«Essays on Collaboration, Innovation, and Network Change in Organizations by Russell James Funk A dissertation submitted in partial fulﬁllment of ...»
Essays on Collaboration, Innovation, and Network
Change in Organizations
Russell James Funk
A dissertation submitted in partial fulﬁllment
of the requirements for the degree of
Doctor of Philosophy
in The University of Michigan
Professor Jason D. Owen-Smith, Chair
Professor Gautam Ahuja
Professor Mark S. Mizruchi
Assistant Professor Maxim Vitalyevich Sytch
c Russell James Funk 2014
All Rights Reserved For Kylee, whose love and encouragement made this dissertation possible.
ACKNOWLEDGEMENTSA central thesis of this dissertation is that ideas are often stimulated and developed with the help of a good network, and this dissertation itself is no exception to that claim. Since coming to the University of Michigan, I have been incredibly fortunate to have the support of an outstanding network of mentors, collaborators, and friends who, beyond their contributions to this dissertation, helped me develop as a scholar and person.
I am especially grateful to Jason Owen-Smith, my dissertation committee chair, for his outstanding mentorship. I ﬁrst ran into Jason’s research while working on my undergraduate thesis. I have a distinct memory of sitting in the Regenstein library, reading some of his articles, and thinking, Wouldn’t it be neat if I could do research like that someday? Little did I know that I’d have my chance just a few short years later. Jason made my decision to accept Michigan’s oﬀer to enroll in their sociology PhD program a no-brainer. After I moved to Ann Arbor in the summer of 2008, he graciously invited me to work on his U.S. Knowledge Economy (USKE) project, even before I had set foot inside a graduate school classroom. It is hard to imagine a better way to learn the craft of research than simply digging in and getting one’s hands dirty.
Although I’m sure it must have slowed his own progress, Jason encouraged this style of learning on the USKE project and all of our subsequent collaborations. I am deeply thankful to Jason for being so generous with his time, for his always-helpful feedback, and most of all, for pushing me to ask questions and pursue ideas far beyond what I had thought I was capable of when starting the PhD program. I hope that I can be iii the same kind of mentor to my future students.
I am also incredibly appreciative of the support and encouragement of the other members of my dissertation committee, Mark Mizruchi, Maxim Sytch, and Gautam Ahuja. Mark’s intellectual curiosity, breadth of knowledge, and methodological contributions have been deeply inspiring to me as a budding academic. I am also grateful to Mark for organizing Michigan’s economic sociology community, which has served as my primary intellectual home throughout graduate school. Anyone who has had the privilege of sitting in a workshop with Mark will know that he has an uncanny ability to (constructively) sniﬀ out weak assumptions and faulty logic. Whatever ability I have to make good arguments owes much to watching Mark ply this craft and to his generous feedback over the years on my own written work and presentations.
Maxim pushed me in many ways to expand my methodological horizons. Although his inﬂuence pervades each chapter of this dissertation, it is especially apparent in the last two, which build on insights from his research on network communities, network dynamics, and their respective connections to innovation. Over the past few years, I’ve also had the privilege of working with Maxim on several other projects. These collaborations have given me an incredibly valuable window into Maxim’s systematic and rigorous approach to scholarship and, in so doing, improved this dissertation and my capabilities as a researcher.
Last but not least, I am indebted to Gautam for teaching me about the interconnections between sociology and strategy. I ﬁrst began to see the exciting possibilities at the intersection of these two ﬁelds in Gautam’s doctoral seminar on corporate strategy. That class is also where I completed the ﬁrst drafts of Chapter II, which beneﬁted greatly from his helpful guidance. Additionally, I am grateful to Gautam for always pushing me to look for bigger, higher impact theoretical contributions.
Beyond research, I owe Gautam special thanks for opening the doors to Michigan Strategy for me, which in many ways came to feel like a second home department.
iv To Jason, Mark, Maxim, and Gautam—I hope the completion of this dissertation marks just the beginning of many exciting collaborative projects.
I beneﬁted from the guidance and support of many other faculty members at the University of Michigan. Jerry Davis has in many ways been like an unoﬃcial member of my committee, who not only provided feedback on early drafts of some chapters, but also took the lead in organizing the Interdisciplinary Committee on Organizational Studies (ICOS) and other related Michigan communities that played a major role in shaping my research. I am also thankful to Sue Ashford, Bob Axelrod, Sarah Burgard, Michael Heaney, Greta Krippner, Sandy Levitsky, Ned Smith, Sara Soderstrom, Brian Wu, Mayer Zald, and Minyuan Zhao for valuable comments on drafts of chapters and closely related projects. John Hollingsworth deserves special thanks for giving me the opportunity to explore many of the ideas developed in this dissertation in the health care domain. I am also grateful to Brian Noble and Sharon Broude Geva for introductions in the scientiﬁc computing community.
Graduate school would not have been the same without my amazing group of friends and fellow PhD students. I am especially grateful to Dan Hirschman, for outstanding friendship and many fun collaborations, and Helena Buhr, for paving institutional trails at Michigan that helped make this dissertation possible. For camaraderie and intellectual support, I also thank Jon Atwell, Johan Chu, Natalie Cotton-Nessler, Maria Farkas, Spencer Garrison, Mikell Hyman, Julian Katz-Samuels, Heeyon Kim, Yong Hyun Kim, Suntae Kim, Sun Park, Tristan Revells, Kelly Russell, Todd Schifeling, Lotus Seeley, and Matt Sullivan.
This dissertation was made possible by generous ﬁnancial and technical support. I am particularly indebted to the National Science Foundation, the Rackham Graduate School, and Michigan’s Department of Sociology for fellowships that gave me time to build new skills and the opportunity to take risks on data collection and analysis.
Rick Smoke spent countless hours helping me with what must have appeared to v be an endless and seemingly random array of technical challenges, from interview transcription to web server administration. I also thank Mark Montague for his patience with my many questions about Michigan’s Flux computing cluster.
Finally, I thank my family. My mother, Diane, has always supported and inspired me in whatever I do, and instilled in me the love of learning from a very young age.
My father, Jim, together with my aunt, Linna Place, opened me to the idea of a career in academia and created big shoes that I am always seeking to ﬁll. I’m also very grateful to my brothers, Todd and Chris, and members of my extended family, including John, Wyn, and Bill Hyzer, for encouragement and inspiring conversations.
I dedicate this dissertation to my wife and partner, Kylee, for her endless support, for always inspiring me to do better, and for ensuring we enjoyed the journey through the ups and downs of graduate school. To Kylee—I look forward to sharing many more adventures in the next exciting chapter of our lives.
Chair: Jason D. Owen-Smith This dissertation examines how internal communication and collaboration networks inﬂuence organizations’ performance at innovation. Because some conﬁgurations may be better than others, I also consider strategies for changing networks. I structure my investigation around three studies.
The ﬁrst study examines the eﬀects of diﬀerent networks in diﬀerent geographic settings. Using data on 454 ﬁrms active in nanotechnology, I ﬁnd that sparse networks of inventors help geographically isolated ﬁrms retain diverse knowledge and promote innovation. By contrast, ﬁrms located close to industry peers beneﬁt from highly connected networks among their inventors that facilitate information processing.
In the second study, I examine the eﬀects of network structure in an investigation of brokers. A broker is a person connected to people who are not tied to each other.
Studies ﬁnd that brokers have better performance on many metrics. However, little is known about how brokers aﬀect their contacts. Using data on the networks of over 18,000 inventors at 37 pharmaceutical ﬁrms, I examine the eﬀect of connection to a broker. To disentangle causality, I focus on changes among inventor’s existing contacts, where the decision to connect was made before the contact became a broker xiii and therefore is exogenous to performance. I ﬁnd that although becoming a broker positively aﬀects performance, the opposite is true for having a connection to one.
After focusing on performance in the ﬁrst two studies, the ﬁnal study considers reshaping networks. Using data on 23 million exchanges among 1.3 million members of 25 technical communities, I examine how a common organizational feature— knowledge categorization systems—inﬂuences bridging. Bridging ties create and strengthen connections among otherwise distant people and therefore are powerful tools for adapting networks. Categorization systems facilitate bridging by helping people locate distant peers. However, they may also inhibit bridging. First, as a categorization system grows large, it becomes harder to use and people are less able to establish distant ties. Second, as a categorization system decouples from real expertise, its value for bridging diminishes. Finally, as norms of evaluation vary more widely in an organization, people make fewer exchanges with unfamiliar peers. All three ideas are supported.
Innovation is a social activity. Although we sometimes imagine that breakthroughs emerge from people working on their own—the scientist toiling away in her lab, the budding entrepreneur tinkering in his garage—today, perhaps more than ever, revolutionary new discoveries, products, and ideas are the result of collaborations.
It is easy to see why—social relationships help people achieve better outcomes.
Decades of research demonstrate that working with diverse collaborators helps enhance creativity by exposing people to diﬀerent ideas and perspectives. Highly connected individuals work more eﬃciently because they have knowledge about the capabilities of others inside their organizations and know where to look for assistance. And cohesive networks, with many dense interconnections, promote creative risk taking by fostering supportive environments where people feel comfortable sharing unconventional ideas.
As evidence accumulates about the importance of relationships, many organizations are taking an active role in cultivating and managing internal communication and collaboration networks. For example, in November 2013, Microsoft announced the end of its infamous stack ranking system, whereby managers were required to rank the relative performance of their employees on a ﬁxed distribution, regardless of whether everyone had met or even exceeded expectations. One reason for abandoning the system was to promote a more connected company, which was limited by employee’s fears of working with others who may outshine them at ratings time.
Illustrations may also be found outside the corporate sector. In May 2012, for instance, the University of Michigan launched MCubed, a funding initiative designed to stimulate projects involving researchers from diﬀerent disciplines, who may otherwise not have the opportunity to work together.
Although there is substantial research showing how and why communication and collaboration ties help individuals perform better, surprisingly little is known about the eﬀects of larger internal network conﬁgurations—the aggregate of individual members’ interconnections—on broader organizational eﬀectiveness. Existing work, for example, oﬀers few theoretical tools to explain whether greater internal connectivity— like that sought by Microsoft and the University of Michigan—is likely to result in better outcomes. Will a connected Microsoft produce more breakthroughs? Scholars also have yet to consider the conditions under which deliberate eﬀorts at network change are likely to succeed. Will the University of Michigan’s bottom-up, researcherdriven approach to promoting integration lead to enduring ties across disciplines?
The purpose of this dissertation is to develop a conceptual framework and methodological approach and to present some preliminary evidence that help address these kinds of questions.
Intraorganizational networks, like all social networks, are deﬁned by a set of actors and the relations among them. Relevant actors may include divisions, teams, or people (Guler and Nerkar, 2012; Hansen, 1999; Hargadon and Sutton, 1997; Mizruchi et al., 2011). Unlike other networks, the membership and activities of intraorganizational networks are circumscribed by the boundaries of an organization.
Researchers have long understood the importance of internal networks for organizational processes and outcomes. For example, pioneering work by March and Simon (1958) argued that patterns of communication are related to an organization’s ability to manage uncertainty and to the distribution of power and inﬂuence among its units.
Similarly, in an early study of organization–environment relations, Burns and Stalker (1961) found that ﬂexible channels of internal communication (as opposed to rigid hierarchies) are beneﬁcial for ﬁrms that operate in more dynamic industrial sectors.