Self-confirming dynamics in Hollywood

Research by David Waguespack

In recent years, academic research has emphasized the benefits of embedded exchange, that web of friendships and prior relationships that drives business transactions in many different fields. The entertainment industry is particularly notorious in this respect; it is common wisdom that to be successful in Hollywood, “it’s not what you know, it’s who you know.” But this may be keeping the film industry from maximizing its own profit potential.

“Social Structure and Exchange: Self-Confirming Dynamics in Hollywood,” by David Waguespack, assistant professor of management and organization, with co-author Olav Sorenson, University of Toronto, examines the relationships between film distributors and production companies as a means of understanding the effects of self-confirming dynamics in embedded exchange. Film distributors receive a percentage of all domestic and foreign revenue from the films they distribute, so they have a strong incentive to maximize the revenue of the projects they support. But their own biased expectations may keep them from choosing the best and most profitable film projects.

The authors used data archived in the online Internet Movie Database (imdb.com) to follow the career paths of key figures in film-making, including producers, writers, lead actors and directors. Imdb, a volunteer- run site, tracks thousands of films and more than 2 million individuals in the motion picture industry throughout the course of their careers. The authors had access to a vast amount of data, including information about budgets, promotion and release timing for each film.

Waguespack used information from the industry daily Variety to track the films’ weekly box office performance. He also interviewed three film producers and five current or former senior distribution company executives and used a variety of secondary sources to describe the ways the industry works.

Examining more than 5,200 movies made between 1982 and 2004, Waguespack found that film distributors tend to work repeatedly with the same team of principals at a level much higher than could be expected due to chance alone. Films in which the production teams had prior interactions with the distributors stayed in theaters longer and earned more than four times as much as other projects in average ticket sales.

On the surface, this seems to indicate that distributors benefit from working with production teams they know. But Waguespack argues that this is actually a self-confirming dynamic—an effect that occurs because the distributors, believing that the films produced by their friends have greater commercial potential, create the circumstances for the film’s success.

This occurs because distributors have control of three key factors to a film’s success: budget, promotion and release date. Films with larger budgets, intensive promotion and favorable release dates are predisposed to be successful. After accounting for the factors that represent distributor effort, Waguespack found that films in which the principals have prior relations with the distributor actually performed worse at the box office.

What accounts for this negative effect? Marketing dollars and favorable release dates benefit higher quality films more than lower quality films, and it may be that distributors overallocate these scarce resources to production teams they know because they overestimate the quality of the work being produced. People tend to overestimate the quality of work offered by prior exchange partners. As they continue to work with the same partners, confirmation bias prevents them from changing their initial opinion of their partner’s work quality. Their beliefs become even more biased as they interpret ambiguous evidence in their partner’s favor.

Because distributors give preferential treatment to those with whom they have prior relationships, they may be forgoing a more beneficial relationship that would result in greater success and thus greater profits. In the end, the problem is not that the films of prior partners don’t do well—it is that another project might do even better. So relying on the web of prior relationships prevents the film industry from maximizing profits for both distributors and production teams.

Waguespack believes that the ways expectations produce self-confirming effects may be felt in a wide variety of markets that exhibit embedded exchange. Film-making is an entrepreneurial enterprise, with each film being a new project requiring the industry equivalent of venture capital. Perhaps venture capitalists are also creating or limiting the successes of businesses presented to them based on their own expectations. “What is really driving performance?” says Waguespack. “Is it that venture capitalists are really good at picking the best new businesses? Or is that success based on the effort that the investor puts into the company afterward?”

“Social Structure and Exchange: Self-Confirming Dynamics in Hollywood” was published in Administrative Science Quarterly. For more information about this research, please contact dwaguesp@rhsmith.umd.edu.