This is a lay summary for the article <Repeated Interactions and Improved Outcomes: An Empirical Analysis of Movie Production in the United States> that was first published on <Marketing Science> on February 2016. Authors of this research include Vishal Narayan and Vrinda Kadiyali.
PROBLEM
Movie production involves significant artistic and organizational coordination among the producer, director, screenwriter, actors, etc. Members of a production team might have worked in prior movies with the focal team’s and other teams’ members. Although it might seem intuitive that such repeated interactions among team members is positively related to team production success, much remains unclear and unexplored about the precise role of repeated interactions. This is the first study to empirically estimate the effect of team members’ repeated interactions in the movies industry.
DATA AND MODEL
This empirical work proceeds as follows. Researchers assemble a data set of 4,117 individuals who feature in 1,123 movies released in the 1999–2005 period. They measure the repeated interactions and experiences of five team members in each movie—producer, director, screenwriter, lead actor, and lead actress. Consistent with the movies literature, they measure a team’s production output by box office revenues. In addition, they also include a variety of other descriptors of movie revenues (like advertising expenditure, production costs, seasonal dummies, etc.)
To deal with the endogeneity issue while still accounting for unobserved heterogeneity across movies, Researchers apply GMM style estimators of dynamic panel data models that exploit lags and lagged differences of explanatory variables as instruments. These methods have the advantage of not relying on the availability of robust exogenous instruments. Methodologically our work is most closely related to this paper since our endogenous variables of interest do not vary over time either.
Results
Findings of this research are as follows:
1. repeated interactions have an economically significant impact on movie revenues.
2. The number of within-team repeated interactions influences revenue even after controlling for success of these interactions, experience, and success of individual team members as well members’ overall team experience (with those not in this focal team) and various movie character is tics. It is the number of repeated interactions that affects current productivity, not the success of these interactions.
3. The number of within-team repeated interactions matters more than the number of outside- team repeated interactions; Repeated interactions with members outside a focal team have no impact on focal product success. it suggests that the past interactions of team members with each other are a larger determinant of success than is their individual experience. Bringing a less experienced person with a strong history of collaboration with other members to the team is perhaps more beneficial than recruiting a more experienced new-to-the-team player.
4. Within-team repeated interactions matter more than experience or success of individual team members. it suggest that members who are not viewed by the consumer can have an important role in enhancing product success due to their past interactions. More generally, for product development teams (e.g., software and advertising) and for other team oriented business activities (e.g., business consulting, investment banking).
5. The producer’s within-team repeated interactions matter more than other members’ within-team repeated interactions,The interactions of the producer have the greatest revenue impact. We find that repeated interactions between the following pairs of team members in a focal team improve team output: producer–lead female, producer–director, and lead male–lead female.
An obvious view of how the producer affects movie success would be via her role of organizing the production process, e.g., by obtaining financing and distribution for a movie. However, this route of the producer’s impact of movie success is likely to be measured by her individual descriptors (number of movies made, success of past movies, etc.). We find that the producer’s repeated interactions with team members are most salient for movie success. This is consistent with the following view: the three drivers of improved outcomes in team production (lesser agency, lower transaction costs, and learning by doing) are most salient in pairwise repeated inter- actions featuring the producer. This finding points to the role of leaders in teams (the producer is akin to the CEO) rather than the day-to-day operational head (the director is akin to the COO). This finding also underlines the relative importance of the revenue- enhancing role of team members who consumers do not view on the screen.
Limitation
This is the first study to empirically estimate the effect of team members’ repeated interactions in the movies industry by making use of appropriate methods of controlling for endogeneity and unobserved heterogeneity in panel data on a wide range of movies released over seven years. Findings of this research are relevant for team-based production in other industries, like music, TV shows, theatre, etc.
This empirical research has the following limitations. Research have not examined repeated interactions among the production team outside of the five mem- bers, e.g., costume designers, cinematographers, etc. Additionally, they have not been able to measure the process of selecting team mem- bers. Also, they have been unable to conclusively estimate the extent to which agency is reduced because of repeated inter- actions or the extent of learning by doing or investments in relationship-specific assets.
Reference