Collaboration and Creativity

“What a small world!” is a common response to discovering that a new acquaintance knows someone you do. The small-world phenomenon is so common, in fact, that scholars and researchers who specialize in network analysis study what they call “small world networks” to explain the dynamics of various social systems, including friendships, corporate alliances, scientific collaborations, the Internet, and business production teams.

Brian Uzzi, professor of management and organizations at the Kellogg School of Management, in his recent work with co-author Jarrett Spiro (Stanford University), has applied the concept of the small world network to the creative world of Broadway musical production teams. Uzzi and Spiro wanted to test their theory that general truths about how social networks organize and govern success can be applied to the study of different social systems. Published in the American Journal of Sociology, their paper is the first of its kind to give insight into those networks in terms of their performance and rates of success versus failure. This research is especially important for creating a more thorough understanding of the key properties and conditions that lead to different effects of small world networks.

“As the world is moving away from the predominance of organizations organizing activities to the rise of freelancers, including in places like Hollywood and Silicon Valley,” said Uzzi, “it is especially important for us to understand how those economies perform, and what helps them to perform better as they change over time.”

Having grown up in New York City and feeling an affinity for Broadway, Uzzi is interested in artistic creativity as well as the study of social networks. “My goal is to have people recognize that success isn’t just based on internal talent and knowledge,” explained Uzzi. “Success is partially derived from relationships with other people, through whom they get access to expertise and capabilities beyond themselves.” Through his two main research areas, social network analysis and complexity theory, Uzzi models innovation and creativity in different industries by looking closely “at how individual capabilities are limited and transcended through the network.” In other words, it’s not just what you know—it’s also who you know.

Rather than using the conceptual framework of “six degrees of separation” popularized by Stanley Milgram’s well-known 1967 study of social networks, Uzzi and Spiro argue that Broadway networks are closer to the parlor game “Six Degrees of Kevin Bacon,” in which players try to connect the actor to other actors based on what movies they have performed in together. The production of Broadway musicals requires six key, freelance production professionals: a composer, a lyricist, a librettist who writes the plot and dialogue, a choreographer, a director, and a producer. They operate as teams that often work together on multiple shows and serve as professional networks for each other. As Uzzi and Spiro demonstrate, these small world networks, or teams, govern behavior that shapes the level of connectivity and cohesion among the social actors.

They find that the success of new musical productions in the twentieth century relies upon two key parameters: the ratio of new blood versus industry veterans, and the degree to which incumbents involve their former collaborators and serve as brokers for new combinations of production teams. The success of a musical production depends on the team, rather than individuals, and introduces new methodological questions for how to study such collaborative networks and successful innovation.

To test their model and its theoretical hypotheses, Uzzi and Spiro studied a data sample consisting of 2,092 people who worked on 474 musicals from 1945 to 1989. Their statistical model used new methods of examining small world networks by supposing that people working on musicals interact in dense and overlapping groups or clusters, and that social ties are repeated over time. Their analysis also controlled for alternative factors that affect a show’s success, including talent, economic and geographic conditions, and artists that were inactive for lengthy periods of time. They also included rare data on forty-nine musicals that died in pre-production. Failure data is often inaccessible to researchers studying networks and decreases potential bias in their analyses. Uzzi and Spiro measured the success or failure of a show based on the industry gold standard—critics’ reviews and whether the production was a financial success (if it recouped its costs). These clustered networks, as Uzzi and Spiro show, allow for the specialized resources of teams to blend and mingle, inspiring creativity and innovation in producing new musical hits.

However, as the researchers argue, small worlds can become too insulated. Uzzi and Spiro learned from their statistical analysis that the small world networks that help to create success or failure in Broadway musicals, at both artistic and financial levels, have a parabolic effect and a “bliss point,” after which the small world networks of musical production are too connected. In these cases, the networks face liabilities in the realms of innovation and collaboration that impede their creating new, successful musical hits. In other words, too much small-worldliness can undermine the very benefits it creates at more moderate levels, due to a decrease in artists’ ability to innovate and break convention.

Reflecting on the results, Uzzi remarked, “This paper shows that if you take a talented person, then you can see that his talent becomes amplified or inhibited based on his connections. Also, we show that mediocre talent is brought up by the structure of these networks more than the superstar, even though it helps the superstars, too. The key takeaway is that everybody does better.”

Uzzi did a follow-up study, published in Science in December 2005, to test his model’s application to the world of peer-reviewed science publishing. He examined articles and authors publishing in four different areas of science: social psychology, economics, ecology, and astronomy. He and his collaborators found similar patterns of small world networks and success in science publishing, as well. “We can now ask the question, ‘What areas of science are going to produce the big hit?’” he said. What was especially interesting in comparing the different disciplines was that each had a different pattern of team clustering. In publishing peer-reviewed articles on astronomy, small world networks are most productive at a highly clustered part of the curve, while publishing articles on social psychology is more productive at the bliss point.

In terms of its practical application to the business world, Uzzi said, “This information translates into who does better or worse, makes money or doesn’t. That’s quite important in today’s business environment. Business firms and organizations are increasingly thinking about their strategies and strategic alliances, so thinking about and analyzing their networks will be especially significant in these ways.”

Further readings:

Guimerà, Roger, Brian Uzzi, Jarrett Spiro, and Luís A. Nunes Amaral (2005). “Team Assembly Mechanisms Determine Collaboration Network Structure and Team Performance.” Science 308(5722): 697-702.

Milgram, Stanley (1967). “The Small World.” Psychology Today, 2: 60–67.