *** (0.0002) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) Sequels 0.181 (0.123) 0.355* (0.145) 0.288* (0.121) 0.356* (0.147) 0.364* (0.142) 0.372** (0.140) Production Budget ( logged) 1.565*** (0.108) 1.163** * (0.138) 1.564*** (0.108) 1.551*** (0.105) 1.558*** (0.100) Number of Stars 0.082* (0.036) 0.255*** (0.060) 0.083* (0.036) 0.075* (0.037) 0.070 (0.037) Number of Partners - 0.052 (0.066) - 0.011 (0.093) - 0.052 (0.066) - 0.050 (0.065) - 0.051 (0.067) Bro adly - defined Experience (BDE , 5 years) 0.059 (0.053) 0.041 (0.051) 0.060 (0.055) 0.073 (0.056) 0.061 (0.051) Narrowly - defined Experience (NDE , 5 years) 0.066 (0.039) 0.067 (0.041) 0.060 (0.079) 0.059 (0.040) 0.024 (0.044) Budget x Budget - 0.147* (0. 059) Stars x Stars - 0.083** (0.025) Partners x Partners - 0.015 (0.029) NDE x NDE (5 years) 0.002 (0.018) Budget x BDE (5 years) 0.036 (0.055) Stars x BDE (5 years) - 0.094* (0.042) Partners x BDE (5 years) 0.022 (0.0 54) Budget x NDE (5 years) 0.123 (0.071) Stars x NDE (5 years) - 0.119** (0.038) Partners x NDE (5 years) - 0.013 (0.026) Intercept 17.896*** (0.459) 19.117*** (0.508) 18.938*** (0.497) 19.114*** (0.522) 19.091*** (0.475) 19.157*** (0.4 10) n = 664 in model 2 th r ough 6, n=722 in model 1 *** p <0.001; ** p <0.01; * p <0.05; p < .10 Robust standard errors are in parentheses; year dummy variables included 65 Hypothes i s 4 predict s that an increase in the broadly - defined experience of a s tudio is positively related to the box office revenue of its films. Although the coefficient of this variable is positive , as shown in Model 2, it is not statistically significant (b= 0.0 6 ; p=0. 27 ) . Thus, hyp othesis 4 is not supported. Hypotheses 5a through 5c predict that the broadly - defined experience of a studio moderates the relationships , in different patterns, between each of the predictors including production budget, top stars, and production partners, with box office revenue of its films, respective ly. In order to test these hypotheses, I include these interactions in Model 5. Specifically, h ypothesis 5a arg ues that a higher level of broadly - defined experience of a studio can enhance the positive effect of the production budget on box office revenue of its film s . As indicated in Table 3, the coefficient of broadly - defined experience interaction with the size of production budget is positive but not statistically significant (b=0.0 4 , p=0. 5 1). Thus, the result fails to provide support for this predictio n. Similarly, hypothesis 5b predicts that a firm can generate more revenue with more top stars when it ha s a higher level of broadly - defined experience. Contrary to my prediction, t he interaction effect of broadly - defined experience with the number of top stars is negatively significant (b= - 0. 09 , p<0.0 5 ) . I plotted this interaction in Figure 4 . As F igure 4 indicates, having more broadly - defined experience with managing film projects in different genres does not enable the firm to obtain more value from mor e top stars 18 . Thus, hypothesis 5b is not su pported. Furthermore, hypothesis 5c suggests that the positive effect of production partners on box office revenue is diminished when the studio has a higher level of broadly - defined experience. 18 Though Figure 4 suggests that a higher level of broadly - defined experience could potentially buffer the adverse impact of not having any top star in a film . This will be discussed further in the following chapter. 66 However, the inter action effect of broadly - defined experience with the number of production partners is non - significant (b=0.0 2 , p=0. 69 ) . T hus , this result fails to provide support for hypothesis 5c. Figure 4 Interaction Effect of Broadly - defined Experience with Number of Stars Predicting Domestic Box Office Revenue (logged) Hypothesis 6 argues that there is a diminishing marginal return on box office revenue associated with an increase in the narrowly - defined experience of a stu dio. In order to test this hypothesis, I include the quadratic term of narrowly - defined experience in Model 4. The coefficient of its quadratic term is not statistically significant (b = 0.0 2 , p=0.9 1 ). Thus, I do not find support for this prediction . Moreover, h ypotheses 7a through 7c predict that the narrowly - defined experience of a studio moderates the impact on box office revenue of production budget, top stars, and production partners in different ways . To test these hypothes es , I include their int eractions in Model 6. Sp ecifically, hypothesis 7a predicts that the positive effect of the production budget on 67 box office revenue of films can be enhanced when the studio has a higher level of narrowly - defined experience. As shown in Table 3, the coefficient of narrowly - defined experience interaction with the production budget is positive and marginally significant (b=0.12, p=0.09). This result is plotted in Figure 5 illustrating its interaction effect . According to Figure 5 , the narrowly - defin ed experience of a studio amplifies the positive impact of production budget on the box office revenue of its films such that a studio that has more experience within a certain genre can extract more value from an increase in budget for films within that g enre. Though a relatively inexperien ced studio with a certain genre of films can make more use of a constrained production budget in making films within that genre. Overall, t h e result shows some support for this prediction . Figure 5 Interaction Effect of Narrowly - defined Experience with Production Budget Predicting Domestic Box Office Revenue (logged) Hypothesis 7b predicts that the positive impact of top stars on box office revenue is reduced when the studio has a higher level of narrowly - defined experience wi th films within that 68 genre. The result provide s support for this prediction that the coefficient of the interaction between the number of top stars featured in a film and its box office revenue is negative and statistically significant (b= - 0.12, p<0.01). Th is interaction is plotted in Figure 6 to further illustrate this result. As shown in this figure, having more narrowly - ability to extract more value from the use of multiple top stars. This f inding is consistent with my overriding argument that the narrowly - defined experience of a studio hinders its ability to continue to derive more value from adding more top stars in its film projects. In addition, this result appears to be quite robust with different time windows of experience and analytical methods 19 across severa l robustness checks. Figure 6 Interaction Effect of Narrowly - defined Experience with Number of Stars Predicting Domestic Box Office Revenue (logged) 19 The resul ts of these additi onal robustness checks are presented in the supplemental analyses section . 69 Meanwhile, other results do not provide support for hypothesis 7c and 7d . Specifically , I theorized that the positive impact of adding production partners on box office revenue of a film will be reduced when the studio has a higher level of narrowly - define d experience with films within that genre. The coefficient of the interaction term between these two variables is negative but not statistically significant. Thus, hypothesis 7c is not supported. In addition, hypothesis 7d predicts that the positive effec t of production partners on box office revenue is reduced even mo re when the studio has greater broadly - defined experience than when it has greater narrowly - defined experience. Because there is no significant moderating effect of broadly - defined experience or narrowly - defined experience with the number of partners , I was not able to find support for hypothesis 7d. To summarize, of the 1 5 hypothesized relationships, only four were supported based on the primary analyses using GEE regression . Consistent with prior research though, the effects of production budget an d top stars appear to be quite strong. Of particular interest in this study, there are limits to the benefits that can be derived from adding more production budget and assigning more top stars to a film. In addition , there were some interesting findings for the effect of experience on the value that could be extracted from resources. Specifically , prior experience of a studio can allow it to extract more value from its production budget when it ha s accumulated a higher level of experience with films only within that genre . Nonetheless , the positive effect of top stars on the box office revenue of a film was negatively impacted by broad as well as narrow experience. To ensure the validity of these f indings, I con ducted multiple robustness checks and present the results in the supplemental analysis section that follows. 70 Supplemental Analyses I conducted a large number of supplemental analyses testing my model s using alternative measures of the ind ependent variable s 20 and moderator s. In addition, I also utilized other analytical techniques. In what follows, I present the results of some of these supplemental analyses in two tables. First, Model 7 - 1 1 in Table 4 feature the results using alternative me asures of moderators . Next , I present the results using OLS regression with robust standard errors in Model 1 2 - 16 in Table 5 . The first set of supplemental analyses focused on the moderating variables. Specifically, I created an alternative measure of b roadly - defined experience of a studio with different film genres in the previous 10 years following the HHI type of approach . In particular, I obtained the squared values of the percentage s 10 ye ars and then subtract ed the aggregated components of the squared value s from 1 . 21 Similarly, the narrowly - defined experience of a studio is measured by the number of films it has produced in the same genre in the previous 10 years . Applying this wider time window could potentially allow me to examine whether firm experience becomes irrelevant quickly and how fast learning decays. These analyses worth exploring given the rapid development in information technology in recent decades an d 20 I applied an alternative list of top stars in measuring the number of stars assigned to a film as another robustness check. In particular, it defines the list of stars in any given year based on a two - year window following his/her most recent success (as opposed to three years). The results were highly consistent with the primary analyses presented in Table 3. To save space, th ese results are not presented here. 21 I also measured broadly - defined experience of a studio by counting the total number of different genre of films that it has produced in the past 5 years , and 10 years, respectively, as additional robustness checks . I do not discuss the results of these additional robustness checks since the results were highly consistent. 71 advancement in product ion processes in many different sectors including the entertainment industry. I present the results of this set of supplemental analyses in Table 4 beginning with the baseline model. As indicated in Model 7, the coefficients of b oth broadly - defined expe rience and narrowly - defined experience remain ed non - significant. Model 8 includes the quadratic terms of resource variables. Consistent with Model 3 in Table 3, the coefficients of quadratic terms of production budget and top stars remain ed negative and st atistically significant (b= - 0.15; p<0. 0 5 ; b= - 0.0 9 , p<0.0 5 respectively), providing support to hypothesis 1 a and 2 a . Hypothesis 3 a , predicting diminishing return on box office revenue by adding more partners, was not supported (b= - 0. 0 2 ; p=0. 61 ). Furthermore , no support was found for hypotheses 1b, 2b, and 3b after reconstructing linear spline functions using mkspline in Stata 13. Specifically, I was unable to conclusively demonstrate a change in slopes from positive to negative for these variables. No supp ort was found for hypothesis 4 that the effect of broadly - defined experience on box office revenue was positive but not significant. Model 10 presents the results of interaction effec ts regarding broadly - defined experience, addressing hypotheses 5 a throug h c. Again, no support was found in this model. Model 9 includes the quadratic term of narrowly - defined experience. Although the coefficient of the quadratic term of narrowly - defined experience became negative, comparing to Model 4 in Table 3, it remained non - significant (b= - 0.01; p=0. 34 ) thus failed to provide support to hypothesis 6. Model 11 concerns the interaction effects of narrowly - defined experience with budget, stars, and prod uction partners. Similar to Model 6 in Table 3, only hypothes e s 7a and 7b received some support. In particular, the coefficient of the interaction term of narrowly - defined experience with production budget was positive and marginally significant (b=0.11, p =0.06). Hypothesis 7b was supported as indicated by the 72 coefficient that was negative and statistically significant (b= - 0.11, p<0.001). In short, results from this set of supplemental analyses are consistent with the primary analyses . In addition , the main effect of broadly - defined experience remained non - significant while the main effect of narrowly - defined experience shrank but bec ame non - significant (b= 0.02 , p=0.62). As such, there were no meaningful differences between the five - year and 10 - year time window in terms of experience effects. Next, I present another set of supp lemental analyses in which I utilized hierarchical OLS regression with robust standard errors to account for the potential influence of heteroskedasticity (Wooldridge, 1989) . The results of OLS regression analyses are presented in Model 12 through 16 in Ta ble 6. To begin with , Model 12 includes the baseline predictor variables. The results are largely in line with primary analyses. Pro duction budget is positive and statistica lly related to domestic box office revenue (p<0.001) while the coefficient of top s tars becomes marginally significant (p=0.052). Model 13 includes quadratic terms of the resource variables . T he results were consistent with Model 4 in Table 3. The square d term of production budget is negatively significant related to box office revenue (b= - 0.15, p<0.0 0 1) supporting hypothesis 1 a . Similarly, the negatively significant square term of top stars (b= - 0.08, p<0.01) provides support for hypothesis 2 a . Hypothesis 3a , however, is still not supported (b= - 0.02, p=0.60). Consi stent with the primary analyses, I was not able to find support for hypotheses 1b, 2b, and 3b. Furthermore, Model 14 includes the quadratic term of narrowly - defined experience and its coefficient remains non - significant (b=0.00 2 , p=0.9 1 ). Model 15 includes interactions of broad ly - defined experience with budget s , top stars, and production partners, addressing hypotheses 5 a through c. Same as model 5 in Table 3, these hypotheses are not supported. 73 Finally, Model 16 shows the interaction effects of narrowly - defined experience rega rding hypotheses 7a through 7 c . Of note, only hypothes e s 7 a and 7 b are supported as indicated by the positive, and negative coefficient s, respectively ( b=0.12, p<0.05; b= - 0.12, p<0.01). Since no significant moderating effect was found for broadly - defined e xperience or narrowly - defined experience with production partners, hypothesis 7d was not supported. O ther results were almost identical comparing to Model 6 in Table 3. In short, results from OLS regression analyses are largely consistent with the primary analyses. Taken together, my findings from the primary analyses are generally robust across these sup plemental analyses. 74 Table 4 GEE Regression Coefficients and Robust Standard Errors Predicting Domestic Box Of fice (with a 10 - year time window of experience variables) Model 7 Model 8 Model 9 Model 10 Model 11 Domestic Box Office Revenue (logged) Screens 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) Sequels 0.362 ** (0.138) 0.297* (0.117) 0.352* (0.137) 0.378** (0.135) 0.380** (0.133) Production Budget ( logged) 1.552*** (0.107) 1.148*** (0.136) 1.555*** (0.108) 1.518*** (0.100) 1.549*** (0.096) Number of Stars 0.089* (0.037) 0.269*** (0.061) 0.088 * (0.037) 0. 091* (0.035) 0.076* (0.038) Number of Partners - 0.041 (0.065) - 0.060 (0.091) - 0.040 (0.065) - 0.039 (0.065) - 0.039 (0.064) Broadly - defined Experience (BDE , 10 years) 0.077 (0.053) 0.027 (0.053) 0.071 (0.053) 0.076 (0.054) 0.073 (0.050) Narrowly - de fined Experience (NDE , 10 years) 0.023 (0.047) 0.063 (0.051) 0.060 (0.076) 0.020 (0.047) - 0.011 (0.039) Budget x Budget - 0.149* (0.060) Stars x Stars - 0.088* (0.026) Partners x Partners - 0.015 (0.029) NDE x NDE (10 years) - 0.012 (0. 012) Budget x BDE (10 years) 0.008 (0.051) Stars x BDE (10 years) - 0.122* (0.042) Partners x BDE (10 years) 0.012 (0.060) Budget x NDE (10 years) 0.112 (0.059) Stars x NDE (10 years) - 0.108** (0.034) Partners x NDE (10 years) - 0.021 (0.025) Intercept 19.124*** (0.482) 18.957*** (0.464) 19.152*** (0.485) 19.073*** (0.440) 19.188*** (0.361) n = 6 78 *** p <0.001; ** p <0.01; * p <0.05; p < .10 Robust standard errors are in parentheses; year dummy variables included 75 Table 5 OLS Regression Coefficients and Robust Standard Errors Predicting Domestic Box Office ( with a 5 year time window of experience variables ) Model 12 Model 13 Model 14 Model 15 Model 16 Domestic Bo x Office Revenue (logged) Screens 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) 0.001*** (0.0001) Sequels 0.355* (0.138) 0.288* (0.135) 0.356* (0.139) 0.364** (0.135) 0.372** (0.133) Production Budget ( logged) 1.565*** (0. 074) 1.163*** (0.142) 1.564*** (0.075) 1.551*** (0.080) 1.558*** (0.073) Number of Stars 0.082 (0.042) 0.255*** (0.067) 0.083* (0.042) 0.075 (0.042) 0.070 (0.043) Number of Partners - 0.052 (0.048) - 0.012 (0.064) - 0.052 (0.048) - 0.050 (0.049) - 0 .051 (0.048) Broadly - defined Experience (BDE , 5 years) 0.059 (0.047) 0.041 (0.045) 0.060 (0.048) 0.073 (0.050) 0.061 (0.047) Narrowly - defined Experience (NDE , 5 years) 0.066 (0.040) 0.067 (0.040) 0.060 (0.070) 0.059 (0.040) 0.024 (0.047) Budget x Budget - 0.147*** (0.041) Stars x Stars - 0.083** (0.025) Partners x Partners - 0.015 (0.029) NDE x NDE (5 years) - 0.002 (0.020) Budget x BDE (5 years) 0.036 (0.056) Stars x BDE (5 years) - 0.094* (0.045) Partners x B DE (5 years) 0.022 (0.055) Budget x NDE (5 years) 0.123* (0.056) Stars x NDE (5 years) - 0.119** (0.042) Partners x NDE (5 years) - 0.013 (0.029) Intercept 19.117*** (0.564) 18.938*** (0.562) 19.114*** (0.566) 19.091*** (0.540) 19.157*** (0.461) n = 664 *** p <0.001; ** p <0.01; * p <0.05; p < .10 Robust standard errors are in parentheses; year dummy variables included 76 DISCUSS ION T he conventional RBV view emphasizes control of VRIN resources as a sufficient condit ion for firms to achieve SCA (Barney, 1991; Peteraf, 1993) . However, this traditional view of RBV has impeded our understanding about how firms can derive the most benefits from the se resources (Kraaijenbrink et al., 2010) . This is especially important in project - based industries whe re firms rely on access to commonly available resources in order to improve the performance of their projects. Since such resources are available to all the firms within the industry , each firm must search for ways to increase t he value that they can derive from them. To begin with, w hereas the RBV literature emphasizes the possession of VRIN resources as a sufficient c ondition for firms to achieve SCA, I argue that there are limits to the benefits that can be derived from simply having access to VRIN resources. In other words, this dissertation challenges the traditional RBV view which implies that firms could continue to improve performance by increasing the amount of VRIN resources that they draw upon. Specifically, I examined the extent to which firms can benefit from the addition of such resources and demonstrated that adding more production budget and assigning more top stars will generate a diminishing marginal return to box office revenue of these films. Furthermore , I soug ht to explore the different ways that a firm can obtain value from its resources using the RBV literature (Barney, 1991; Peteraf & Barney, 2003 ) and research on firm experience ( Cuypers et al. , 2017; Egelman et al., 2016 ) . In doing so , I developed and test ed arguments how a its performance and how prior experience influences the value that it may be able to obtain from its resources . Overall , this dissertation makes several important contribut ions to the st rategic management literature as well as to the motion picture research. 77 Moreover, this research unpacks the distinct mechanisms through which studios derive more value from the different types of resources by leveraging their prior experience. My finding s show that the benefits to be obtained from financial resources and human resources are both tied to prior experience of the firm but through distinct ways. Drawing from these findings , this study sheds lights on the processes through which the role of re source access and the role of prior experience in extracting value from resources contribute to firm performance . The contributions and implications of this study are presented as follows. The Role of Resource Ac c ess First , this study sheds lights on the extent to which a focal firm can keep benefiting from an increasing amount of financial resources that it has access to. Prior studies in the motion picture research ha ve long argued that bigger production budget can allow more f reedom in filmmaking and en ables the studios to add more appealing attributes to their films to improve their box office popul arity (Litman & Ahn, 1998; Basuroy et al., 2003 ; Elberse & Eliashberg, 2003 ) . However, my results show that the extent to which firm can continue to improve performance by simply adding more financial resources is limited. The quadratic term of the production budget is negatively significant indicating a diminishing marginal return of production budget on box office revenue. Moreover , t he statistics from multi ple spline function test indicate th e slope for the interval of product budget above 1.26 standard deviation (or above 86 percentile) becomes non - significant , suggesting that additional financial investments above this th reshold will not generate additiona l returns . There are important managerial implications associated with this finding. In recent years, the average size of the production budget has been continually rising in the film industry 22 . 22 For instance, the average production budget for a Bollywood films has risen from less than half a million dollars to more than 3 million USD betw een 2000 to 2014. Similarly, the Moti on Picture 78 Scholars commented that studios have been more and more fasc inated with pursuing the next massively successful blockbuster by increasingly co mmitting to making big - budget, global - marketed, event films ( Perren & Schatz, 2004). As my finding indicate s , however, any additional investment in the production budget abo ve a certain threshold will not lead to higher box office revenue. In other words, this challenges some scholarly contention that studios can assign bigger budgets to films as an insurance policy to attract more audiences since they cannot predict market t rend (Basuroy et al., 2003). To the contrary, t his research highlights that additional s pending on lavish p lots, visual effects, stunning locations , that typically characterize event films are likely to be risky decisions for the studios . Second , my findi ng shows that there are limits to the degree to which a firm can continue to improve performance by assigning more top stars to its film projects . T he effects of diminishing marginal return from an additional increase in the number of top stars on box offi ce revenue appear to be quite robust across different robust ness checks. Research on motion picture economics has long debated on whether or not studios should hire top stars (Elberse, 2007; Liu, Mazumdar, & Li, 2014; Ravid, 1999) . This study provides addi tional insights to managerial practice s in the motion picture industry regarding the benefits to be obtained from working with multiple top stars. Specifically, this study shows that while a studio might obtain benefits from the addition of one or two star s to a film, the addition of more stars will le ad to diminishing returns in terms of box office revenue. This implies what Pierce and Aguinis (2013) described as the Too - Much - of - a - Good - Thing effect when the additional allocation of beneficial resources cea ses to have a significant impact on the outcome beyond the inflection point and produce suboptimal result due to economic inefficiency. Association of America estimated the average production budget has increased by 36% from $55 million in 2002 to $75 million in 2007 . 79 Taken together, these findings provide valuable insights to the RBV literature regarding the extent to which having acc ess to a greater deal of resources can enable the firm to achieve SCA. Prior research in the RBV vein has raised concerns among managemen t scholars that criticize one of the underlying assumptions more is always better (Kraaijenbrink et al., 2010). However , few studies have examined whether firms can keep benefiting from increasing the amount of VRIN resources. This study is the first, to m y knowledge, to challenge this commonly held assumption and empirically examine whether firms can continue to improve p erformance by keep adding more of these resources. Furthermore, this study shows that the extent to which a firm can derive value from re sources it has access to is also influenced by the prior experience of the firm. In the section that follows, I discuss the findings and contributions regarding the role of a in extracting value from different types of resources . The Role of Prior Experience in Extracting Value from Resource s In t his study , I also explored how firm attributes may play a role in extracting more value from using the resources that it has access to and examine d how different types of firm prior experience affect its ability in obtaining value from its resources . In so doing, I buil t o n the growing body of research that increasingly emphasize s the importance of firm experience and explores the role of a firm in developing a competitive adva ntage from using available resources ( Holcomb et al, 2009; Cuypers et al., 2017; Sirmon et al, 2011). In particular, my overridin g arguments suggest that firms can leverage their broader experience to enhance their ability to make more effective use of their financial resources and human resources but not their partner resources while their narrower experience might be helpful to extract value from financial resources but n ot from their human resources or partner resources . 80 Although the findings show that the prior experience of a firm does have important implications on the degree to which the firm can benefit from its resource s , h owever, my predictions are mostly unsupported. T he results do indicate that having more narrow experience within a certain gen re can have a positive impact on the relationship between production budget and box office revenue . However, this relationship does not hold when the firm has broader experience across different film genres. This finding suggests that merely having more exposure to a diverse range of product categories is insufficient for firms to develop a better understanding of how to use fina ncial resources in different contexts. Instead, firms can learn to make better use of financial resources within the same p roduct category because firms gradually accumulate experience within a specific domain and develop their routines in performing simil ar tasks that are tied to this domain (Narayanan et al., 2009) . Again , this finding has important implication s for managerial practice s . Specifically, my results show that the prior experience of a firm can enhance the contribution of financial resources to box office revenue . This also suggests that the adverse impact of working with a constrained production budget i s even more pronounced among the experienced firms. In other words , more experienced firms can further improve performance when they have acc ess to an adequate amount of financial resources. In overall terms, the use of production budget complements the p rior experience of a firm . Stated differently , sufficient financial resources must be ensured in the presence of accumulated narrowly - defined experience in order to achieve a desirable outcome in terms of box office revenue . This is because firms gradually develop certain routines in performing tasks on the basis of having access to financial resources. As such, prior experience of firms can be 81 leveraged upon to perform these tasks in better ways only if the needed financial r esources have been secured. At the same time , the extent to which can influence its ability to obtain value from human resources is not contingen t upon the nature of its prior experience. In other words, the value to be obtained from the use of top stars is tied to all of the prior experience of the firm. Specifically, prior experience of a firm, defined either broadly or narrowly, negatively moder ates the relationship between the number of top stars and box office revenue. As predicted, the prior experience of a studio undermines its ability from making more use of top stars . In fact , the results indicate that greater e xperience of a firm could co mpensate for the lack of top stars on a film . Stated differently, stars are much more important to a film when the studio that is associated with it has less experience. In other words, this finding suggests a substitution effect between prior exp erience and the use of top stars in its films . While f eaturing top stars in a film experience. As scholars noted, top stars can bring box appeal to their fans (W a llace et a l., 1997). To some extent, t he idiosyncratic processes of working with the assigned top stars may even conflict with . This finding provides valuable insights to managerial practices . First, it highli ghts the importance of recruiting top stars in film projects amongst the inexperienced studios. Specifically, the use of top stars can enable the studio to improve its project performance in the absence of prior experience . The experienced firm s , however, should give very careful consideration about its reliance on top stars. Second, while a firm can best benefit from a higher production budget only when it has a higher level of experience , i t may be able to derive more 82 benefits from top stars when it has l ower levels of experience. This is consistent with recent research that suggests the prior experience affect the value to be obtained from different types of resources in distinct ways as a result of specific characteristics of these resources (Mannor et a l., 2016). Limitations T here are several limitations of this dissertation that I would like to acknowledge. Some of these limitations might contribute to the fact that a number of hypotheses did not receive support in this study. First, I theorized earlie r that enlisting a production partner could enable the focal firm to draw on a broader base of resources in managing a film project and lever age its partner s expertise to generate more box office revenues . In addition, it was predicted that the increase i n the number of production partners will lead to diminishing marginal retur ns . I n this study, however, the impact of having one or more production partners in a film project on its box office revenue was no t statistically significant. I t is possible that t his null finding can be explained by a competing theoretical explanation such that c ollaborating with another firm can incur communication and coordination issues which may neutralize the potential benefits to be obtained from working with others. However , I suspect that the empirical limitation associated with the measurement of produc tion partners in this study also contribute s to this null finding. When multiple production companies were listed on IMDB and no further clarification could be obtained from their own websites or Wikipedia, it was difficult to establish the primary production firm and to separate it from partners. In these cases, I select ed the primary production company as the one which had the most experience. In addition, some of these fil ms were jointly produced by up to eight production companies . It is possible that some of these production partners ma d e significant 83 contribution while others only played peripheral roles on these films . However, I was not able to explore the specific role s and the respective contributions of each partner in these partnership s due to data unavailability. Second, while there was some su pport for hypothesis 2a that predicted diminishing marginal returns from adding more top stars on the box office revenue of films in which they appear, there were relatively few films that had more than one or two stars. This made it difficult for me to demonstrate support for hypothesis 2b that adding more top stars above this level will lead to declined performance. The seem ingly inverted - U shaped relationship between the number of top stars in a film and the box office perf ormance of this film is not statistically significant. In particular, o nly 16 films in my sample had four or more top stars assigned to any given film. Th is is likely because I followed stringent rules in compiling the list of top stars in my sample after top stars that I adopted in this study is a lot s horter than many other studies in the motion picture industry that examine the impacts of stars ( Basur oy et al., 2003; Wallace, Seigerman, & Holbrook 1993 ). Finally , the overall lack of support for the interaction effects between project resources and fir m prior experience might be explained , to some extent, by measurement error of firm experience. Specifically, t he different types of experience of a studio w ere measured by its prior experience with managing film projects on the basis of film genres. While this is consistent with prior research using film gen re s to study the experience of studios (e.g. Shamsie et al., 2009), the identification of film genres in this study is mostly relying on BOI which only identified the primary genre of film s . Indeed, s om e scholars argue that the primary genre of a film determines the basis of the kinds of skills and expertise needed to produce the film (Perretti & Negro, 2007; 84 Gomery & Pafort - Overduin, 2011) . However, the assumption that the primary genre s of films listed on BOI always fully and a ppropriately describe each of the distinct product categories of films is unwarranted. In fact, some of the other sites such as IMDB may assign as many as three genres to a film. In addition, it is also possible th at a studio can gradually gain exposure to other kinds of films from the different elements of its prior films that we re outside of its focal domain. This , perhaps, can be better illustrated with an example. Rudraksh 23 , was categorized as an action film but it contains ce rtain elements of a Sci - Fi and crime/thriller film . The company, Karma Entertainment , that produced this film might have accumulated some experience with managing Sci - Fi and thriller films that can be leveraged upon in making its next film Tathastu that wa s categorized as c rime/ t hriller. As such, th e influence of a experience that can be leveraged upon in extracting value from its resources might be undermined by the measurement error of the experience. Future Directions Several addition al avenues can be identified for future research related to this dissertation. First, valuable insights can be obtained from f urther exploration of how partnership s with other firms contribute to project performance by adopting a more nuanced measure of p r oduction partners and delineating the specific roles of each partner . Specifically, management scholars can more convincingly ascertain the role of partnership in managing a project through uncovering the companies that take on l ead roles in collaborating with other s . In this regard, interviewing company executives to learn how each of the partners contribute s to a collaborative venture can significantly improve our understanding about how a focal firm can 23 Rudraksh (2004) was categorized as an Action film on BOI and Action, Fant asy, Sci - Fi on IMDB 85 leverage its partners as resources in man aging its projects. This will also allow us to further untangle the processes through which the benefits to be obtained from collaborat ing with others can be enha nced by leveraging its prior experience with different product markets. Next, how firms selec t their production partners on the basis of their prior experience is another interesting future avenue. Previous research has shown evidence that prior experienc e of collaborations of the focal firm has an impact on its ability to derive value from subseq uent collaborative ventures (Heimeriks & Duysters, 2007; Hoang & Rothaermel, 2005). It is rojects of different kinds can enable it to continue to derive more value from having collabor ative experience with other firms. In other words, it is not clear whether the experience of collaborating with others can be complemented or substituted by the f different types. Further exploration in this vein can also yield important insights to managerial work with ot her partners rather than work on its own. Finally , future studies in motion picture research can benefit from different types of measures perhaps through the use of other analytical approaches, to more fully capture the role of a experience with managing film projects of different kinds. Experience can be me asured on the basis of the production budgets of prior movies among other aspects. This will also us to examine the distinct mechani sms through which different kinds of prior experience produce distinct learning outcomes in different contexts (March, 2010; Eggers, 2012) . 86 REFERENCES 87 REFERENCES Acquaah, M. (2012). Social firm performance in a transition economy: A comparative analysis of family owned and nonfamily firms. Strategic Management Journal , 33(10), 1215 - 1228. Abernathy, W. J. & Wayne, K. (1974). L imits of the learning curve. Harvard Business Review, 52, 109 - 119 Aime, F., Johnson, S., Ridge, J. W., & Hill, A. D. (2010). The routine ma y be stable but the advantage is not: Competitive implications of key employee mobility. Strategic Management Journa l , 31(1), 75 - 87. Albers, S., Wohlgezogen, F., & Zajac, E. J. (2016). Strategic alliance structures: An organization design perspective. Jou rnal of Management , 42(3), 582 - 614. Albert, S. (1999). Movie stars and the distribution of financially successful films in the motion industry. Journal of Cultural Economics , 23(4), 325 - 329. Altman, R. (1984). A semantic/syntactic approach to film genre . Cinema Journal , 6 - 18. Amabile, T. M. (1997). Motivating creativity in organizations: On doing what you love and loving what you do. California Management Review , 40(1), 39 - 58. Amit, R., & Schoemaker, P. J. (1993). Strategic assets and organizational r ent. Strategic Management Journal , 14 (1), 33 - 46. Arend, R. J. (2006). Tests of the resource - based view: do the empirics have any clothes? Strategic Organization, 4(4): 409 - 422 Argote, L., Beckman, S. L., & Epple, D. (1990). The persistence and transfer of learning in industrial Management Science, 36(2), 140 - 154. Argote, L., McEvily, B., & Reagans, R. (2003). Managing knowledge in organizations: An integrative framework and review of emerging themes. Management Science , 49(4), 571 - 582. Argote, L., & Miron - Spektor, E. (2011). O rganizational learning: From experience to knowledge. Organization Science , 22 (5), 1123 - 1137. Armstrong, C. E., & Shimizu, K. (2007). A review of approaches to empirical research on the resource based view of the firm. Journal of Management , 33(6), 959 - 986. 88 Audia, P. G., & Goncalo, J. A. (2007). Past success and creativity over time: A study of inventors in the hard disk drive industry. Management Science , 53(1), 1 - 15. Bain, J. S. (1956). Barriers to new competition (Vol. 3, p. 55). Camb ridge, MA: Harvard University Press. Bain, J. S. (1964). The impact on industrial organization. The American Economic Review , 54(3), 28 - 32. Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1) , 99 - 120. Barney, J. B. (1995). Looking inside for competitive advantage. Academy of Management Perspectives , 9(4), 49 - 61. Barney, J. B. (2001a). Is the resource - management research? Yes. Academy of Manag ement Review , 26(1), 41 - 56. Barney, J. B. (2001b). Resource - based theories of competitive advantage: A ten - year retrospectiv e on the resource - based view. Journal of Management , 27(6), 643 - 650. Barney, J. B. (2002). Strategic management: From informed co nversation to academic discipline. Academy of Management Perspectives , 16(2), 53 - 57. Basuroy, S., Desai, K. K., & Talukdar, D. (2006). An empirical investigation of signaling in the motion picture industry. Journal of Marketing Research , 43(2), 287 - 295. Basuroy, S., Chatterjee, S., & Ravid, S. A. (2003). How critical are critical reviews? The box office effects of film criti cs, star power, and budgets. Journal of Marketing , 67(4), 103 - 117. Bendersky, C., & Hays, N. A. (2012). Status conflict in groups . Organization Science , 23 (2), 323 - 340. Benner, M. J., & Tushman, M. L. (2003). Exploitation, exploration, and process management: The productivity dilemma revisited. Academy of Management Review , 28(2), 238 - 256. Bissola, R., & Imperatori, B. (2011). O rganizing individual and collective creativity: Flying in the face of creativity clichés. Creativity and Innovation Management , 20 (2), 77 - 89. Bjorvatn, T., & Wald, A. (2018). Project complexity and team - level absorptive capacity as drivers of project ma nagement performance. International Journal of Project Management, 36(6), 876 - 888. Buscombe, E. (1970). The idea of genre in the American cinem a. Screen, 11(2), 33 - 45. 89 Choi, S., & McNamara, G. (2018). Repeating a familiar pattern in a new way: The effect of exploitation and exploration on knowledge leverage behaviors in technology acquisitions. Strategic Management Journal , 39(2), 356 - 378. Cicmil, S., Cooke - Davies, T., Crawford, L., & Richardson, K. (2009). Exploring the complexity of projects: Implic ations of complexity theory for project management practice. Project Management Institute. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences. Collis, D. (1994). How Val uable Are Organizational Capabilities? Strategic Management Journal , 15, 143 - 152 Conner, K. R. (1991). A historical comparison of resource - b ased theory and five schools of thought within industrial organization economics: do we have a new theory of the firm? Journal of Management , 17(1), 121 - 154. Clark, J. R., & Huckman, R. S. (2012). Broadening focus: Spillovers, complementarities, and specialization in the hospital industry. Management Science , 58(4), 708 - 722. Clark, J. R., Kuppuswamy, V., & Staats , B. R. (2018). Goal Relatedness and Learning: Evidence from Hospitals. Organization Science , 29 (1), 100 - 117. Crook, T. R., Ketchen Jr, D. J., Combs, J. G., & Todd, S. Y. (2008). Strategic resources and Strategic Management Journal , 29(11), 1141 - 1154. Cuypers, I. R., C uypers, Y., & Martin, X. (2017). When the target may know better: Effects of experience and information asymmetries on value from mergers and acquisitions. S trategic Management Journal , 38 (3), 609 - 625. Danneels, E. (2002). The dynamics of product innova tion and firm competences. Strategic Management Journal , 23(12), 1095 - 1121. Danneels, E. (2011). Trying to become a different type of company: Dynamic capabi lity at Smith Corona. Strategic Management Journal , 32(1), 1 - 31. Das, T. K., & Teng, B. S. (2002). Alliance constellations: A social exchange perspective . Academy of Management Review , 27 (3), 445 - 456. DeFillippi, R. J., & Arthur, M. B. (1998). Paradox in project - based enterprise: The case of film making. California Management Review , 40 (2), 125 - 139. De Silva, I. (1998). Consumer selection of motion pictures. The Motion Picture Mega - industry, 144 - 171. 90 De Vany, A., & Walls, W. D. (1999). Uncertainty i n the movie industry: Does star power reduce the terror of the box office? Journal of Cultural Economics , 23(4), 285 - 318. Desai, K. K., & Basuroy, S. (2005). Interactive influenc e of genre familiarity, star power, and critics' reviews in the cultural g oods industry: The case of motion pictures. Psychology & Marketing , 22(3), 203 - 223. Dosi, G., Nelson, R., & Winter, S. (Eds.). (2001). The nature and dynamics of organizational c apabilities. OUP Oxford. Egelman, C. D., Epple, D., Argote, L., & Fuchs, E . R. (2016). Learning by doing in multiproduct manufacturing: Variety, customizations, and overlapping product generations. Management Science , 63 (2), 405 - 423. Eggers, J. P. (2012). All experience is not created equal: learning, adapting, and focusing i n product portfolio management. St rategic Management Journal , 33(3), 315 - 335. Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: what are they? Strategic Management Journal , - 1121. Elberse, A., & Eliashberg, J. (2003). Dema nd and supply dynamics for sequentially released products in international markets: The case of motion pictures. Marketing Science , 22(3), 329 - 354. Elberse, A. (2007). The power of stars: Do star actors drive the success of movies? Journal of Marketing , 71(4), 102 - 120. Eliashberg, J., Elberse, A., & Leenders, M. A. (2006). The motion picture industry: Critical issues in practice, current research, and new research directions. Marketing Science , 25(6), 638 - 661. Ethiraj, S. K., Kale, P., Krishnan, M. S. , & Singh, J. V. (2005). Where do capabilities come from and how do they matter? A study in the software services industry. Strategic Management Journal , 26(1), 25 - 45. Fang, Y., Wade, M., Delios, A., & Beamish, P. W. (2007). International diversificatio n, subsidiary performance, and the mobility of knowledge resources. Strategic Management Journal , 28(10), 1053 - 1064. Finler, J. W. (2003). The Hollywood Story . Wallflowe r Press. Fong Boh, W., Slaughter, S. A., & Espinosa, J. A. (2007). Learning from exp erience in software development: A multilevel analysis. Management Science , 53(8), 1315 - 1331. tainable explanation of competitive advantage. Managerial and Decision Economics , 24 (4), 291 - 307. 91 Garzón - Vico, A., Gibbons, P., McNamara, P., & Rosier, J. (2016). Technological area experience, experience diversity and innovation. Technology Analysis & Strategic Management , 28 (9), 1041 - 1055. Goldberg, F. (1991). Motion picture marketin g and distribution. Getting movies. Gomery, D., & Pafort - Overduin, C. (2011). Movie history: A survey . Taylor & Francis. Gonzalez, C., & Madhavan, P. (2011). Diversity during training enhances detection of novel stimuli. Journal of Cognitive Psychology , 23(3), 342 - 350. Grant, B. K. (Ed.). (1977). Film genre: Theory and criticism . Scarecrow Press. Grant, B. K. (1986). Experience and meaning in genre films. Film genre reader , 114 - 128. Greve, H. R. (2003). A behavioral theory of R&D expenditures and inno vations: Evidence from shipbuilding. Academy of management journal , 46 (6), 685 - 702. Gruber, M., Heinemann, F., Brettel, M., & Hungeling, S. (2010). Configurations of resources and capabilities and their performance implications: an exploratory study on technology ventures. Strategic Management Journal , 31(12), 1337 - 1356. Gulati, R., & S ingh, H. (1998). The architecture of cooperation: Managing coordination costs and appropriation concerns in strategic alliances. Administrative Science Quarterly , 781 - 81 4. Gulati, R., Wohlgezogen, F., & Zhelyazkov, P. (2012). The two facets of collaborat ion: Cooperation and coordination in strategic alliances. The Academy of Management Annals , 6 (1 ), 531 - 583. Haas, M. R., & Hansen, M. T. (2005). When using knowledge can hurt performance: The value of organizational capabilities in a management consulting company. Strategic Management Journal , 26(1), 1 - 24. Harrell, F. E. (2001). Regression modeling strategies, with applications to linear models, survival analysis and logistic regression. GET ADDRESS: Springer . Haunschild, P. R., & Sullivan, B. N. (2002). Learning from complexity: Effects of prior accidents and incidents on airlines' learning . Administrative Science Quarterly , 47(4), 609 - 643. Heimeriks, K. H., & Duy sters, G. (2007). Alliance capability as a me diator between experience and alliance performance: An empirical investigation into the alliance capability development process. Journal of Management Studies , 44 (1), 25 - 49. 92 Helfat, C. E., & Peteraf, M. A. ( lifecycles. Strategic Management Journal , 24(10), 997 - 1010. Hoang, H., & Rothaermel, F. T. (2005). The effect of general and partner - specific alliance experience on join t R&D project performance. Academ y of Management Journal , 48 (2), 332 - 345. Holcomb, T. R., Holmes Jr, R. M., & Connelly, B. L. (2009). Making the most of what you have: Managerial ability as a source of resource value creation. Strategic Management Journal , 30(5), 457 - 485. Hoskisson, R . E., Wan, W. P., Yiu, D., & Hitt, M. A. (1999). Theory and research in strategic management: Swings of a pendulum. Journal of Management , 25(3), 417 - 456. Huber, G. P. (1991). Organizational learning: The contributing processes and the literatures. Orga nization Science , 2(1), 88 - 115. Huesch, M. D. (2013). Are there always synergies between productive resources and resource deployment capabilities? Strateg ic Management Journal , 34(11), 1288 - 1313. Strategic Management Journal 18: 75 98 Jauss, H. R. (1982) . Theory of Genres and Medieval Literature. Toward an aesthetic of reception , 76 - 109. Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. The American E conomic R eview , 76 (2), 323 - 329. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics , 3(4), 305 - 360. John, K., Ravid, S. A., & Sunder, J. (2017). Managerial ability and success: Evidence from the career paths of film directors. Journal of Corporate Finance , 44, 425 - 439. Joseph, J., & Gaba, V. (2015). The fog of feed back: Ambiguity and firm responses to multiple aspiration levels. Strategic Management Journal , 36 (13), 1960 - 1978. King, A. W., & Zeithaml, C. P. (2001). Competencies and firm performance: Examining the causal ambiguity paradox. Strategic Management Jou rnal , 22(1), 75 - 99. Klein, B., Crawford, R. G., & Alchian, A. A. (1978). Vertical integration, appropriable rents, and the competitive contracting process. The Journal of Law and Economics , 21(2), 297 - 326. 93 Koch, M. J., & McGrath, R. G. (1996). Improving labor productivity: Human resource management policies do matter. Strategic Management Journal , 17(5), 335 - 354. product of research creativity. Journal of Managemen t Studies , 37(1) Kor view of strategic management. Journal of Management Studies , 41(1), 183 - 191. Kor, Y. Y., & Mahoney, J. T. (2005). How dynamics, management, and governance of res ource Strategic Management Journal , 26(5), 489 - 496. Kor, Y. Y., Mahoney, J. T., & Michael, S. C. (2007). Resources, capabilities and entrepreneurial perceptions. Journal of Management St udies , 44(7), 1187 - 12 12. Kraaijenbrink, J., Spender, J. C., & Groen, A. J. (2010). The resource - based view: A review and assessment of its critiques. Journal of Management , 36(1), 349 - 372. Lampel, J., & Shamsie, J. (2003). Capabilities in motion: New or ganizational forms and the reshaping of the Hollywood movie industry. Journal of Management Studies , 40(8), 2189 - 2210. Lavie, D. (2006). The competitive advantage of interconnected firms: An extension of the resource - based view. Academy of Management Re view , 31(3), 638 - 658. Lawrence, M. (2018). Taking Stock of the Ability to Change: The Effect of Prior Experience. Organization Science , 29 (3), 489 - 506. Lee, B. K., & Lee, W . N. (2004). The effect of information overload on consumer choice quality in an Psychology & Marketing , 21(3), 159 - 183. Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of Sociology , 14 (1), 319 - 338. Levin, A. M., Levin, I. P., & Heath, C. E. (1997). Movie stars and authors as brand nam es: Measuring brand equity in experiential products. ACR North American Advances. Litman, B. R., & Ahn, H. (1998). Predicting Financial Success of Motion Pictures. BR Litman The Motion Picture Mega - Industry . Litman, B. R., & Kohl, L. S. (1989). Predict ing financial success of motion pictures: The'80s experience. Journal of Media Economics , 2(2), 35 - 50. Litman, B. R. (1983). Predicting success of theatrical movies: A n empirical study. The Journal of Popular Culture , 16(4), 159 - 175. 94 Liu, A., Mazumdar, T., & Li, B. (2014). Counterfactual decomposition of movie star effects with star selection. Management Science , 61(7), 1704 - 1721. Lorenzen, M., & Frederiksen, L. (200 5). The management of projects and product experimentation: examples from the music in dustry. European Management Review , 2(3), 198 - 211. Lorenzen, M., & Mudambi, R. (2012). Clusters, connectivity and catch - up: Bollywood and Bangalore in the global economy. Journal of Economic Geography , 13(3), 501 - 534. Magee, J. C., & Galinsky, A. D. (20 f power and status . Academy of Management annals , 2 (1), 351 - 398. Mahoney, J. T. (2004). Economic Foundations of Strategy . Sage Publications. view within the conversation of strategic management. Strategic Management Journal , 13(5), 363 - 380. Mannor, M. J., Shamsie, J., & Conlon, D. E. (2016). Does experience help or hinder top manage rs? Working with different types of resources in Hollywood. Strategic Management Journal , 37(7), 1330 - 1340. Malhotra, N. K. (1982). Information load and consumer decision making. Journal of Consumer Research , 8(4), 419 - 430. Malhotra, N. K. (1984). Reflections on the information overload paradigm in consumer deci sion making. Journal of Consumer Research , 10(4), 436 - 440. March, J. G. (2010). The ambiguities of experience . Cornell University Press. Mehra, A. (1996). Resource and market b ased determinants of performance in the US banking industry. Strategic Manag ement Journal , 17(4), 307 - 322. Miller, D., & Shamsie, J. (1996). The resource - based view of the firm in two environments: The Hollywood film studios from 1936 to 1965. Academy of Management Journal , 39(3), 519 - 543. Moliterno, T. P., & Wiersema, M. F. (2 007). Firm performance, rent appropriation, and the strategic resource divestment capability. Strategic Management Journal , 28(11), 1065 - 1087. rcing nature of power and status. Academy of Management annals , 2 (1), 351 - 398. 95 Nag, R., & Gioia, D. A. (2012). From common to uncommon knowledge: Foundations of firm - specific use of knowledge as a resource. Academy of Management Journal , 55(2), 421 - 4 57. Narayanan, S., Balasubramanian, S., & Swaminathan, J. M. (2009). A matter of balance: Specialization, task variety, and individual learning in a software maintenance environmen t. Management Science , 55(11), 1861 - 1876. Neale, S. (2000). Hollywood an d Genre . New York and London: Routledge Nelson, R., & Winter, S. (1982). An evolutionary theory of the firm. Belknap, Harvard, 41. Nerkar, A., & Roberts, P. W. (2004). Technological success of new product introductions in the pharmaceutical industry. Strategic Management Journal , - 799. and suggestions for future research. Strategic Management Journal , 28(2), 121 - 146. Newbert, S. L. (2008). Value, rareness, competitive advantage, and performa Strategi c Management Journal , 29(7), 745 - 768. Park, K. M. (2007). Antecedents of convergence and divergence in strategic positioning: The effects of performance and aspiration on the direction of strategic change. Organization Science , 18 (3), 386 - 402. Paulus, P. B., Dzindolet, M., & Kohn, N. W. (2012). Collaborative creativity Group creativity and team innovation. In Handbook of Organizational Creativity (pp. 327 - 357). Academic Press. Penrose, E. T. (1959). The theory of the growth of the firm. New York: Shar pe . Perren, A., & Schatz, T. (2004). Hollywood . Published in the Sage Handbook of Media Studies, (2004), pp. 495 - 515. Perretti, F., & Negro, G. (2007). Mixing genres and matching people: A study in innovation and team composition in Hollywood. Journal of Organizational Behavior: The International Journal of Industrial, Occupational and Organizational Psychology and Behavior , 28(5), 563 - 586. Managerial and Decision Economics , 24(4), 309 - 323. view. Strategic Management Journal , 14(3) , 179 - 191. 96 Pierce, J. R., & Aguinis, H. (2013). The too - much - of - a - good - thing effect in management. Journal of Management , 39 (2), 313 - 338. Pitelis, C. N. (2004). Edith Penrose and the resource - based view of (international) business strategy. Internationa l Business Review , 13(4), 523 - 532. Porter, M. E. (1980). Competitive strategy: Techniques for analyzing industries. Competitors, The Free Press. New York. Porter, M. E. (1985). Competitive advantage: creating and sustaining superior performance. 1985. New York: F ree Press, 43, 214. Priem, R. L., & Butler, J. E. (2001). Is the resource - strategic management research? Academy of Management Review , 26(1), 22 - 40. Ravid, S. A. (1999). Information, blockbusters, and sta rs: A study of the film industry. The Journal of Business , 72(4), 463 - 492. Rossman, G., & Schilke, O. (2014). Close, but no cigar: The bimodal rewards to prize - seeking. American Sociological Review , 79(1), 86 - 108. Rothaermel, F. T. (2001). Complementar advantage: an empirical study of industry and firm effects in the biopharmaceutical industry. Research Policy , 30(8), 1235 - 1251. Rumelt, R. P. (1984) 'Toward a strategic theory of the firm'. In R. Lamb (ed.), Competitive Strategic Management , Prentice - Hall, Englewood Cliffs, NJ, pp. 556 - 570. Salvato, C. (2009). The contribution of event - sequence analysis to the study of organizati onal routines (pp. 68 - 102). Northampton, Massachusetts: Edward Elgar Pub lishing. Schatz, T. (1981). Hollywood genres: Formulas, filmmaking, and the studio system . McGraw - Hill Humanities/Social Sciences/Languages. Schilling, M. A., Vidal, P., Ployhart, R. E., & Marangoni, A. (2003). Learning by doing something else: Variati on, relatedness, and the learning curve. Management Science , 49(1), 39 - 56. Shamsie, J., & Mannor, M. J. (2013). Looking inside the dream team: Probing into the contributions of tac it knowledge as an organizational resource . Organization Science , 24(2), 513 - 529. Shamsie, J., Martin, X., & Miller, D. (2009). In with the old, in with the new: Capabilities, strategies, and performance among the Hollywood studios . Strategic Management Journal , 30(13), 1440 - 1452. 97 Shane, S. (2000). Prior knowledge and the d iscovery of entrepreneurial opportunities. Organization Science , 11(4), 448 - 469. Sirmon, D. G., Hitt, M. A., Ireland, R. D., & Gilbert, B. A. (2011). Resource orchestration to create competitive advantage: Breadth, depth, and life cycle effects. Journal of Management , 37(5), 1390 - 1412. Sirmon, D. G., Hitt, M. A., & Ireland, R. D. (20 07). Managing firm resources in dynamic environments to create value: Looking inside the black box. Academy of Management Review, 32(1), 273 - 292. Smith, K. G., Smith, K. A. , Olian, J. D., Sims Jr, H. P., O'Bannon, D. P., & Scully, J. A. (1994). Top management team demography and process: The role of social integration and communication. Administrative Science Quarterly , 412 - 438. Sobchack, T. (1975). Genre Film: A Classica l Experience. Literature/Film Quarterly , 3 (3), 196 - 204. Sobchack, T. (1988). The adventure film. Handbook of American film genres , 9 - 24. Sochay, S. (1994). Predicting the performance of motion pictures. Journal of Media Economics , 7(4), 1 - 20. Sonenshe in, S. (2014). How organizations foster the creative use of resources. Academy of Management Journal , 57(3), 814 - 848. Staats, B. R., & Gino, F. (2012). Specialization and variety in repetitive tasks: Evidence from a Japanese bank. Management Science , 58 (6), 1141 - 1159. Tatikonda, M. V., & Rosenthal, S. R. (2000). Technology novelty, project complexity, and product development project execution success: a deeper look at task uncertainty in product innovation. IEEE Transactions on engineering management , 47(1), 74 - 87. Ta ylor, A., & Greve, H. R. (2006). Superman or the fantastic four? Knowledge combination and experience in innovative teams. Academy of Management Journal , 49(4), 723 - 740. Thompson, K., & Bordwell, D. (1994). Film history . An introduction, 2 . McGraw - Hill: New York Vandaie, R., & Zaheer, A. (2014). Alliance partners and firm capability: Evidence from the motion picture industry. Organization Science , 26(1), 22 - 36. Wallace, W. T., Seigerman, A., & Holbrook, M. B. (1993). The role of actor s and actresses i n the success of films: How much is a movie star worth? Journal of Cultural Economics , 17(1), 1 - 27. 98 Strategic Management Journal , 5(2), 171 - 180. Winter, S. G. (1995). Four Rs of profitability: rents, resources, routines, and replication. In Resource - based and evolutionary theories of the firm: Towards a synthesis (pp. 147 - 178). Springer, Boston, MA. Wooldridge, J. M. (1989). A computationally sim ple heteroskedasticity and serial correlation robust standard error for the linear regression model. Economics Letters , 31 (3), 239 - 243. Yeoh, P. L., & Roth, K. (1999). An empirical analysis of sustained advantage in the US pharmaceutical industry: Impact of firm resources and capabilities. Strategic Management Journal, 20(7), 637 - 653. Zander, U., & Kogut, B. (1995). Knowledge and the spe ed of the tra nsfer and imitation of organizational capabilities: An empirical test. Organization Science , 6(1), 76 - 92. Zollo, M., & Winter, S. G. (2002). Deliberate learning and the evolution of dynamic capabilities. Organization Science , 13(3), 339 - 351 .