University of Southern California Dissertations and Theses

Interorganizational knowledge networks: the case of the biotechnology industry

(USC Thesis Other)

Interorganizational knowledge networks: the case of the biotechnology industry

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

INTERORGANIZATIONAL KNOWELDGE NETWORKS: THE CASE OF

THE BIOTECHNOLOGY INDUSTRY

by

Lu Tang

__________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)

August 2007

Copyright 2007 Lu Tang
Acknowledgements
First and foremost, my gratitude goes to those scholars who have guided me in my
study. I would like to thank my advisor, Dr. Patricia Riley, who always
encouraged me to pursue research topics I am truly interested in, provided
insightful comments on my ideas, and always had confidence in me. I would also
like to thank Dr. Janet Fulk, whose advice and support have been invaluable in
sharpening my ideas, and Dr. Peter Monge, who taught me that “network” can not
only be a method but also a theory. Dr. Ian Mitroff has been an inspiration. He
always challenged me to think about social science research in an “all-level, all
quadrant” way.
Special thanks go to Dr. Gurinder Shahi, who introduced me to the
fascinating world of biotechnology and helped me get to know so many brilliant
scientists and business leaders, some of whom turned out to be my interviewees.
I also want to thank my fellow doctoral students at Annenberg. Friends
from my cohort: Li Ying, Anna Kostygina, and Wei Peng. I feel so happy that we
went through the ups and downs in graduate school with each other’s support and
so honored that we are now embarking on our new careers together. My
collaborators: Graig Hayden, Seungyoon Lee, and Bettina Heiss. It is so much fun
working with you!
ii
Finally, I am very grateful to my family, especially my husband, Degui
Zhi. Thank you for being interested in my study, helping me manage and
understand such a huge database, and putting up with me during the past 5 years.
iii
Table of Contents
Acknowledgements ii
List of Tables viii
List of Figures ix
Abstract x
Chapter 1. Interorganizational knowledge sharing in the biotech sector:
An introduction 1
Biotechnology industry and interorganizational knowledge network 3
Academic-industry knowledge sharing in the biotech industry 9
Preview of the following chapters 13
Chapter 2. Literature Review 16
Interorganizational relationship literature 16
Motivation of interorganizational relationship 17
Formation of interorganizational relationship 20
Forms of interorganizational relationship 21
Theoretical perspectives on interorganizational relationship 24
Knowledge sharing literature 34
The nature of knowledge 35
Characteristics of knowledge 40
Interorganizational knowledge sharing 46
Culture and knowledge sharing 50
The outcome of knowledge sharing 58
Academic-industry knowledge sharing 63
Firm characteristics 64
Characteristics of academic institutions 66
Geographic proximity 69
Professional culture 71
Trust 72
Summary 73
Chapter 3. Academic-industry knowledge sharing: an interview study 74
Theoretical development and research questions 74
Research Questions 76
Methodology for the interview study 79
Access and sampling 79
Semi-structured interview and interview protocol 81
Data collection 84
Data analysis 87
Chapter 4. Research Findings 89
RQ1: Why do academic research institutes and commercial biotech
companies share knowledge? 89
Biotech companies 90
iv
Academic institutions 92
Academic scientists 95
RQ2: How is knowledge shared between academia and industry? 97
RO2a: What knowledge do academia and industry share? 98
RQ2b: How is the knowledge shared between academia and
industry protected as intellectual properties? 104
RQ2c: What are the channels of academic-industry
knowledge sharing? 105
RQ3: How does professional culture (of academia and industry)
affect the knowledge sharing between the two communities? 118
The divergence of academic and industrial culture 119
The convergence of academic and business culture 127
Professional culture and knowledge sharing 130
Subcultures of knowledge sharing 134
Intraorganizational sharing vs. interorganizational sharing 139
Summary 140
RQ4: What is the role of informal knowledge sharing in academic-
industry knowledge sharing in biotech industry? 142
Goals of informal knowledge sharing 144
Trust and informal knowledge sharing 147
Communication technology and informal knowledge
sharing 151
Conclusion and limitation 152
Summary 152
Reliability and validity 156
Chapter 5. A theory of knowledge networks 158
Knowledge sharing in interorganizational networks 158
Strategic alliance network 160
Interlocking directorate network 161
Supply-chain network 162
Citation networks 162
Co-authorship network 164
Critique 165
Theoretical development: Transactional vs. interactive knowledge
networks 168
Characteristics of the knowledge sharing process 171
Characteristics of the knowledge networks 176
Methodology 184
Data 184
Operationalization 185
Sampling 186
Data analysis 186
v
Chapter 6. Result of network analysis 189
Descriptive statistics 189
Hypotheses testing 194
H1: The growth of interactive knowledge network follows
the logic of preferential attachment, while the growth of
transactional knowledge network does not show this pattern. 194
H2: An interactive knowledge network will demonstrate
higher (a) global-level centralization and (b) within-cluster
centralization than a transactional knowledge network. 196
H3a: In the interactive knowledge network, two nodes that
have ties in the past are more likely to share ties in the
future. 197
H3b: In the transactional knowledge network, two nodes
that have ties in the past are more likely to share ties in the
future. 197
H4a: When two nodes have transactional ties in the past,
they are more likely to have interactive ties in the future. 201
H4b: When two nodes have interactive types in the past,
they are more likely to have transactional ties in the future. 201
Discussion 202
Chapter 7. Discussion and conclusion 205
Theoretical framework 205
Research contexts and methods 206
Discussion of main findings 207
RQ1: what are the goals of academic-industry knowledge
sharing? 207
RQ2: How is knowledge shared between academia and
industry? 209
RQ3: How do the professional cultures of academia and
industry affect the knowledge sharing between the two
communities? 210
RQ4: What is the role of informal knowledge sharing in
academic-industry knowledge sharing in biotech industry? 213
H1: The growth of interactive knowledge network follows
the logic of preferential attachment, while the growth of
transactional knowledge network does not show this pattern. 213
H2: Interactive knowledge network will demonstrate higher
(a) global-level centralization and (b) within-cluster
centralization than transactional knowledge network. 214
H3a: In the interactive knowledge network, two nodes that
have ties in the past are more likely to share ties in the
future. 215
vi
H3b: In the transactional knowledge network, two nodes
that have ties in the past are more likely to share ties in the
future. 215
H4a: When two nodes have transactional ties in the past,
they are more likely to have interactive ties in the future. 215
H4b: When two nodes have interactive types in the past,
they are more likely to have transactional ties in the future. 215
Theoretical implications 215
Practical implications 217
For those participating in academic-industry knowledge
sharing 217
For those making policies regarding knowledge-intensive
industries 218
Limitations and directions for future research 219
References. 222
Appendices 253
Appendix A: Interview protocol-industry 253
Appendix B: Interview protocol-academia 256
Appendix C: List of Codes with Frequencies 259
Appendix D: Sample Codes and Quotations 263
Appendix E: List of Conceptual Networks Created 265
Appendix F: Visualization of interactive and transactional knowledge
network 1995, 2000, and 2006 266
vii
List of Tables
Table 2-1. A developmental model of knowledge (Styhre, 2003) 38
Table 2-2. An all-quadrant model of knowledge 40
Table 2-3. Characteristics of knowledge 40
Table 3-1. Demographic information of interviewees 85
Table 4-1. Channels of academic-industry knowledge sharing 106
Table 4-2. Characteristics of academic culture and industry culture 120
Table 5-1. Characteristics of the knowledge sharing process 176
Table 5-2. Number of deals 185
Table 6-1. Descriptive statistics of two networks 190
Table 6-2. Number of nodes by category 190
Table 6-3. 20 most connected organizations in two networks 191
Table 6-4. Number of ties 193
Table 6-5. Descriptive statistics of AIDS networks, 2002-2006 194
Table 6-6. Regression analysis results for hypotheses 4a and 4b 196
Table 6-7. Clustering coefficient of two networks 198
Table 6-8. MRQAP results for hypotheses 3 and 4 200

viii
List of Figures
Figure 1-1. The relationahip among major players in the biotech sector 9
Figure 2-1. Four modes of knowledge creation 61
Figure 4-1. Goals of academic-industry knowledge sharing 90
Figure 4-2. Relationship between the formality of the channel and the
articulatedness of knowledge 116
Figure 4-3. The relationship between the formality of channel and maturity
of the knowledge 117
Figure 4-4. Informal knowledge sharing 144
Figure 6-1. Number of license and research deals 1990-2006 191
Figure 6-2. Longitudinal centralization scores of two networks 193

ix
Abstract
We live in a knowledge society where intellectual capital is the main driving force
of social development and wealth creation. The competitiveness of today’s
organization lies in its ability to create, transfer, assemble, integrate, protect, and
exploit knowledge. This dissertation studies the interorganizational knowledge
sharing in the biotech industry.
Despite the fact that interorganizational knowledge sharing is inherently a
communication phenomenon, organizational communication literature has paid
little attention to understanding the process of interorganizational knowledge
sharing as a communicative process. This dissertation aims at further our
understanding of interorganizational knowledge sharing in the biotech industry
from an organizational communication approach. It develops a theory of
knowledge network that explicates the characteristics of interorganizational
knowledge sharing on both micro/individual and macro/organizational levels of
analysis. Furthermore, it applies the knowledge network theory to
interorganizational knowledge sharing in the booming biotech industry, a highly
knowledge-intensive industry, by looking at how knowledge is shared and created
in the process of academic-industry and industry-industry interactions.
This dissertation consists of two studies. Further, I used an interpretive
approach to analyze the knowledge sharing behaviors, attitudes, and perceptions of
those engaged in interorganizational knowledge sharing, especially in academic-
x
industry knowledge sharing. Interviews supplemented with ethnographic
observation offered insights into how different players in the knowledge sharing
process: company executives, company scientists, academic scientists, and
technology transfer specialists engage in and make sense of the academic-industry
knowledge sharing in the biotech industry. Special attention was paid to
understand how professional cultures affect the knowledge sharing attitudes and
behaviors.
The second part of the dissertation looked beyond academic-industry
knowledge sharing to study interorganizational knowledge sharing in a more
general sense. Based on an analysis of existing literature and the findings of the
interview study, I developed a theory of knowledge network identified two
distinctive types of knowledge networks: interactive knowledge network and
transactional knowledge network. Medtrack, an industry database that contains the
information on the interorganizational ties in biotechnology sector from 1989 to
2007 was used to examine the proposed theory.
xi

Chapter 1. Interorganizational knowledge sharing in the biotech sector: An
introduction
When knowledge is broadly distributed and brings a competitive advantage,
the locus of innovation is found in a network of interorganizational
relationships. To stay current in a rapidly moving field requires that an
organization have a hand in the research process. Passive recipients of new
knowledge are less likely to appreciate its value or to be able to respond
rapidly. In industries in which know-how is critical, companies must be
expert at both in-house research and cooperative research with such external
partners as university scientists, research hospitals, and skilled competitors
(Powell, Koput, & Smith-Doerr, 1996, p. 119).

We live in a knowledge society where intellectual capital is the central driving
force of social development and wealth creation (Teece, 2000). The
competitiveness of today’s organization lies in its ability to create, transfer,
assemble, integrate, protect, and exploit knowledge. Interorganizational knowledge
sharing constitutes an important part of knowledge management (Powell et al.,
1996).
Most of the existing research on interorganizational knowledge is found in
the industrial economics literature, the technology transfer literature and the R&D
literature. This line of research is usually guided by one of the two theories:
transaction-cost economics theory and social network theory. The transaction-cost
approach understands interorganizational knowledge sharing as economic
transactions based on formal contracts and attributes the motivation for
interorganizational knowledge sharing to the fact that acquiring knowledge from
external source is often cheaper than developing it in-house (e.g. Arora &
1

Gambardella, 1990, 1994). Knowledge sharing through a transaction-cost
economics approach is considered to be an economic interaction. Mutual self-
interest and reciprocity is the safeguard against IP infringement (Kogut, 1989). On
the other hand, the social network approach perceives interorganizational
knowledge sharing as regulated by social norms and values and argues that
contemporary companies, especially knowledge-intensive firms, make the strategic
decision to embed themselves in an interorganizational knowledge network.
Knowledge sharing, from this perspective is perceived to be a social interaction.
Despite the fact that interorganizational knowledge sharing is inherently a
communication phenomenon, the organizational communication literature has paid
little attention to understanding the communicative aspects of interorganizational
knowledge sharing. First, there is a lack of effort in trying to understand
interorganizational knowledge sharing on the individual level: how people involved
in the process perceive, make sense of, and participate in interorganizational
knowledge sharing. Second, most of the existing studies look at interorganizational
knowledge sharing as knowledge sharing through formal contractual channels, such
as patent licensing or citation without examining its informal dimensions.
This dissertation aims to further our understanding of interorganizational
knowledge sharing through an organizational communication approach. First, it
develops a theory of knowledge networks that explicates interorganizational
knowledge sharing on both micro/individual and macro/organizational levels of
2

analysis. Second, it applies the theory of knowledge networks to interorganizational
knowledge sharing in the booming biotech industry as an exemplar of a
knowledge-based industry by looking at knowledge that is shared and created in the
process of university-company and company-company interactions.
Biotechnology industry and interorganizational knowledge network
Dubbed as the technology of the 21
st
century, modern biotechnology had its seed
planted in 1973 with the breakthroughs on recombinant DNA (rDNA) technology
made by Hebert Boyer and Stanley Cohen (Powell et al., 1996; Quere, 2003). This
discovery led to the founding of Genentech, the first modern biotech company in
1976. Since then, the biotech industry underwent steady development in the 1980s
and a boom in late 1990s in concurrence with the IT industry. Today, more than
3000 biotech firms have been founded in the United States and 5000 worldwide
(Shahi, 2004).
Modern biotechnology, also known as the 3
rd
generation of biotechnology,
refers to those technologies that allow the manipulation of organisms on the
molecular level (Schacter, 1999). While the biotech industry started as an
alternative to the traditional approach to drug development and was based on the
selection of small molecules (Gambardella, 1995), today’s biotech industry has
expanded its scope to include five major areas: pharmaceuticals
1
, medical devices,

1
Drug development and manufacturing is by far the largest component of the modern biotechnology
industry. The traditional pharmaceutical industry relies on chemically synthesized molecules
derived from natural organisms. Scientists have to sift through literally millions and millions of
3

agribiotechnology
2
, environmental biotechnology
3
, and industrial biotechnology
4
.
Some industry analysts also include computing, such as bioinformatics as an
emerging area of the biotech industry (e.g., Bergeron & Chan, 2004).

small molecules to find a few that possess the therapeutic effect they are looking for. As a result, the
process of drug development is long and expensive. It takes an average of 12 years and an
investment between $200-$800 million to successfully bring a drug from the lab to the market
(Bergeron & Chan, 2004). This blockbuster model of drug development proved to be very
successful in the 1980s and 90s. However, it is difficult to sustain this large scale, capital-intensive
model of drug development today because more and more blockbuster drugs are going off patent
and the high cost of drug development has made drug prices prohibitively high. Biotechnology
becomes an increasingly important alternative to the traditional approach to drug development.
Rather than sifting though an astronomical number of molecules that are derived either from natural
organism or chemical synthesis, molecular biologists can design a drug to have a certain effect
based on their understanding of the relationship between structure and function at the molecular
level. This new approach promises to significantly reduce the time and cost of developing a new
drug (Shahi, 2004). As a result, traditional pharmaceutical companies, such as Pfizer, Merck, Eli
Lily, and GlaxoKlineSmith, are the second largest investors in biotechnology after the US federal
government.
2 Agribiotechnology refers to the application of genetic engineering techniques to the breeding of
new crops and new plants. For over a thousand years, people have been manipulating and
selectively breeding new varieties of existing species of crops to increase the productivity of these
crops. The new genetic engineering technique enables the manipulation of the genetic composition
of animals and crops in a direct and controlled manner and a much faster pace to produce desirable
traits such as high yield, pest-resistant, herbicide tolerant, draught resistant, disease resistant,
extended shelf life, and improved nutrition (Juanillo, 2001; Shahi, 2004). According to data from
the International Service for the Acquisition of Agri-biotech Applications (ISAAA), the acreage of
genetically modified crops has increased from almost nothing in 1996 to more than 200 million
acres in 2005.
3 Environmental biotechnology refers to the technologies using living organisms such as bacteria
for waste management and environmental clean-up. This process is also called “bioremediation.”
Microbial remediation and phytoremediation are two major types of bioremediation (Shahi, 2004).
Microbial remediation refers to the use of bacteria and fungi that breakdown hazardous waste and
pollutants into harmless components. Phytoremediation, on the other hand, utilizes the natural
functions of engineered plants to treat toxic chemicals.
4 Industrial biotechnology includes the use of the technologies provided by the advancement of
molecular biology and genetic engineering for a broad range of industrial purposes, mostly the
production of biomaterial and biofuel. Biomaterial is the material artificially constructed to replace
part of a living system, such as human organs, tissues, and skin (Bergeron & Chan, 2004). New
biotechnologies now enable the production of fuels based on clean and renewable natural living
organisms, such as agriculture products. The production of biofuels decreases the world’s
dependence on petroleum, which is both limited and unrenewable (Bergeron & Chan, 2004).
4

The biotech industry is one of the most knowledge-intensive industries in
the contemporary economy (Gertler & Levitte, 2005). The emergence and
development of modern biotechnology industry is fundamentally driven by the
advancement of life sciences research. Biotechnology companies are knowledge-
intensive firms capitalizing on their scientific and intellectual resources. An
interorganizational knowledge network that is composed of biotech companies and
academic research institutions creates a synergistic force that enabled the boom
times of the biotech industry in the past decade (Powell, 1998).
Biotech companies are knowledge-intensive firms (Alvesson, 1995, 2000)
that have a highly educated and highly professional work force that engages in
state-of-the-art scientific research (Shahi, 2004). More importantly, the market
value of biotech companies is based more on their intellectual capital rather than
conventional assets (Brown & Duguid, 1998) and their success comes from the
successful creation, development, and application of frontier scientific discoveries.
Knowledge sharing is essential to the survival of biotechnology firms. Due
to the high intensity of innovation and the soaring speed of knowledge turnover,
biotech companies cannot operate solely based on the knowledge generated
internally. Instead, they need to constantly acquire knowledge from outside the
company (Powell, 1998; Quere & Saviotti, 2002).
The biotech industry consists of several major players: large established
firms, dedicated biotech firms, academic research institutes and funding agencies,
5

and industrial investors. These players form a closely knit interorganizational
network engaged in resource exchange and knowledge sharing that simultaneously
fuels the exponential development of life science research and the biotech industry.
Figure 1.1 represents the structure of such a network, with solid arrows
representing the flow of financial resources and white arrows representing the flow
of knowledge.
Large established firms: Because of the promise of biotechnology, many
traditional pharmaceutical and chemical companies have begun using
biotechnology as an alternative and cheaper way to produce their traditional
products, such as drugs, chemicals, or agricultural products. These companies are
often repositories of enormous financial, technological, and human resources. Most
of them have a diversified portfolio of technologies and products (Quere &
Saviotti, 2002). On the other hand, some early biotech companies such as Amgen
and Genentech have developed into generalist organizations similar to
pharmaceutical companies. Established companies dominated the life science
industries until the emergence of dedicated biotech firms in the 1980s.
Dedicated biotech firms (DBF): DBFs are new entrants to the market
(Powell et al., 1996). These firms primarily utilize frontier life science discoveries
to produce goods or services. Very often DBFs are small companies with a narrow
portfolio, concentrating on one or at most a few technologies, and are thus
specialist organizations. They are often involved in biotechnology R&D and are the
6

“true pioneers of new technology” without much vertical integration (Quere &
Saviotti, 2002. p. 5).
Investors: Investors fuel the development of the biotech industry by
providing funding to start-up biotech companies. Two primary types of investors
energize the biotech industry: angel investors and venture capitals. Angel investors
usually are wealthy individuals with extensive knowledge and experience of the
biotech industry. They provide DBFs with a modest amount of initial funding to
develop early-stage technology, ranging from several hundred thousand to several
million dollars. They face extremely high risks, with the upside potential of
harvesting enormous financial returns. Compared to angel investors, venture capital
groups control much more of the money but they often come in later, when the
technology is more developed. In this way, they avoid the high risk of early stage
technology development, but at the same time, they often receive a lower rate of
financial return. Later when a biotech company goes public, the general public can
start to invest in the company through company stocks.
Academic research community: The academic research community is a vital
component of the biotech industry. It is the primary source of intellectual and
human capital essential to the development of the biotech industry. Academic
research institutes such as universities, medical research centers, and research
hospitals provide a fertile ground for basic and applied research in the life sciences.
7

Advances in basic research in life sciences are often driven by public
funding on both the federal level and the state level. Long-term public funding is
especially important for life science research because of the long research cycle and
high cost of such research (Bartholomew, 1997). The US government’s heavy
investment on health related sciences in mid twentieth century has arguably planted
the seeds of the modern biotech industry (Kenny, 1986). Federal funding for
academic research increased 12 times from 1973 to 2002 and reached a whopping
19 billion dollars in 2002, more than two third of which went to life science and
medical research (National Science Foundation, 2003)
The major players discussed above form a close-knit network of resource
exchange and dependency as demonstrated in Figure 1.1 (Solid arrows represent
the flow of financial resources and the white arrows represent the flow of
knowledge). In the biotech sector, the cycle of innovation usually starts in
academic labs in major universities and research institutions. Initially, basic
research is often funded by federal/state funding agencies or large established
companies through grants and research collaboration agreements. When this
research results in an early stage technology, a DBF would license the technology
from academia and develop it to the point when it is close to production.
Alternatively, university scientists can choose to establish spin-off companies,
which are a special type of DBF, to commercialize their own intellectual properties.
During this process, DBFs rely heavily on angel investors and venture capitalists.
8

Once the technology is developed to the point where it demonstrates clear market
potential, an established company would step in and purchase the technology or
perhaps the entire DBF in order to further develop, manufacture, market, and sell
the products. Many established companies have successfully internalized new
biotechnologies by partnering with or purchasing promising DBFs (Grabowski &
Vernon, 1994).

Established
firms
Dedicated
biotech firms
Academic
institutes
Investors
Research funding agencies

Figure 1-1. The relationahip among major players in the biotech sector
Academic-industry knowledge sharing in the biotech industry
The biotechnology industry is closely coupled with the academic research
community (Dalpe, 2003). Prior to the 1940s, there was little formal interaction
between academia and industry in the US, as the two communities carried out
parallel research independently (Geiger, 2004). The decades between 1940s and
1980s saw an increasing trend of universities feeding industry with new
9

knowledge, technology, and human resources. However, academic-industry
knowledge sharing during this period of time was highly sporadic and
uninstitutionalized. This can partly be attributed to the fact that at that point in time
the intellectual properties derived from publicly funded research projects belonged
to the funding agency, and consequently, academic institutions, industrial
companies, and investors lacked the motivation to collaborate with each other to
commercialize the university-based scientific discoveries.
In the early 1980s, the US congress pushed for the passage of a series of
legislation to encourage the commercialization of scientific discoveries made in
academia. The two most important ones are: the Bayh-Dole Act and the Stevenson-
Wydler Act (Rand, 2003). The University & Small Business Patent Procedure Act,
commonly known as the Bayh-Dole Act, gives universities and commercial
companies the rights to commercialize technologies that were developed using
federal money and the rights to nonpatentable intellectual properties such as trade
secrets (Schacter, 1999). As an amendment to Bayh-Dole Act, the Trademark
Clarification Act (1984) gave the same treatment to labs that are funded by federal
money operated by nonfederal entities. The Executive Order 12597 of 1987
provided the same provision to large corporations conducting R&D with federal
money. The Stevenson-Wydler Technology Innovation Act of 1980 promotes the
transfer of technologies from federal agencies to non-federal entities by requiring
each federal agency to establish an Office of Research and Technology application.
10

As a result of these legislative changes, the number of academic-industry
knowledge transfer deals skyrocketed in late 1980s and 1990s and academic-
industry knowledge sharing became very intensive. To illustrate the scale of this
development, from 1985 to 2001, the number of patents granted to US universities
rose from 589 to 3200, and industrial funding of university research increased from
$630 million to $1.896 billion (National Science Board, 2004). Universities
harvested enormous financial gains by collaborating with industry and
commercializing their intellectual property. For instance, Stanford University and
the University of California, San Francisco have made more than 200 million
dollars from their patents on rDNA discovered by Cohen and Boyer (Dalpe, 2003).
A close tie between the academic community and the biotech industry
enables the smooth transfer of new scientific discoveries from university labs to
commercial biotech companies and facilitates the transformation of the fruits of
basic scientific research into marketable products. The biotech industry depends on
academia more than any other industry, as Kenney (1986) pointed out in one of the
earliest books written on biotechnology industry,
Biotechnology, a science that is capable of being commercialized, has been
totally dependent on university research. In no other fledging industry have
university scientists played such an all-encompassing role. Other sciences
(organic chemistry, electrical engineering, computer science, and physics)
have undergone a transformation from a science to a technology, with some
scientists leaving academe to start companies. But none of their earlier
technologies was developed entirely in academia. The computer industry,
for example, attracted developers who were not university professor or
PhDs. Furthermore, professors left universities to launch computer firms;
they generally did not remain in the university, nor was it common for
11

administrators to attempt use professors’ work in a private industry as a
source of university income (p. 4-5).

The close tie between the academic community and the biotech community
leads to a very unique phenomenon: the emergence of biotech clusters around
centers of scientific excellence, i.e., leading universities and research institutes
(Audretsch & Stephan, 1996). Some of the most successful biotech clusters in the
US are located in California (the San Francisco Bay Area, San Diego, and Los
Angles), the Boston Area, and North Carolina. In order for a biotech cluster to
flourish, it usually contains a vibrant and productive research center with star
scientists (Audretsch & Stephan, 1996; Zucker, Darby, & Armstrong, 1998), a
visionary venture capital community (Audretsch, 2001), a group of private
companies that develop the scientific discoveries from the life sciences into
marketable products, a supportive policy environment (Audretsch, 2001), and a
mature business infrastructure (Shahi, 2004). Furthermore, research has shown that
industrial R&D is more clustered than manufacturing (Baptista, 1998). Some
consider the clustering of biotech puzzling because geographical distance should
not have played such a big role in the development of the biotech industry, or any
other high tech industry for that matter, when the cost of communication and
transportation has been reduced tremendously due to the development of advanced
communication technology and transportation technology (Audretsch & Stephan,
1996).
12

The afore-mentioned characteristics of the biotech industry make it an ideal
context for the communicative study of interorganizational knowledge sharing and
academic-industry knowledge sharing.
Preview of the following chapters
This dissertation consists of two studies. In the first study, I use an
interpretive approach to analyze the knowledge sharing behavior, attitudes, and
perceptions of those engaged in interorganizational knowledge sharing—
particularly with respect to academic-industry knowledge sharing in biotechnology.
Interviews are supplemented with ethnographic observation and together they offer
insight into the ways different players behave in the knowledge sharing process:
specifically, how company executives, company scientists, academic scientists, and
technology transfer specialists engage in and make sense of academic-industry
knowledge sharing in the biotech industry. Special attention is paid to professional
cultures and the ways they affect knowledge sharing attitudes and behaviors. A
symbolic-interpretive approach is used because it allows knowledge sharing to be
studied as a process rather than an outcome (Eisenberg & Riley, 2001; Weick,
1979). It also allows for a more subtle understanding of varied attitudes and
behaviors of the different players engaging in academic-industry knowledge
sharing while avoiding an essentialized view of academia and industry.
The second part of the dissertation supplements our understanding of
knowledge sharing by studying knowledge sharing networks in biotechnology.
13

Based on an analysis of the existing literature and the findings of the interview
study, I will develop a theory of knowledge networks that describes and compares
different types of knowledge networks. Medtrack, an industry database that
contains information on the interorganizational ties in biotechnology sector from
1989 to 2007 will be used to examine the proposed theory. The following gives an
overview of the subsequent chapters of the dissertation.
Chapter II provides a review of existing literature that informs an
understanding of interorganizational knowledge sharing through a communication
perspective, including the interorganizational relations literature, the knowledge
management literature, and the technology transfer literature.
Chapter III discusses the research questions that emerge from the
integration and analysis of literature, and presents the methodology of the first
study that utilizes interviews supplemented by ethnographic observations.
Chapter IV reports the findings of the interview study on academic-industry
knowledge sharing. Twenty-four in-depth interviews were conducted with four
panels of respondents: biotech business executives, industry scientists, academic
scientists/faculties, and technology transfer specialists from academic institutes.
The Atlas-ti program is used to aid in the thematic and content analysis of the
interview transcripts.
Chapter V develops a theory of interorganizational network behavior by
examining organizational networks as they engage in knowledge sharing. The
14

second part of the chapter presents the methodology for a network analysis of
interorganizational knowledge networks, based on a biomedical industry database.
Chapter VI reports and interprets the result of the network analysis.
Chapter VII summarizes the rationale, theories, and findings of the
qualitative and network studies. It discusses the theoretical and practical
implications of this dissertation. It also addresses the limitations and explores the
directions of future research.

15

Chapter 2. Literature Review
Interorganizational knowledge sharing, especially the sharing of knowledge
between academic institutions and commercial companies, is a powerful force
driving the development of knowledge-intensive industries such as the biotech
industry (Bartholomew, 1997; Cooke, 2002; Quere & Saviotti, 2002). It has
attracted the attention of researchers from a variety of disciplines such as
management, economics, and communication. Several bodies of literature
potentially contribute to the understanding of interorganizational knowledge
sharing, including the interorganizational relationship literature, knowledge
management literature and technology transfer literature. This chapter will be
devoted to a review of these three areas of research.
Interorganizational relationship literature
One line of research that informs to our understanding of interorganizational
knowledge sharing is the interorganizational relationship (IOR) literature.
Interorganizational relationship refers to the collaboration between two or more
organizations in which they try to achieve predetermined goals by performing
certain activities together. In the first section of the chapter, I will review some
important studies on the goals, formation, and forms of interorganizational
relationship as well as the theoretical frameworks underlying the study of IORs.
16

Motivation of interorganizational relationship
A significant part of the existing research on interorganizational relationships has
been focused on the motivations or antecedents of interorganizational relationships
(e.g. Oliver, 1990; Ring & Van De Ven, 1994; Schermerhorn, 1975). These studies
try to explain why and under what circumstances organizations that could
otherwise be competitors form cooperative interorganizational ties. Several major
categories of motivations have been identified, including: pooling resources,
sharing risk and costs, increasing productivity, and achieving institutional
legitimacy.
First and foremost, as identified by Schermerhor (1975)’s early study,
organizations form IORs to share a wide variety of valuable resources, including
financial resources (e.g. Blodgett, 1991; Chi, 1994; Faulkner, 1995; Hagedoorn,
1993b; Harrigan, 1986; Hennart, 1988), political resources (Howard Aldrich &
Mueller, 1982; Atler & Hage, 1993; A. K. Gupta & Lad, 1983; Kanter, 1989;
Oliver, 1990), and knowledge (Bartholomew, 1997; Doz, 1996; Hamel, 1991;
Hamel, Doz, & Prahaland, 1989; Inkpen & Crossan, 1995; Kogut, 1988; Powell,
1998; Powell et al., 1996).
The sharing of risks and costs is another major motivation that drives
organizations into IORs (Bartholomew, 1997; Contractor & Lorange, 1988; Das &
Teng, 1998; Hamel et al., 1989; Mariti & Smiley, 1983; Ohmae, 1987; Rockwood,
17

1983). Interorganizational relationships allow organizations to decrease the risk and
cost of a particular business venture.
Furthermore, organizations engage in IORs to increase their performance in
terms of speed and flexibility (Barringer & Harrison, 2000). On the one hand,
organizations pool their resources and technological skills so that their products
will reach the market faster and they can capture the first mover advantage
(Badaracco, 1991; Doz & Hamel, 1998; Faulkner, 1995; Hamel et al., 1989;
Larson, 1991). On the other hand, IORs provide an alternative to market and
hierarchies, which are more heavily regulated so that organizations have more
flexibility in dealing with the environment, compared to other governance
structures such as an acquisition or merger (Badaracco, 1991; Barringer &
Harrison, 2000; Hennart, 1988; Kanter, 1989; Powell, 1990; Segil, 1998).
In addition, organizations build IORs for institutional/legitimacy reasons,
especially when such a relationship is associated with a positive normative value
(Schermerhorn, 1975). For instance, companies seek to build IORs when they are
required to do so by third party organizations such as government agencies. Finally
organizations build IORs to block the progress of a mutual competitor (Badaracco,
1991; Contractor & Lorange, 1988; Koh & Venkatraman, 1991; Ohmae, 1989;
Shaprio & Willing, 1990).
While researchers try to explain the motivation for IORs in general, special
attention has been paid to understanding the motivations of a specific type of
18

IORs: technology alliances. In reviewing this body of literature, Hagedoorn
(1993b) identified three major categories of motives for interorganizational
technology cooperation: research, innovation, and market opportunity.
First, organizations build technology alliances to advance their own
research, whether it basic research or applied research (Hagedoorn, 1993b). Today,
with the quickly expanding knowledge base of high tech industries, many
companies, even large diversified companies, find it necessary to collaborate with
other companies and academic research institutions in order to gain access to the
knowledge and competence that is not internally available. While R&D is often
highly costly and risky, technology alliances allow companies to share the cost and
risk. Secondly, organizations engage in technology alliances to facilitate their
innovation projects (Hagedoorn, 1993b). On the one hand, organizations aim at
capturing the tacit knowledge of their partners in the process of collaboration. On
the other hand, there is an explicit technology transfer agreement whereby one
organization acquires a piece of knowledge or technology from another
organization. This explicit and implicit learning shortens the product development
cycle and speeds up the innovation process (Kale, Singh, & Perlmutter, 2000).
Third, technology alliances are built for the purpose of market entry and expansion
(Hagedoorn, 1993b). In a comparative analysis of a wide variety of industries
ranging from a high technology industry such as biotechnology, new materials, or
computers, to medium tech industries, including automotive, chemicals, and
19

consumer electronics, and from low tech industries such as food and beverage,
Hagedoorn (1993b) found out that high tech companies, compared to companies in
other industries, are more likely to form technology alliances for technology
complementarity and time reduction. On the contrary, market entry is seldom a
motivation for high tech companies to enter technology alliances.
Formation of interorganizational relationship
While a large number of studies have been focused on the antecedents of IORs, the
process by which such relationships are established remains relatively
understudied. One exception is Ring and Van de Ven (1994). Examining the
developmental process of cooperative interorganizational relationships, Ring and
Van de Ven (1994) proposed a model of IOR formation that involves three
repetitive stages: negotiation, commitment, and execution. In the negotiation stage,
organizations engage in formal bargaining and informal sense-making of each
other’s goals, potential investments, and the risks of the IOR. This is followed by
the commitment stage in which organizations reach agreements about the
governance and structure of the relationship based on formal and legal contracts as
well as informal “handshake” agreements. Finally, organizations execute the rules
established in building the IOR. During this process, as organizations and people
involved get to know each other better, they may base their interactions
increasingly on interpersonal and informal rules and less on formal contracts.
However, the execution stage is not the end of an IOR. In the process of
20

cooperation, when there are problems, disagreements, and conflicts, this process of
negotiation-commitment-execution will be enacted repeatedly. Finally, the
relationship dissolves with two possible results: either the goal of the IOR is
achieved successfully or the IOR fails to produce the desired result.
Forms of interorganizational relationship
An interorganizational relationship assumes a variety of forms. Different forms of
IORs are used to fulfill the different goals of IORs discussed previously. The IOR
literature identified several prominent forms of interorganizational relationships,
including: joint ventures, networks, consortia, alliances, trade associations, and
interlocking directorates (Barringer & Harrison, 2000).
Joint venture: Two or more organizations put certain parts of their resources
together to build a new and separate organization (Inkpen & Crossan, 1995). In
some cases, multinational corporations build joint ventures with local partners
when they want to enter a foreign market (Inkpen & Li, 1999). In other cases, joint
ventures are created to develop a critical mass of resources that is unavailable in
individual parent organizations (Hennart, 1988).
Network: A group of organizations form a network through social contracts
rather than legal contracts (Atler & Hage, 1993). Most of the research on
interorganizational networks looks at a focal organization and its relationship with
other organizations in the network, i.e., the ego network (Harrigan, 1986; Jarillo,
1988). Compared to vertical integration, an interorganizational network allows
21

organizations to focus on their core activities and outsource secondary activities to
others and in turn, achieve goals such as flexibility, speed, product development,
learning, blocking competition, etc. (Barringer & Harrison, 2000). One exemplary
network is the supply network of Toyota (Dyer & Nobeoka, 2000).
Consortia: A consortium is a specific type of joint venture. It involves a
group of organizations within a particular industry that have similar needs.
Consortia are most common in high technology industries (Kanter, 1989). They
foster cooperation among for-profit companies, government agencies, and non-
profit organizations, such as academic research institutions (Barringer & Harrison,
2000). Consortia provide a venue in which competitive companies can pool their
resources to conduct pre-competitive R&D (Child & Faulkner, 1998; Kanter,
1989). They also facilitates collective lobbying and the sharing of trade information
(Barringer & Harrison, 2000).
Alliance: Alliances refer to agreements among organizations to establish an
exchange or collaborative relationship without the sharing of ownership (as in joint
ventures) and without a central administrative structure (as in consortia) (Dickson
& Weaver, 1997). Organizations involved in alliances are only loosely coupled
with each other (Barringer & Harrison, 2000). Two of the most common categories
of alliances are marketing alliances and technological alliances (Das, Sen, &
Sengupta, 1998).
22

Trade association: Trade associations are loose organizations formed by
companies in the same industry with the goal of sharing industry information,
conducting training programs, and providing legal and technical advice (Barringer
& Harrison, 2000). One central mission of trade associations is political lobbying,
using the collective resource and influences of an industry to influence
governmental policy (A. K. Gupta & Lad, 1983; Oliver, 1990).
Interlocking directorates refer to the arrangement by which the executives
or board members of one company sit on the board of another company or the
executives of two companies sit together on the board of a third company (Burt,
1983). An interlocking directorate helps companies to gain access to resources
essential to their survival (e.g. Kaplan & Milton, 1994) and promotes the diffusion
of innovation across organizational boundaries (Granovetter, 1985; Rogers, 1983).
All these forms of interorganizational relationships can be evaluated along
four dimensions: degree of formalization, degree of intensity, degree of reciprocity,
and degree of standardization (Marrett, 1971). Formalization refers to the degree to
which IORs are officially sanctioned by participating organizations. The existence
of an intervening unit is an indicator of structural formalization. While
interorganizational ties differ in formality, they also differ in intensity, i.e., how
involved organizations are in the IORs. For instance, joint ventures and consortia
are more intense than trade associations. Reciprocity, as many theorists have
pointed out, is another important dimension of IORs. It refers to the extent to which
23

the exchanges between organizations are bi-directional or one-directional. The
degree of reciprocity is measured on three aspects: direction of the exchange
(unilateral or bilateral), the extent of agreement, and the power balance between
interacting organizations. Finally, standardization is the degree to which routines of
interorganizational exchange are established and exercised.
Theoretical perspectives on interorganizational relationship
Several theories inform the study of interorganizational relationships. Next, I will
review the three major theoretical frameworks that are important to the current
study on interorganizational relationships in general and interorganizational
knowledge sharing in specific: transaction cost economics, resource dependency
theory, and institutional theory.
Transaction-cost economics
Transaction-cost economics theory explains companies’ boundary expanding
activities through an analysis of the cost of entering into transactions (Williamson,
1975, 1985). The theory is based on two assumptions: bounded rationality and
opportunism (Williamson, 1993). Bounded rationality refers to the fact that when
making a decision, people or organizations have limited intellectual capacity, which
prevents them from scrutinizing all the information available and comparing all
possible solutions—or in other words, prevents them from being perfectly rational
(March & Simon, 1958). As a result, people often make suboptimal decisions.
24

Opportunism means that people are self-interested and will be untruthful and take
advantage of others when there is no safeguard against such behaviors (Williamson,
1993). Not everyone is opportunistic, however; the problem is that we cannot
predict who, exactly, is going to be opportunistic.
Williamson’s theory on transaction-cost economics started from an analysis
of the complex forms of legal contracts used in economic transactions and the
transaction costs of these contracts. Traditional economics mainly examined what
he called the “classical contract,” which refers to a strict short-term formal contract
between two organizations (Williamson, 1979, 1991). This is the form of contracts
used in an ideal market in which a particular organization that needs certain
resources from the environment at a particular time has plenty of potential sources
from which it can acquire the resources. Thus there is no dependence between this
company and any other company in the environment. The second type of contract
is the so-called neoclassic contract (Macneil, 1978; Williamson, 1979, 1991).
Sometimes one organization depends on another organization because the resource
the focal organization acquires from the other is not available from other sources.
In this situation, the focal organization wants to have repeated interactions with the
second organization. Negotiating and bargaining for each contract could be very
costly and impractical. As a result, organizations build semi-formal neoclassic
contracts that allow them to form a long-term relationship (Williamson, 1979,
25

1991). Finally, organizations can engage in relational contracting that is based more
on norms than legal contracts (Macneil, 1978; Williamson, 1979).
Different types of contracts are invented to meet the requirements of
different governance structures of transactions. Williamson identified three major
governance structures: market, hierarchy, and hybrid (Williamson, 1975, 1991).
The transaction cost of a particular governance structure is determined by three
factors: frequency of transaction, uncertainty in transaction, and asset specificity
(Williamson, 1985).
An interorganizational relationship is an alternative to pure market or
hierarchy (Barringer & Harrison, 2000). Interorganizational relationships, such as
joint ventures or strategic alliances, can help organizations prevent opportunism
and reduce transaction costs. For instance, Jarillo (1988) proposed that using
strategic networks as governance structures helps companies reduce transaction
costs. A strategic network is composed of a hub company and a few peripheral
companies. The hub company concentrates on those activities that are central to
their competitive advantage and outsource nonessential activities to other
companies in the network. All companies in the network can benefit from
specialization. At the same time, companies build long-term relationships in such a
network, which prevents opportunism (Jarillo, 1988).
26

Resource dependency theory
J. Kenneth Benson first proposed an understanding of interorganizational
relationships as a power relation. How much power an organization has is
determined by its position in the network (Benson, 1975). Following the logic of
Benson, Pfeffer and Salancik (1978) developed resource dependency theory, which
is based on the assumption that the most important goal of an organization is to
survive, which requires different types of resources that cannot all be generated
internally. Organizations thus depend on their environments for those resources
(Scott, 1992). Survival, therefore, is based on an organizations’ ability to manage
their relationships with the environment and other organizations in it (Mizruchi &
Yoo, 2002). According to Pfeffer and Salancik (1978), organizations can use either
network extension or network consolidation to control their dependency on other
organizations.
The social control of an organization comes from the interdependent
relationship between this organization and other organizations in its environment.
In social systems, interdependence is a common phenomenon that exists whenever
one social actor does not have complete control over the necessary conditions for
the achievement of its action goals. In studying interdependence, one needs to
differentiate behavior interdependence from outcome interdependence. Behavior
interdependence refers to the interdependence between the actions of two social
actors, while outcome interdependence refers to the interdependence between the
27

action outcomes of two social actors. Two types of outcome interdependence are
identified: competitive and symbiotic (Howard Aldrich, 1999). The
interdependence between two social actors is often asymmetrical and constantly
changing.
In order to survive, an organization needs to respond to the demands of its
environment, which are often in conflict with one another. In dealing with the
contradictory messages from its environment, the organization gives priority to the
demands of those elements of the environment on which it depends most (Pfeffer &
Salancik, 1978). Three factors decide the extent to which organization A depends
on organization B (Pfeffer & Salancik, 1978). The dependence of organization A
on organization B is first determined by the importance of the resource held by B to
A, which has two dimensions: the relative magnitude of the exchange and the
criticality of the resource. The second major determinant of dependence is B’s
discretion over the allocation of resource needed by A. This discretion is a major
source of power. B’s control of the resource is manifested in four ways: through its
possession of the resource, through its access to the resource, through its use of the
resource, and through its ability to regulate the possession, allocation, and the use
of the resource and in enforcing the regulation. Third, organization A depends on
organization B if B has the ability to possess and to allocate those resources
essential to the survival of A that are otherwise unavailable.
28

The dependence between two organizations is often two directional: while
B has power over A by controlling the resource important to A, A can have
countervailing power over B by possessing the resource B needs (Pfeffer &
Salancik, 1978). For the dependence between A and B to provide B with real power
over A, there must be asymmetry in exchange relationship. In other words,
organization B has power over organization A when A depends on B more than B
depends on A.
As the environment is composed of organizations of diversified interests, it
sends the focal organization conflicting messages, and the focal organization faces
the problem of managing the contradictory messages. The focal organization deals
with the conflicting demands by paying attention to those demands in sequence, by
concealing information, and by playing one influencing organization against
another (Pfeffer & Salancik, 1978).
While balancing the conflicting demands of the external environment may
secure the autonomy of the focal organization to some extent, the more effective
way of avoiding constraints is to avoid the conditions which demand compliance in
the first place (Pfeffer & Salancik, 1978). First of all, the focal organization can
control the generation of demands by manipulating the illusion of satisfaction to
avoid further demands by the environment, by restricting the information available
to influencing organizations, and by participating in determining the environmental
demand. Second, the focal organization can manipulate the definition of
29

satisfaction. Third, the focal organization can control the formation of demands
through professionalization and self-regulation, through its involvement in setting
standards and regulatory policies, and through advertising and other marketing
methods. Fourth, the focal organization can just not comply with the environment,
but this strategy might not be effective when the focal organization operates in a
tightly regulated, or very competitive environment or an environment with other
powerful organizations. Finally, the focal organization can avoid the control of its
environment by reducing the visibility of its behavior though information control.
While the previously mentioned techniques are useful in avoiding the
influence of relatively less powerful groups, organizations need to use another set
of methods to manage the interdependence when they are faced with powerful
external organizations (Pfeffer & Salancik, 1978). First, the organization can adopt
organizational change strategies. The focal organization can adapt to the
environment or attempt to change the environment. By adapting to the
environment, the focal organization will assess the need of its environment and
adapt its production process to this need. Or, the focal organization can manage its
environment, for instance, by choosing the market in which it will compete.
Second, the focal organization can use strategies to avoid resource dependence by
developing sustainable exchanges and through diversification. Third, the focal
organization can use strategies to avoid the control of powerful organizations
30

though the use of laws, norms, or strategies of absorption and regulation such as
mergers, cooptation, and the exchange of personnel.
Based on resource dependency argument, a company enters
interorganizational relationships in order to decrease its dependence on other
organizations and increase other companies’ dependence on the focal organization
(Pfeffer & Salancik, 1978). For instance, large pharmaceutical companies often
enter interorganizational relationships with small dedicated biotech firms to take
advantage the cutting edge R&D in biotech firms, while dedicated biotech firms
establish IORs with large pharmaceutical companies to utilize other types of
resources, such as financial resources, manufacturing and marketing capabilities,
etc. (Fisher, 1996).
Institutional theory
According to classical organization theories, organizations operate in pursuit of
efficiency, however, the core argument of institutional theory is that organizations
are not only motivated by a need for efficiency but also a need for legitimacy. In
their seminal paper, Meyer & Brown (1977) laid down the basic assumptions of
institutional theory. First, the adoption of formal structure is influenced not only by
the concerns of internal technical efficiency but also by external/environmental
forces. Second, the observation of dominant formal structure can improve the social
evaluation and increase the survival chance of organizations. Third, there might be
a gap between the formal structure of the organization and the actual activities of
31

organizational members. Following the work of Meyer and Brown (1977),
institutional theory has developed in two major directions: institutional
isomorphism and loose-coupling theory (Donaldson, 1995).
One branch of institutional theory focuses on institutional isomorphism.
Asking why today’s organizations look so similar to each other, DiMaggio and
Powell (1983) proposed that organizations are shaped by three types of institutional
forces: coercive isomorphism, mimetic isomorphism, and normative isomorphism.
Coercive isomorphism refers to organizations’ adoption of certain features due to
regulations from state or other regulating agencies. Mimetic isomorphism refers to
organizations’ implementation of certain features to appear similar to other
organizations in the same industry or with similar functions. Organizations are
influenced by normative isomorphism when they adopt certain structures or
features to comply with professional norms. Organizations are especially likely to
engage in institutional isomorphism when organizational goals are ambiguous,
when many organizations depend on the vital resources provided by a few
organizations, and when technologies are uncertain (DiMaggio & Powell, 1983).
Zucker and collogues studied the adoption of certain organizational features, such
as civil service regulations and discovered that while early adopters are often
motivated by technical efficiency concerns, later adopters were simply conforming
to institutional norms set by early adopters (Tolbert & Zucker, 1983; Zucker,
32

1987). Several empirical studies supported this theory (e.g. Donaldson, 1987;
Rumelt, 1974).
The second branch of institutional theory offers a loose-coupling theory of
organizations. For instance, Meyer and Scott (1983) pointed out that when
organizations are forced to adopt a particular organizational form, strategy or
technology under institutional pressure, there is often a gap between the appearance
of conformity and the actual operation.
According to the institutional theory, organizations’ participation in
interorganizational relationships is more out of their pursuit of legitimacy than their
concern of efficiency (Barringer & Harrison, 2000). A newly established company
can increase its visibility, reputation, and legitimacy in general by establishing
strategic relationships with large and well-known companies, which is important
for the survival of new companies (e.g. Wiewel & Hunter, 1985). Furthermore,
partnering with an organization with socially desirable goals, such as a charity
organization or a university, may enhance the legitimacy of commercial companies.
Finally, companies want to participate in trade organizations and consortia to
appear normal (Barringer & Harrison, 2000).
The three theories reviewed above provide different theoretical frameworks
for the understanding of the motivations and characteristics of interorganizational
relationships. One special type of IOR is interorganizational knowledge sharing. In
33

the next section of the chapter, I will review the knowledge sharing literature with
special attention paid to interorganizational knowledge sharing.
Knowledge sharing literature
We live in a knowledge society in which intellectual capital is the central driving
force of social development and wealth creation (D. Bell, 1973; Castells, 1996;
Hardt & Negri, 2000). Manuel Castells (1996) calls it the “informational society,”
arguing that the creation, processing, and transmission of information is the
“fundamental source of productivity and power” (p. 21). Knowledge becomes a
source of competitive advantage for today’s organizations and knowledge
management becomes an essential part of management (Argote & Ingram, 2000;
Badaracco, 1991; Kalling & Styhre, 2003; Nonaka, 1994; Nonaka & Takeuchi,
1995; Teece, 2000). As a result, communication becomes a central process in
today’s society (Castells, 1996). Knowledge-intensive firms emerge as a distinctive
feature of the knowledge economy. As Teece (2000) aptly stated, “the essence of
the firm in the new economy is its ability to create, transfer, assemble, integrate,
protect, and exploit knowledge assets. Knowledge assets underpin competences,
and competences in turn underpin the firm’s product and service offerings in the
market” (p. 29).
Within such a historical context, how to manage knowledge effectively
becomes a major concern of researchers from disciplines such as management,
information system, economics, and organizational communication. Knowledge
34

management includes three inseparable aspects: knowledge creation, knowledge
transfer, and knowledge utilization (Schein, 2000).
Most of the existing studies on knowledge management present either an
information-based approach or an interactional approach. Studies taking the
information-based perspective examine how knowledge is captured, archived,
retrieved, and diffused (Demarest, 1997). This approach is often criticized for
reducing knowledge to information and for understanding knowledge management
outside of its organizational, social, and cultural contexts (Von Krogh, Ichijo, &
Nonaka, 2000). On the other hand, an interactional approach to knowledge
management looks at knowledge and knowledge management as embedded in
organizational and social contexts and focuses on the “people” aspect of knowledge
management: how people are connected and what motivate them to share
knowledge or not to share it (Demarest, 1997; Iverson & McPhee, 2002). In the
following section of the chapter, I will review the knowledge management
literature, focusing on three aspects of knowledge management: the nature of
knowledge shared, the mechanism of knowledge sharing, and the outcome of
knowledge sharing. Special attention will be paid to interorganizational knowledge
sharing.
The nature of knowledge
Knowledge is probably one of the most difficult concepts to pin down.
Philosophers have been debating about epistemology, or the nature of knowledge
35

for millenniums (Nonaka & Takeuchi, 1995). The rationalism tradition, started by
Plato, believes that knowledge is the result of mental exercise and logical
reasoning, while empiricist thinkers argue that there is no prior knowledge and
instead knowledge comes from our sensory experience of the world (Nonaka &
Takeuchi, 1995).
Most of the existing research on knowledge management makes the
distinction among data, information, and knowledge. Boisot (1998) summarized the
relationship among these three concepts in a succinct way: “knowledge builds on
information that is extracted from data” (p. 12). Data are the observation of actual
events and entities and are raw facts on which higher levels of knowledge are
constructed (Boisot, 1998; G. B. Davis & Olson, 1985). Data are meaningless until
they are interpreted by whoever possesses the data. This process of interpreting and
sense-making transforms data into information (Boisot, 1998). Information,
according to Nonaka (1994), “is a flow of messages or meanings, which might add
to, reconstruct, or change knowledge” (p. 15). One characteristic of information is
that it is transferable and can be communicated (Bierly, Kessler, & Christensen,
2000). Finally, knowledge implies a clear understanding of information and the
ability to use it. In other words, information is the bridge between data and
knowledge.
Most of the knowledge management studies are limited in that they are
based on a very restricted and incomplete definition of knowledge. These studies
36

usually adopt the Cartesian paradigm that is based on a split between mind and
body and based on a flatland view of the universe (Nonaka & Takeuchi, 1995;
Wilber, 2000). Ken Wilber (2000) criticizes the dissociation of mind and body in
today’s academic research, attributing it to the flatland view of Kosmos in those
scientific and philosophic traditions such as naturalism, physicalism, and scientific
materialism. Two important arguments made by Wilber will help us understand
knowledge in a more holistic way. First, he captures the state of existence of human
beings through the “Great Nest of Being.” This nest includes five levels of
existence: matter/physics, life/biology, mind/psychology, soul/theology, and
spirit/mysticism. The nest also delineates the developmental potentials of human
beings and human knowledge. Second, Wilber (2000) argues that modernity is
characterized by "the differentiation of art, morals, and science," the three of which
were previously fused (p. 60). While this differentiation has enabled unprecedented
developments in these three aspects respectively, it has also led to the
predominance of scientific materialism, which reduced the Great Nest of Being into
a mere "system of matters" (p. 61). To solve this problem, Wilber (2000) introduces
the Four Quadrants as the scheme of understanding humans and human society
through two dimensions: individual/the collective and interior/exterior. He
emphasizes the inadequacy of traditional science and argues that the interior and
subjective dimension of both the individual and the collective is as important as the
exterior and objective dimension that can be easily observed and measured.
37

A holistic model of knowledge includes two aspects. First, it is important to
understand knowledge through a developmental framework. In the past, knowledge
management literature looked at knowledge as a developmental continuum from
data to information to knowledge. Recently, some researchers proposed the concept
of wisdom as a form of knowledge on a higher level of development. Bierly,
Kessler, and Christensen (2000) defined organizational wisdom as “the judgment,
selection, and use of knowledge for a specific context” and it involves “both the
collection, transference and integration of individuals' wisdom and the use of
institutional and social processes (e.g., structure, culture, routines) for storage” (p.
597). The major distinctions between data, information, knowledge and wisdom are
summarized in Table 2.1. Interorganizational knowledge sharing literature has been
focusing on the sharing of lower level knowledge such as data or information.
Knowledge as an integration of information is sometime addressed, while wisdom,
as a higher form of knowledge, still remains unexamined.
Level Definition Learning process Outcome
Data Raw data Accumulating truths Memorization
(databank)
Information Meaningful, useful
data
Giving form and
functionality
Comprehension
(Information bank)
Knowledge Clear understanding
of information
Analysis and
synthesis
Understanding
(Knowledge bank)
Wisdom Using knowledge to
establish and
achieve goals
Discerning
judgment and
taking appropriate
action
Better living/success
(Wisdom bank)
Table 2-1. A developmental model of knowledge (Styhre, 2003)
38

Secondly, it is essential to understand knowledge in an all-quadrant way by
looking at the individual/collective and interior/exterior dimensions of knowledge
(See Table 2-2). On the upper left quadrant is individual-interior knowledge, i.e.,
tacit knowledge possessed by individual organizational members. On the upper
right quadrant is the individual-exterior knowledge, such as people’s technical
know-how that has been verbalized and can be easily transferred. On the lower
right quadrant is the collective-exterior knowledge, i.e., the explicit knowledge
possessed by the organization, such as company’s patents, products, and
technologies. Finally on the lower left quadrant is collective-interior knowledge
such as organizational norms, procedures, routines that are embedded in the daily
operation of the organization. For instance, Szulanski (1996; 2003) studied the
sharing of this type of knowledge as “best practices.” While the knowledge
management literature has tapped into all four quadrants of knowledge, the
literature on interorganizational knowledge sharing is still primitive in that it
mainly focuses on the lower-right quadrant, i.e., organizational knowledge that is
clearly articulated and can be easily transferred across organizational boundaries.

39

Upper left
Interior-individual I

Individual tacit knowledge
Upper right
IT Exterior-individual
Technical know-hows
Lower left
Interior-collective WE

Organizational procedures, routines,
best practices
Lower right
ITS Exterior-collective
Organization’s patents, products,
technologies
Table 2-2. An all-quadrant model of knowledge
Characteristics of knowledge

Scholars have made many attempts to explicate the complex nature of knowledge.
Based on current knowledge management literature, knowledge can be understood
in reference to the following dimensions: tacit vs. codified, rationalized vs.
embedded, migratory vs. embedded, knowing what vs. knowing how, individual vs.
collective, public vs. private, and stickiness. Table 2-3 presents a list of these
theories of knowledge and research studies related to these theories.
Categories of knowledge Authors
Tacit vs. explicit Polany (1958), Nonaka (1994)
Articulable vs. observable Winter (1987)
Rationalized vs. embedded Weiss (1999)
Migratory vs. embedded Badaracco (1991)
Individual vs. collective Matusik & Hill (1998)
Public vs. private Matusik & Hill (1998)
Knowing what vs. knowing how Ryle (1949), von Hippel (1988), Kogut &
Zander (1992), Quinn et al. (1996)
Table 2-3. Characteristics of knowledge
40

Tacit knowledge vs. codified knowledge
The concept of “tacit knowledge” was first proposed by Michael Polanyi (1958) to
refer to the knowledge we know but cannot tell. According to Polanyi (1958),
people form a holistic understanding of the world, part of which they don’t even
recognize and cannot articulate, and as a result, traditional study of knowledge as
codified information offers only a very limited understanding of human knowledge.
Later the term “tacit knowledge” came to be broadly used to describe any
“uncodified” knowledge, in opposition to codified information (Cowan, David, &
Foray, 2000, p. 212). Sometimes researchers use the terms “knowledge” and
“information” to distinguish between tacit and codified knowledge. In this case,
information is “knowledge reduced to messages that can be transmitted” (Ancori,
Bureth, & Cohendet, 2000, p. 259).
Tacit knowledge is inherently more difficult to share for a number of
reasons. First, as tacit knowledge is often embedded in the social, cultural, and
cognitive contexts, and people are often unaware of the tacit knowledge they
possess and thus unable to articulate it (Choo, 1998). Furthermore, individuals do
not need to externalize their tacit knowledge in order to benefit from it, and as a
result, they usually do not have the incentive to codify their tacit knowledge
(Stenmark, 2000-2001). Finally, the articulation and sharing of tacit knowledge
might cost people their competitive advantage and power, which further
41

discourages the sharing of tacit knowledge (Bordum, 2002; Leonard & Sensiper,
1998).
By converting its tacit knowledge into explicit knowledge, a company can
transform the personal knowledge and skills of its employees into organizational
knowledge that can be shared more easily and utilized by more people (Inkpen &
Dinur, 1998). However, in doing this, the company risks this knowledge being
replicated by its competitors and loosing its competitive advantage.
The dichotomy between tacit and explicit knowledge has its own critics.
First, it is pointed out that very often researchers fail to draw a clear line between
tacit knowledge and explicit knowledge. Wenger (1998) criticized this dichotomy
arguing that that the two are inseparable, “Explicit knowledge is … not freed from
the tacit. Formal processes are not freed from the informal. In fact, in terms of
meaningfulness, the opposite is more likely … In general, viewed as reification, a
more abstract formulation will require more intense and specific participation to
remain meaningful, not less.” (p. 67). Secondly, some researchers criticized the
overuse and mystification of the concept of tacit knowledge. For instance,
Donaldson (2001) pointed out that many researchers exaggerated the importance of
tacit knowledge while ignoring the fact that the dominant trend in knowledge-
intensive organizations is the rationalization and formalization of knowledge and
knowledge sharing process.
42

Rationalized vs. embedded knowledge
Weiss (1999) offers another criticism of the common belief that codified
knowledge is easier to share while tacit knowledge is more difficult to share by
arguing that the ease of knowledge sharing is a function of people’s willingness to
share it. As a result, she proposed a different way of categorizing knowledge:
rationalized knowledge and embedded knowledge. According to Weiss (1999),
rationalized knowledge is general and standardized knowledge that is independent
from a context. Rationalized knowledge is not related to its original source and thus
can be easily transferred. It often has wide applications. On the other hand,
embedded knowledge is context-specific, unstandardized, narrowly applicable and
may be personally or professionally sensitive. Due to its highly personalized nature,
embedded knowledge cannot be shared without the cooperation of the person who
originally possesses the knowledge.
Migratory vs. embedded knowledge
Very similar to the dichotomy between rationalized and embedded knowledge is
Badaracco’s notion of migratory and embedded knowledge. Badaracco (1991)
defined migratory knowledge as knowledge that is “packaged, articulated, and
mobile” (p. 35). Migratory knowledge is often transferred in products and designs.
Embedded knowledge, on the other hand, is context specific and harder to transfer
across organizational boundaries. Badaracco (1991) refers to the alliances that
enable the sharing of migratory knowledge “product links” and the alliances that
43

allow the sharing of embedded knowledge “knowledge links” (p. 79). An
organization’s embedded knowledge resides on different levels: individual, team,
and organization. Embedded knowledge presents a challenge to managers who are
used to acquire migratory knowledge through arm’s length transactions in the
market place.
Knowing what vs. knowing how
Another distinction commonly found in the literature is that between know-what
and know-how (e.g. Kogut & Zander, 1992; Quinn et al., 1996; Von Hippel, 1988).
Know-what refers to the information that can be codified, transferred, and decoded,
or in other words, the fact (Kogut & Zander, 1992). Know-what is often proprietary.
On the other hand, know-how is the accumulated and learned expertise or skills
that allow one to do something efficiently (Von Hippel, 1988).
Individual knowledge vs. collective knowledge
Another way of characterizing knowledge is to approach it in terms of individual
knowledge and collective knowledge. Individual knowledge refers to the total of
knowledge, information, and expertise possessed by individual organizational
members, while collective knowledge is the knowledge that resides on the
organizational level, including organizational routines, norms, practices (Matusik &
Hill, 1998; Zander & Kogut, 1995). A similar pair of concepts is personal
knowledge and organizational knowledge (Kogut & Zander, 1992). Individual
44

knowledge enables the proper functioning of individual organizational members,
while organizational knowledge allows the operation of an organization as a whole
(Bennet & Bennet, 2003).
Public vs. private knowledge
Another pair of dichotomies used in categorizing knowledge is public knowledge
vs. private knowledge (Matusik & Hill, 1998). Private knowledge only belongs to a
specific company while public knowledge is available to many companies, and
thus, public good. Only private knowledge can give a company its competitive
advantage (Matusik & Hill, 1998). Public knowledge is not a source of competitive
knowledge but the failure to use public knowledge is likely to lead to competitive
disadvantage (Matusik & Hill, 1998).
Stickiness of knowledge
In an attempt to answer the question, “why don’t best practices spread” within a
company, Szulanski (1996) proposed the concept of stickiness. Knowledge
stickiness is influenced by four factors: the nature of knowledge, the characteristics
of sources and recipients, and the knowledge-sharing context. In terms of the nature
of knowledge, its stickiness is a function of two properties: causal ambiguity and
usefulness (Szulanski, 1996). Causal ambiguity refers to the uncertainty in the
relationship between causes and outcome in the replication of best practices
(Szulanski, 1996). Causal ambiguity could be attributed to a failure to replicate
45

tacit skills involved or to a failure to understand the special characteristics of the
new context to which the practice is applied (Szulanski, 1996). Beyond the nature
of knowledge, stickiness is a function of the characteristics of the source, including
motivation and credibility, and the characteristics of recipients, including their
motivation, absorptive capacity and retentive capacity. Finally, the stickiness of
knowledge should be understood in the context of knowledge sharing, which
includes two dimensions: a barren organizational context and an arduous
relationship between source and recipient.
Interorganizational knowledge sharing
While traditional knowledge management research often centers on the creation
and sharing of knowledge within a particular organization (e.g. Argote, 1999;
Argote & Ingram, 2000; Argote, Ingram, Levine, & Moreland, 2000; Nonaka,
1994; Nonaka & Takeuchi, 1995), another trend has emerged that emphasizes the
study of knowledge sharing across organizational boundaries (e.g. Dyer &
Nobeoka, 2000; Owen-Smith & Powell, 2004; Oxley, 2004; Powell, 1998; Powell
& Brantley, 1992; Powell et al., 1996).
While the IOR literature examines the different governance structures of
interorganizational relations, similar theoretical frameworks have been used to
understand the mechanisms that underlie interorganizational knowledge sharing.
The knowledge management literature identifies three major mechanisms through
46

which knowledge is shared across organizational borders: the market, hierarchy,
and community/networks.
Market
Knowledge can be a commodity bought and sold in the economic market. In such a
market, interorganizational knowledge sharing is controlled by price mechanisms
(Adler, 2001). Many attribute interorganizational knowledge sharing to the
calculation of costs and benefits (e.g. Appleyard, 1996). It is argued that when the
development of a new technology within an organization is more expensive than
purchasing it from outside, the focal company will make the decision to buy. For
example, an organization can purchase knowledge from external sources through
patent licensing or material transfer agreements. Large pharmaceutical companies
often buy new technologies or patents from dedicated biotech firms when
developing the same knowledge in-house is more expensive. In the ideal situation,
organizations would engage in pure market exchange with other organizations for
the necessary resources that are not internally available.
Hierarchy
When the transaction cost of the market is too high, for instance when there is a
high level of uncertainty or when two organizations need to engage in repeated
interactions, the transaction is likely to take place within the hierarchy of
organization. In other words, the organization in question is more likely to choose
47

vertical integration so that the knowledge sharing is governed by internal
hierarchical forces rather than external market mechanisms. For instance, when a
large pharmaceutical company finds itself constantly negotiating with a dedicated
biotech firm to acquire latter’s technologies, skills, and other resources, the
pharmaceutical company may decide to build this capacity in house or to purchase
the dedicated biotech firm as a whole to internalize the knowledge base, skills, and
human resources in order to reduce the transaction costs.
Community
In today’s knowledge economy with its unprecedented high rate of innovation and
expanding knowledge base, it is impossible for organizations to access knowledge
solely through vertical integration. At the same time, market exchange mechanisms
prevent the sharing of knowledge across organizational borders. As a result, some
theorists emphasize community/network as an alternative mechanism of knowledge
sharing. Such a perspective emphasizes the social interaction and social contracts in
interorganizational knowledge sharing.
Granovetter (1985) argues that it is important to understand economic
transactions in the context of social relationships. He says that human beings and
organizations are embedded in all types of social relations. Market process and
other economic actions are also subject to the influence of social relations for at
least the following two reasons. First, pure market behavior is non-existent. Many
business transactions, including fairly complex ones, are conducted based on the
48

long-lasting relationships between business partners (Macaulay, 1963). Second,
people that engage in transactions usually refrain from resorting to terms of
contract to resolve conflict. Reliance on market standards to solve conflict can be
extremely damaging to business relationships.
Along a similar line, the literature on community of practice offers another
version of knowledge management following the logic of community. Community
of practice (CoP) refers to informal collaboration among a group of people working
more or less within the same area that emerges from their day-to-day
communication and interaction for the purpose of knowledge sharing, problem
solving, and personal development (Wenger, McDermott, & Snyder, 2002).
Communities of practice benefit organizations in a number of ways (Wenger &
Snyder, 2000). They facilitate the transference of best practices and implicit
knowledge, assist the professional development of participating members, help
companies recruit and retain talents, and ultimately lead to new business
opportunities (Wenger & Snyder, 2000, p. 17-20).
More recently, Adler (2001) argues that community is a more appropriate
metaphors for knowledge management in today’s economy. Accordingly, trust,
instead of price, is the most effective controlling mechanism in the community of
knowledge sharing (Adler, 2001).
Bell et al. (2002) conducted participatory action research on three large
multinational firms and concluded that the sharing of process knowledge among
49

these firms follow the logics of both market and community. More specifically,
companies based their decision to share knowledge with other companies on the
rationale of the market, but their actual knowledge sharing practices show elements
of community.
Some researchers have been debating the relative effectiveness of these
different mechanisms of knowledge management in facilitating knowledge flow in
different contexts. For instance, Almeida, Song, and Grant (2002) analyzed the
patent citation pattern in the semiconductor industry and found out that
multinational corporations in the form of hierarchy are more effective than strategic
alliances or markets in promoting cross-border knowledge sharing because they
allow more flexibility and the simultaneous use of multiple channels.
Culture and knowledge sharing
Culture is another variable influencing both intraorganizational and
interorganizational knowledge sharing. Knowledge developed in one cultural
context cannot be easily transplanted into another cultural context, even when these
two cultures are in the same organization (Hutchins, 1996; O'Dell & Grayson,
1998). A relatively small number of studies examined the effect of culture on
knowledge management practices in an attempt to identify the culture most
conducive to productive knowledge management (Schein, 2000). Both national
culture and organizational culture have been studied.
50

National culture
National culture is often considered as one important factor affecting the sharing of
knowledge within and among organizations. Hofstede (1980) defines national
culture as “the collective programming of the mind which distinguishes the
members of one human group from another” (p. 25). He identified five dimensions
of national culture: collectivism/individualism, power distance, uncertainty
avoidance, masculinity/femininity, and orientation towards time (1991, 2003).
Hofstede’s dimensions of national culture, especially the dimensions of
collectivism/individualism and power distance, are widely used in studying the
effect of national culture on organizational knowledge sharing. Chow and
colleagues studied the effects of these national cultural characteristics on
knowledge sharing behaviors in companies in the Western/Anglo cultural context
and the Eastern/Chinese cultural context. For instance, they compared how
subordinates share private knowledge with their superiors in the US and Taiwan
though an experiment and found out that when there is no face-to-face interaction,
Chinese managers are less likely to distort information than their American
counterparts (Chow, Hwang, Liao, & Wu, 1998). In a second study, they explored
the influence of national culture on informal knowledge sharing in Australian and
Taiwanese firms through interviews, focusing on several dimensions of national
culture such as collectivism/individualism, power distance, and the concern about
“face” (Chow, Harrison, McKinnnon, & Wu, 1999). It is found that Taiwanese
51

managers are likely to share private knowledge due to their concerns about
collective responsibility while Australian managers often do so for individual
reasons. However, Taiwanese managers are less likely to share potentially
personally damaging knowledge compared to Australian managers, when there is
the presence of a superior. In a third study, they examined the effect of a culture’s
collectivism/individualism on company manager’s knowledge sharing practices
(2000). They surveyed 142 managers from the US and China and discovered that
Chinese managers are more likely to share knowledge to the benefit of the
collective when there are conflicts between individual interests and collective
interests. At the same time, Chinese managers are less likely to share knowledge
with someone who is not an insider due to their concerns of the interest of their
own group.
The characteristics of national culture affect not only intraorganizational
knowledge sharing but also interorganizational knowledge sharing. Appleyard
(1996) surveyed employees in the semiconductor industry in both the US and Japan
and found that Japanese and American employees engage in interorganizational
knowledge sharing in very different ways. More specifically, Japanese employees
are more likely to use public channels such as newsletters, trade journals, and
conferences, while American employees are more willing to use private channels of
interorganizational knowledge sharing, including telephone and face-to-face
meetings. This difference can be attributed to the constructs of
52

collectivism/individualism and uncertainty avoidance dimensions of national
cultures. Because Japanese culture is more collectivist in nature, Japanese
employees are more likely to share knowledge with colleagues in the same
company, but are less likely to share knowledge with outsiders, especially through
personal/informal channels of communication. Furthermore, because the Japanese
culture emphasizes uncertainty avoidance and thus job security, Japanese
employees usually stay with their employers for a long time, which limits the
number of external contacts they have.
Organizational culture
On the organizational level, organizational culture is a crucial factor in knowledge
management and knowledge sharing (De Long & Fahey, 2000; Knapp & Yu,
1999). The organizational culture literature historically makes the distinction
between the studies that view organizational culture as a root metaphor and those
that view it as a variable (Smircich, 1983).
The study of organizational culture as a root metaphor views organization
as a cultural phenomenon and studies the production of meanings, cognitive maps,
symbols, rituals, metaphors, and folklores in organizations. To them, culture is not
something an organization has, but something an organization is (Smircich, 1983).
In this subfield, the study of organizational culture has been guided by three major
theoretical perspectives: the integration perspective, the differentiation perspective,
53

and the fragmentation perspective (Frost, Moore, Louis, Lundberg, & Martin, 1991;
Martin, 2002).
The integration perspective is characterized by its definition of culture as
consensus and clarity (Martin, 1992). For instance, Schein (1991) defines culture as
shared values which are based on shared underlying assumptions about the
organization and are manifested in artifacts such as organizational structure and
process. According to him, those cultural elements that are not shared do not
belong to the organizational culture. The research using this perspective looks for
consensus, clarity, and stability in organizations.
The differentiation perspective, on the other hand, acknowledges the
contradictions and inconsistencies in the culture of an organization (Martin, 1992).
More often than not, research adopting this perspective conceptualizes
organizational culture as composed of a set of subcultures. For instance, in his
famous study of the culture of Disneyland Theme Park, Van Maanen (1991)
identified the existence of different subcultures among the employees of
Disneyland based on their job status (bilingual tour guide, ride operators, and lowly
sweepers) and addressed the conflicts among these subcultures. While the
organizational culture is characterized by contradictions and inconsistencies, each
subculture is coherent in itself and shared by its members (Martin, 1992).
Finally, the fragmentation perspective emphasizes the ambiguities and
complexities in modern organizations and argues that there is no consensus either
54

on the organizational level or on the subculture level (Martin, 1992). It looks for
those cultural manifestations that are “neither clearly consistent nor clearly
inconsistent with each other” (Frost et al., p. 115). For instance, in her study of
social workers, Meyerson (1991) finds that ambiguity is a normal part in the work
of social workers because of their marginal positions in hospitals. Little consensus
exists on the definition of social work, and people react to this ambiguity in very
different ways.
The study of organizational culture as a variable finds its academic root in
the study of organizational climate in the 60s and 70s (Denison, 1996). The studies
in this tradition treat organizational culture as something that the organization has.
For instance, comparative management studies often treat culture as an independent
variable, a background/contextual factor that is used to explain the difference and
similarities found in other aspects of organizations (Smircich, 1983). Another
example of culture-as-a variable is the corporate culture studies, in which
researchers treat organizational cultures as produced by the organization (thus
manageable) and seek to find out the characteristics of organizational culture that
can be manipulated to enhance the performance and productivity of the
organization (Smircich, 1983).
Organizational culture has always been considered as a critical aspect of
knowledge sharing and knowledge creation (Armbrecht et al., 2001; B. Gupta, Iyer,
& Aronson, 2000; Lopez, Peon, & Ordas, 2004). Most of the existing studies on
55

organizational culture and knowledge sharing have taken a functionalist approach
in trying to identify the objective traits of the kind of organizational culture
conducive to knowledge sharing and knowledge creation. For instance, Park,
Ribiere and Schulte (2004) studied the relationship between dimensions of
organizational culture as defined by Harper and the Knowledge Management
Technology Profile identified by the National Research Council (1994) and found
out that a number of cultural characteristics, including “sharing information freely,”
“working closely with others,” “team oriented work,” “trust,” and “supportive of
employees,” among other attributes, are highly positively correlated to knowledge
sharing, while other cultural characteristics, such as “being calm,” “compliance,”
and “stability” are highly negatively correlated to knowledge sharing (p. 113).
Similarly, Ruppel and Harrington (2001) examined the implementation of the
intranet as a system of knowledge management across a large variety of
organization and identified the organizational cultural characteristics that can
potentially facilitate knowledge sharing through intranet, including an emphasis on
trust, concerns for other people, and a stress on flexibility and innovation.
The cultural characteristics of an organization not only influences how
knowledge is shared internally but also how one organization is likely to share
knowledge with other organizations. Goffe and Jones (1996) described
organizational culture along two dimensions: sociability and solidarity and
identified four organizational culture frameworks. An organization that is highly
56

sociable and highly solitary has a communal culture, while an organization that is
low on sociability and low on solidarity has a fragmented culture. An
organizational culture is networked when there is high sociability and low
solidarity. Finally, an organizational culture that is high on solidarity and low on
sociability is a mercenary culture. Hoffman and Klepper (2000) examined
interorganizational knowledge sharing using the model of organizational culture
proposed by Goffe and Jones and found that the sharing of knowledge between
companies with different organizational cultures often leads to conflicts and
problems.
While most of the existing studies on organizational culture and knowledge
sharing have been focused on identifying and describing the kinds of organizational
culture conducive to knowledge sharing, Schein (2000) cautioned researchers that
they should not “make culture another item on the KM checklist” and suggested
that managers need to understand the specific goals they want to achieve before
they set out to create the culture of knowledge sharing (p. 8). Furthermore, most of
the current research on organizational culture and knowledge sharing failed to
explain how organizational culture affects the knowledge sharing practices of
organizational members. One exception is De Long and Fahey (2000), where the
authors identified four mechanisms by which organizational culture influences
knowledge sharing. First, organizational culture and subcultures define what
knowledge is worth sharing and what knowledge is not. Moreover, organizational
57

culture delineates the relationship between organizational knowledge and
individual knowledge, which in turn, influences organizational members’ decision
on knowledge sharing. Additionally, culture affects the ways in which knowledge
is utilized once it is shared. Finally, organizational culture also determines the
environment in which new knowledge is created, acknowledged, and shared. This
model is partially supported by empirical studies. For instance, several studies
showed that in an organizational culture in which members view knowledge as a
public good belonging to the whole organization rather than individual members,
knowledge sharing is more common (Ardichivili, Page, & Wentling, 2003; McLure
& Faraj, 2000).
The outcome of knowledge sharing
The traditional information-processing model of organizations that has dominated
the western management literature emphasizes the importance of internal rational
decision-making. However, this type of information processing does not always
lead to innovation. Instead, it is argued that organizations need to create knowledge
through repeated action and interaction with their environments. Senge (1990)’s
theory of a “learning organization” represents an early attempt at understanding the
creation of new knowledge through passive, adaptive, active, and generative
learning.
In the knowledge management literature, Nonaka’s model of knowledge
creation provides one of the first and most important theories on how knowledge is
58

created in organizations. According this model, knowledge creation follows an
upward spiral through the transformation of highly subjective, contextual, and
implicit knowledge that organizational members accumulate in their daily work
into objective, independent, and explicit knowledge that can be easily transferred,
which, when applied to the actual daily work of employees, generates more tacit
knowledge (Nonaka, 1994; Nonaka & Konno, 1998; Nonaka & Takeuchi, 1995;
Nonaka, Toyama, & Byosiere, 2001). Nonaka’s model emphasizes the creation of
knowledge in a process of problem solving through the organization’s interaction
with its environment. In this process, tacit knowledge is articulated and becomes
readily transferable. According to this theory, knowledge is created through four
processes: socialization, combination, externalization, and internationalization.
Socialization refers to the sharing of tacit knowledge between two people
through interpersonal interaction, such as observation, imitation, and practice
(Nonaka, 1994). The key to this mode of knowledge creation is the sharing of
experience in specific organizational contexts. One learns from the other the way of
thinking through socialization. One example of socialization is apprentice work.
The process of socialization is often time-consuming and costly (Kogut & Zander,
1992). Furthermore, the knowledge shared and created in this process can hardly be
utilized in an organization-wide approach due to its tacit nature (Kogut & Zander,
1992).
59

Combination refers to the merging of different bodies of explicit knowledge
to create new explicit knowledge or a new knowledge system (Nonaka & Takeuchi,
1995). It often involves a process of analysis and reconfiguration (Nonaka, 1994).
This mode of knowledge creation often happens in meetings, telephone
conversations, and computer-assisted decision making (Nonaka, 1994).
Externalization is the process by which tacit knowledge becomes articulated
and thus transformed into explicit knowledge (Nonaka, 1994). Metaphors and
analogies are important methods of externalization (Nonaka & Takeuchi, 1995).
Internalization refers to the transformation of explicit knowledge into tacit
knowledge through the process of “learning-by-doing” (Nonaka & Takeuchi, 1995,
p. 69). When one internalizes a piece of explicit knowledge, he or she has truly
learned the knowledge because it can be used.
The process of organizational knowledge creation involves the continuous
interaction between the four modes of knowledge creation discussed that follows
the pattern of a knowledge spiral, as shown in Figure 2-1 (Nonaka & Takeuchi,
1995).
60

Figure 2-1. Four modes of knowledge creation

Bierly and Chakrabarti (1996) studied the knowledge strategies of US
pharmaceutical companies between 1977 and 1991. They identified four knowledge
strategy groups: explorers, exploiters, loners, and innovators. Innovators are the
most aggressive learners who have high levels of both internal and external
learning, value both incremental and radical learning, and have the highest speed of
learning. On the other end, loners are the least effective in learning in that they
have the lowest speed of learning and the least number of knowledge links with
other organizations. Somewhere in between are exploiters and explorers. Exploiters
spend very little on R&D and thus do not create much new knowledge on their
own, but they are often very connected to other organizations and are good at
61

absorbing knowledge through external learning. At the same time, they focus on
incremental learning by making small improvements on competitors’ knowledge
and ideas. Explorers, on the other hand, are those companies that emphasize radical
learning and target the creation of blockbuster products. These companies have a
more or less consistent knowledge strategy over time. Innovators and explorers
tend to be more profitable than exploiters and loners.
While the interorganizational knowledge sharing literature shows a clear
information-based bias, the knowledge management literature in general
encompasses both the information-based and interaction-based approaches to
knowledge and knowledge management. Based on the current stage of knowledge
management studies, interorganizational knowledge sharing research needs to
develop in the following directions: (1) adopting a holistic understanding of
knowledge, (2) understanding interorganizational knowledge sharing not only as
transactions but also as interactions, and (3) situating interorganizational
knowledge sharing in the larger industrial, social, legal, and cultural contexts.
While interorganizational collaboration and knowledge sharing between
commercial companies have been extensively studied, one recent trend in
interorganizational knowledge sharing research has been the examination of
knowledge sharing and knowledge transfer between two specific organizational
populations: commercial companies and academic research institutions. The last
section of this chapter will be devoted to a review of the current research on
62

academic-industry knowledge sharing from technology transfer literature and R&D
economics literature.
Academic-industry knowledge sharing
Academic-industry knowledge sharing refers to the exchange of knowledge
between academic research institutions and commercial companies. The topic of
academic-industry knowledge sharing has received great attention in R&D
literature and industrial economics literature. Most studies that focus on academic-
industry knowledge sharing are based on the assumption that the sharing of
knowledge between academia and industry, or more specifically the transfer of
knowledge and intellectual property from academia to industry is beneficial to
industry development. Researchers have been especially interested in
understanding academic-industry knowledge sharing in knowledge-intensive
sectors such as the biotechnology sector (e.g., Jaffe, 1989; Liebeskind, Oliver,
Zucker, & Brewer, 1998) and the semi-conductor sector (e.g., Appleyard, 1996;
Lim, 2000).
Two major theoretical frameworks have dominated current research on
academic-industry knowledge sharing: transaction-cost economics theory and
social network theory. The transaction-cost economics approach understands
academic-industry knowledge sharing as an economic interaction based on formal
contracts. The rationale for this perspective is that acquiring knowledge from
external source is often cheaper than developing it in-house (e.g. 1990; Arora &
63

Gambardella, 1994). On the other hand, the social network approach perceives
academic-industry knowledge sharing as regulated by social norms and values and
argues that embedding themselves in a knowledge network is the fundamental
business strategy of many contemporary companies, especially knowledge-
intensive firms.
The existing academic-industry knowledge sharing literature, or more
accurately, the academic-industry technology transfer literature, assumes that
academic-industry knowledge sharing is a one-way process, i.e., from academia to
industry. It focuses on identifying the determinants of academic-industry
knowledge sharing. These determinants can be categorized into three groups: the
characteristics of companies, the features of academic institutions, and the
characteristics of ties.
Firm characteristics
Since a company is usually at the receiving end of academic-industry knowledge
sharing, its ability to identify, retrieve, receive, utilize, and integrate knowledge
first is an important predictor of the effectiveness of knowledge sharing. Three
major features of companies are identified: absorptive capacity, knowledge
distance, and connectedness.
Absorptive capacity of the company: Cohen and Levinthal (1989, 1990)
proposed that a company’s ability to utilize the discoveries from academic research
institutions is determined by its absorptive capacity. A company’s absorptive
64

capacity is decided by its prior knowledge, which, in turn, is a function of the R&D
investment of the company (Cohen & Levinthal, 1989; Cohen & Levinthal, 1990).
In order to increase its absorptive capacity, the company needs to invest more in
their in-house basic research and encourage their own scientists to publish
(Cockburn & Henderson, 1998).
Knowledge distance: Along a similar line, Liyanage and Barnard (2003) put
forward the concept of knowledge distance, which refers to the distance between a
firm’s prior knowledge and the new knowledge they are acquiring. According to
this theory, a firm’s prior knowledge in an area, as measured by the knowledge
distance, determines the firm’s ability to search for, evaluate, acquire, and utilize
new knowledge (Liyanage & Barnard, 2003).
Company’s connectedness with the scientific community: The concept of
connectedness is proposed by researchers to measure the extent of commercial
firms’ involvement with academic research community (Agrawal, 2001). Cockburn
and Henderson (1998) pointed out that in order to develop most companies’ ability
to absorb the fruits of academic research, increasing their in-house R&D
investment itself (as proposed by Cohen and Levinthal, 1990) is not sufficient.
Instead, companies need to engage themselves proactively with the academic
community, or in other words, to increase their “connectedness” to the academic
community. For instance, companies should encourage their researchers to
collaborate with universities’ scientists (Cockburn & Henderson, 1998). One
65

indicator of the connectedness between firms and universities is co-authorship
(Cockburn & Henderson, 1998). Zucker, Darby, and Armstrong (2002) found that
the number of publications co-authored by scientists from academia and industry is
positively correlated to the number of successful knowledge transfers.
Many later research projects build on the concepts of absorptive capacity
and connectedness in order to explain why some companies are better at sharing
knowledge with academic research institutions. They have identified several
strategies to enhance companies’ ability to make use of external knowledge by
either enhancing their absorptive capacity or by increasing their connectedness with
the public research community, such as hiring gradates and faculties from
universities, using an internal incentive system to encourage publications and
patents, and promoting collaboration between companies and universities by
creating collaborative projects and funding university research (Audretsch, 2000).
Characteristics of academic institutions
Since the passage of Bayh-Dole Act of 1980, these has been a huge increase in the
number of academic patents and licenses in the United States (Henderson, Jaffe, &
Trajtenberg, 1998). Characteristics of academic institutions, including their IP
policies, the characteristics of their technology transfer office (TTO), and the
characteristics of their faculties affect the frequency, the process and the outcome
of academic-industry knowledge sharing.
66

Academic IP policies: First of all, a flexible intellectual property policy
adopted by academic institutions often facilitates successful knowledge sharing, in
the forms of patenting licensing or spin-offs (Bower, 1993; Santoro &
Gopalakrishnan, 2001). For instance, Di Gregorio and Shane (2000) found a
positive relationship between the flexibility of university’s licensing policy and the
number of patents licensed. One example of flexibility is that while traditionally
universities require cash payment in exchange for the rights to commercialize a
university’s intellectual property, now some universities are accepting equity as
royalty payments. In fact, the more experienced and more successful a university is
in technology licensing, the more likely it is going to use equity instead of cash
payments (Feldman, Feller, Bercovitz, & Burton, 2000).
Organizational structure of TTO: As an increasing number of academic
institutions have established Technology Transfer Offices (TTOs) to facilitate the
knowledge sharing and technology transfer between academia and industry.
Researchers have begun to investigate the actual role of TTOs play in this process.
Bercovitz et al. (2001) found that different organizational structures of TTOs (U-
form, M-form, H-form, and Matrix-from
5
) lead to different measures on three
attributes of TTOs: information processing capacity, coordination capability, and

5
Chandler identified two organizational forms: U-Form and M-Form. A U-Form (or unitary form)
organization is centralized and departmentalized. The responsibility of decision-making and
coordination lies with a small group of top executives, while a M-Form (or multidivisional form)
organization is one divided into semi-autonomous departments (Chandler, 1962). H-Form (or
holding company) is also divided into different departments, but has a weak central office
(Williamson, 1975, 1985). Finally, the MX-Form (or matrix form) organization has a structure that
combines elements from two or three of the organizational forms discussed (Bercovitz et al., 2001).
67

incentive alignment, which in turn affect the TTOs’ ability to facilitate technology
transfer. Based on case studies of TTOs of three universities (Duke, Johns Hopkins,
and Pennsylvania State University), Bercovitz et al. (2001) argued that a TTO with
a matrix structure has the highest performance level, while a TTO with a H-form
structure has the lowest level of performance with the a M-form TTO offering a
medium performance.
Characteristics of professor-inventors: The attitudes and characteristics of
academic scientists are another group of variables influencing academic-industry
knowledge sharing. Owen-Smith and Powell (2001) studied the attitudes of
faculties at a major private university and a major state university with comparable
research capacities yet very different rates of patenting, and they discovered that
the perceived ease in dealing with a TTO is a major determinant of faculties’
decision to patent, in addition to perceived benefits of patent protection. They also
discovered that university faculties in different disciplines often patent their
discoveries for different reasons. Life scientists patent their discoveries more often
because they want to get monetary gains from licensing. Physical scientists, on the
other hand, patent their inventions so that they can have the freedom to publish
(Owen-Smith & Powell, 2001b).
The passage of the Bayh-Dole Act of 1990 gave academic scientists great
incentives to patent and license their discoveries. Since then there have been huge
increases in the number of academic patents. However, this does not mean
68

universities are producing higher quality research discoveries or universities are
doing more application-oriented research. Henderson, Jaffe, & Trajtenberg (1998)
found out that the quality of current university patents are not higher than general
patents or university patents in the past by comparing them on two dimensions:
importance and generality. Thursby and Thursby (2000) found that the increase in
the number of academic patents licensed to companies is not due to the shift of
university research from basic research to applied research but due to faculty
members’ increased propensity to license and to the industrial trend of outsourcing
R&D to universities.
Geographic proximity
Geographic proximity has long been considered an important predictor of
successful academic-industry knowledge sharing. Research shows that long-
distance and local knowledge sharing follows different logics (Phene & Tallman,
2002). Audretsch and Stephan (1996) provided empirical support of the
“geographic economics of biotech industry,” i.e., biotech firms and research
universities tend to co-exist in the same geographic location. They attribute this
phenomenon to the existence of university-based scientists. Zucker et al. (1998)
discovered that new biotech companies are likely to locate in areas where star
scientists published papers in gene sequencing. Owen-Smith and Powell (2004)
studied the knowledge network in Boston area and again identified geographic
propinquity as a factor that determines the flow of information. Two explanations
69

have been offered to account for this phenomenon: knowledge spillover and market
exchange.
Most researchers attribute the growth of regional technology clusters to
knowledge spillover. Knowledge spillover refers to the movement of knowledge
from academic institutions to industrial companies through informal
communication and interaction rather than formal channels. Studying patent
licensing at MIT, Agrawal (2000) discovered that geographic distance has a
positive effect on the success of patent licensing mediated the frequency of direct
interaction. In other words, geographic proximity enables more direct
communication and interaction between university and commercial companies,
which, in turn, often results in higher possibility of successful licensing.
Zucker, Darby, and Armstrong (1998) proposed market exchange as another
mechanism underlying the emergence and growth of localized high tech clusters.
Through case studies and interviews, they discovered that the traditional
assumption of knowledge spillover does not explain how knowledge is transfer
from academia to industry. On the contrary, commercial companies, governmental
agencies, as well as star scientists are all engaged in collaboration based on market
exchange relations. In summary, both market exchange and knowledge spillover
contribute to the geographic nature of knowledge sharing as well as the
development of regional high tech clusters.
70

Professional culture
Professional culture is another variable that influences the process and effectiveness
of academic-industry knowledge sharing. The creation and sharing of knowledge
follows different logic in academia and the business world (Gittelman & Kogut,
2003). First of all, the logic of scientific discovery is different from that of the
development of new technologies (Gittelman & Kogut, 2003). Analysis of the
citation patterns in scientific journals and patents shows that scientific discoveries
in the academic community follow an “evolutionary epistemology.” In this
manner, favored ideas are preserved and developed, and disfavored ideas are
deserted, while technological development in the industrial world, on the other
hand, follows the logic of marketability and profitability(Campbell, 1974;
Gittelman & Kogut, 2003).
Furthermore, knowledge sharing is directed by different mechanisms in
these two communities. The institution of science calls for the open sharing of
knowledge for public scrutiny (Merton, 1973; Siegel, Thursby, Thursby, &
Ziedonis, 2001a). On the other hand, the spread of technology in the industrial
world follows a different logic: exclusivity leads to profit (Siegel et al., 2001a).
Companies hoard valuable information, technology, and practice from their
competitors to ensure their own competitive advantage (Han, 2004).
71

Trust
Related to the very different cultures of academia and industry and their conflicting
interests, trust becomes another factor that affects successful knowledge sharing
between universities and industries. Trust is the “willingness of a party to be
vulnerable to the actions of another party based on the expectation that the other
will perform a particular action important to the trustor, irrespective of the ability to
monitor or control the other party” (Mayer, Davis, & Schoorman, 1995, p. 712).
Adler (2001) pointed out that trust is an especially important coordinating
mechanism in today’s capitalist organizations, in which the generation and
diffusion of knowledge has become the key to success. He identified three sources
of trust: familiarity developed through previous contact, calculation based on the
pursuit of interest, and norms. He argued that trust is generated by direct
interpersonal contact, by reputation, and by people’s understanding of the
institutional contexts, and identifies five bases of trust, including consistency,
competence, benevolence, honesty and openness(Adler, 2001).
Santoro and Gopalakrishnan (2001) found that trust is one of the most
important determinants of successful university-industry knowledge transfer by
conducting surveys with 21 research centers in public and private universities and
421 industrial firms that collaborate with these centers. They offered three reasons
explaining the positive relationship between trust and university-industry
knowledge transfer: confidence in the partner’s ability, cooperative instead of
72

opportunistic behavior, and social control mechanism. Adler (2001) introduced the
concept of “reflective trust,” which is different from both calculative and traditional
trust. Reflective trust is based not on either traditional authority or pure economic
calculation but on norms and on the ideal speech situation. It is based on the values
of science, including universalism, communism, disinterestedness, organized
skepticism, etc. (Adler, 2001). This is the trust that guides the knowledge economy
in general.
Summary
This chapter provides a brief review of three bodies of literature:
interorganizational relationship, knowledge management, and academic industry
knowledge sharing. Academic-industry knowledge sharing is a special form of
interorganizational knowledge sharing and an essential part of the knowledge
economy. The next two chapters will be devoted to an interpretive study of
academic-industry knowledge sharing as practiced on both the individual level and
the organizational level.

73

Chapter 3. Academic-industry knowledge sharing: an interview study
As discussed in Chapter 1, academic-industry knowledge sharing is a very special
type of interorganizational knowledge sharing. Due the knowledge-intensive nature
of the biotech industry and the high dependence of biotech companies on the
academic community, the sharing of knowledge between academic institutions and
biotech companies have been studied in several bodies of literature, including
interorganizational relations literature, R&D literature, and technology transfer
literature.
This chapter will introduce the motivation, research questions, and
methodology of an interview study of academic-industry knowledge sharing in the
biotech sector.
Theoretical development and research questions
Researchers from a variety of academic disciplines have been studying
interorganizational knowledge sharing and academic-industry knowledge sharing,
including interorganizational relations, information systems, R&D and technology
transfer. For instance, interorganizational relations literature looks at how
companies form different forms of alliances, such as strategic alliances and joint
ventures, to share knowledge and other types of resources. Information systems
literature examines how the implementation of information systems, such as
intranet and database, facilitates or hinders knowledge sharing. Technology transfer
74

literature examines how the intellectual properties of academic institutions are
transferred to and commercialized by companies.
Most of the existing research on interorganizational knowledge sharing
takes an information-based view of knowledge and a deterministic approach to
knowledge sharing. It tries to explain why, when, and how knowledge is shared
through an examination of the characteristics of companies, universities, faculties
and knowledge (see Agrawal, 2001 for a review of this literature).
This study takes the perspective that academic-industry knowledge sharing,
as well as the interorganizational knowledge sharing in general, is inherently a
communication phenomenon and should be understood in reference to the
following four aspects:
First, most of the existing studies on interorganizational knowledge sharing
and academic-industry knowledge sharing examine them as the interactions
between organizations. In other words, organizations are considered as the main
players and examined as the unit of analysis. There is a lack of effort in trying to
understand interorganizational knowledge sharing and academic-industry
knowledge sharing on the individual level: how people involved in the process,
business executives, company scientists, university professors, and university
technology transfer specialists perceive, make sense of, and behave in sharing
knowledge. So it is important to look at academic-industry knowledge sharing both
on the interorganizational level and on the interpersonal level.
75

Secondly, culture has been identified as one important factor in knowledge
sharing. Past research examined how organizational culture or national cultures
affect both organizational and interorganizational knowledge sharing. In the case of
academic-industry knowledge sharing in the biotech industry, special attention
needs to be paid to understanding how professional culture, i.e. the culture of
academia and culture of industry affect the process and outcome of knowledge
sharing.
Thirdly, while existing business and management literature mainly looks at
how companies and universities share knowledge through formal contractual
channels such as strategic alliances, patent licensing, a communication study of
interorganizational knowledge sharing calls for an understanding of the
social/informal communication involved and an exploration of the interaction
between formal and informal knowledge sharing. These directions of inquiry lead
to the following research questions.
Research Questions
To understand academic-industry knowledge sharing in the biotech sector, it is first
necessary to understand the motivations of knowledge sharing. This leads to the
first research question. Existing research on academic-industry knowledge sharing
has been mainly focused on organizational level motivation, i.e. why organizations
want to share knowledge with other organizations, without exploring individual
level motivation, i.e. why individual organizational members want to share
76

knowledge with others in other organizations. However, academic-industry
knowledge sharing not only involves interorganizational alliances and contracts but
individual level communication and interaction. In an effort to understand the
motivations of academic-industry knowledge sharing, I will examine the
motivations on both individual and organizational levels and look for the
similarities and differences between them. This leads to the following research
questions:
RQ1a: why do academic scientists want to engage in academic-industry
knowledge sharing?
RQ1b: why do company scientists want to engage in academic-industry
knowledge sharing?
RQ1 c: why do academic institutions engage in academic-industry
knowledge sharing?
RQ1d: why do commercial biotech companies engage in academic-industry
knowledge sharing?
Most of the existing studies on academic-industry knowledge sharing adopt a very
narrow conceptualization of knowledge and knowledge sharing, examining the
sharing of codified knowledge through a limited number of channels, including
patent licensing and publication citations. In an effort to understand academic-
industry knowledge sharing as a holistic and communicative process, it is essential
to take a broader approach, to look beyond licensing and citation patterns and to
77

ask the question of what kinds of knowledge are shared between academia and
industry and how they are shared? This leads to the next research question:
RQ2: How is knowledge shared between academia and industry?
To answer this question involves three related questions: nature of the knowledge
shared, intellectual properties used, and channels adopted in the process of
academic-industry knowledge sharing.
RO2a: What knowledge do academia and industry share?
RQ2b: How is the knowledge shared between academia and industry
protected as intellectual properties?
RQ2c: What are the channels of academic-industry knowledge sharing?
Culture is another factor affecting the sharing of knowledge. While a large number
of existing studies aim at explaining how national culture and organizational
culture influence organizational and interorganizational knowledge sharing,
relatively little attention has been paid to the examination of how professional
culture influences the knowledge sharing between individual people and
organizations from the academic community and industry. Certain studies briefly
pointed out that the difference in professional cultures affect knowledge sharing,
especially that between academic organizations and commercial companies (e.g.
Agrawal, 2001). What is missing is a further exploration of how the culture of
academia and the culture of industry come together or go against each other when
78

people from the ivory tower and the office tower come to work together, building
on each other’s knowledge base. This leads to the next research questions:
RQ3: How do the professional cultures of academia and industry affect the
knowledge sharing between the two communities?
Finally, in order to have a comprehensive understanding of academic-industry
knowledge sharing, it is important to go beyond the study of formal knowledge
sharing channels that have been extensively studied in literature and examine the
informal dimension of academic-industry knowledge sharing. This leads to the final
research question:
RQ4: What is the role of informal knowledge sharing in academic-industry
knowledge sharing in biotech industry?
Methodology for the interview study
Access and sampling
I used a combination of interviews and ethnographic observation to make sense of
the academic-industry knowledge sharing in the biotech sector. To that end, I
worked part-time for 13 months from May 2005 to May 2006 at a biotech business
research center affiliated to the business school of a major private research
university. The center was built with the purpose of bridging the gap between
academia and biotech industry and facilitating the commercialization of university
discoveries. During this period, I fully participated in the daily operation of the
79

center. While conducting industry-specific research, I was also responsible for
organizing conferences and seminars that brought together academic research
institutions and biotech companies to discuss issues such as technology transfer,
collaboration, and commercialization. At the same time, I attended several local or
regional conferences on technology transfer and commercialization held in
Southern California and had numerous informal conversations with biotech
professionals, including university technology transfer specialists, faculties,
entrepreneurial-minded PhD students and postdocs, business executives, company
scientists, and leaders of the regional biotech trade organizations. This experience
gave me very intimate understanding of the process of academic-industry
knowledge sharing and technology commercialization. After working at the center
for 9 months, I started to conducted interviews to gather more systematic
information from those professionals who engage intensively in academic-industry
knowledge sharing and collaboration.
Non-probability sampling was used to identify the potential interviewees for
this study. First, I selected interviewees from industry that I knew on a personal
basis, ensuring a representative number from small Dedicated Biotechnology Firms
(DBFs), biotechnology consulting companies, and large established
biotechnology/pharmaceutical companies. At the same time, I got in touch with
technology transfer specialists from major universities and medical research
institutions in California through technology transfer conferences. Finally, I
80

recruited faculties and research scientists conducting research in fields related to
life sciences in major research universities and institutes in Southern California
through personal contacts. In recruiting interviewees from academic institutes (both
universities and research institutions), I paid special attention to include
interviewees from public universities, private universities, and non-teaching
research institutions to get a comprehensive and balanced point of view.
Network sampling or snowball sampling was used to identify potential
interviewees beyond my immediate contacts and connections. During interviews,
my interviewees provided me not only with their experiences and thoughts about
academic-industry knowledge sharing but also with names and contact information
of their friends and contacts that they believed were able to give further insights on
the subject matter, whom were hence contacted for the possibility of an interview.
In the end, the interviewees consist mainly of four panels: (1) business
executives from biotechnology or pharmaceutical companies, (2) scientists working
in the R&D division of biotechnology companies, (3) faculties from universities or
academic research institutions, and (4) technology transfer specialists from
Technology Transfer Offices or Intellectual Property Offices of major research
universities and institutions.
Semi-structured interview and interview protocol
In this study, I used semi-structured interview method because it has the focus of
structured interview and, at the same time, allows for the freedom of unstructured
81

ethnography (Fontana & Frey, 2000). Since the goal of the interviews is
exploratory rather than hypothesis-testing, I am more interested in the generation
and rectification of theoretical concepts rather than testing and confirming
theoretical concepts (Wengraf, 2001, p. 56). To that end, two interview protocols
were developed with both open-ended and close-ended questions regarding
people’s personal experience and opinions about academic-industry knowledge
sharing in the biotech sector. One protocol was designed with interviewees from
industry in mind (See Appendix A) and the other was designed for interviewees
from academia (See Appendix B). The interview protocols ensured that all the
interviews would cover several broader topics, including the goals, the channels,
the effectiveness of academic industry knowledge sharing, and several issues of
interests in the process of knowledge sharing, including: intellectual property
issues, cultural issues, personal communication issues, and the use of
communication technology.
Furthermore, I adapted the interview protocols to the specific characteristics
of each interviewee based on my previous research and on my understanding of the
specific professional experience and qualifications of the interviewee. For instance,
for those interviewees with experience in both industry and academia, I asked them
to compare their experience with the two communities regarding academic-industry
knowledge sharing. Whenever an interviewee made a new and interesting remark, I
probed him/her to elaborate and to give further explanations and clarifications.
82

Interviewees were asked to answer all the questions based specifically on their
experience with life science research and biotechnology industry.
Story-telling technique: As the central topic of the interviews is knowledge
and knowledge sharing, which is often elusive and hard to capture, I utilized the
story-telling technique to help interviewees articulate their understanding of
knowledge and knowledge sharing. Story-telling technique is useful in the study of
knowledge sharing because it offers a peek into the tacit dimension of knowledge
sharing, as Martin (1982) pointed out, “story is a form of implicit communication
within organizations” (p. 261). Through storytelling, people will say more than
what they would normally tell, because "stories permit researchers to examine
perceptions that are often filtered, denied, or not in the subjects' consciousness
during traditional interview" (Hansen & Kahnweiler, 1993, p. 1394).
In the semi-structured interviews, I ask interviews to tell two stories
regarding their most positive and more negative experience of academic-industry
knowledge sharing. Ambrosini and Bowman (2001) recommended the use of this
technique based on the critical incident technique developed by Flanagan (1954)
6
.
Metaphors: Another way of looking into the silent realm of knowledge
sharing, especially the sharing of tacit knowledge, is the study of metaphors.

6
Critical incident technique consists a set of procedures to collect data on organizational behavior
for the purpose of theory building and problem solving. One of the techniques discussed by
Flanagan (1954) in his review of the critical incident techniques is to ask respondents to tell critical
stories related to a particular organizational outcome, and ask them to make sense of the incidents.
For instance, researchers can ask interviewees to discuss a critical incident with positive or negative
result.
83

Metaphors are useful for two reasons. First, they “can communicate meaning when
no explicit language is available, especially in regard to complex ambiguous
experience” (Srivastava & Barrett, 1988, p. 60). Secondly, metaphor is a cognitive
priming mechanism that can be used to predict people’s attitude and behavior
(Gibson & Zellmer-Bruhn, 2001). In my interviews, I paid special attention to the
metaphors used by interviewees and what these metaphors tell about academic-
industry knowledge sharing. Whenever an interviewee used an interesting
metaphor, I encouraged him/her to elaborate on it.
Data collection
After obtaining university IRB approval, I conducted all the interviews between
January and May in 2006. A total of twenty-four semi-structured interviews were
conducted for this study. I traveled in California to meet my interviewees unless
they explicitly chose phone interview over face-to-face interview, as both existing
literature and my personal experience suggested that face-to-face interviews are
often more effective in understanding ambiguous and complicated experience and
in eliciting people’s thoughts and opinions on non-sensitive subjects (Wengraf,
2001). Of the 24 interviews conducted, 9 were phone interviews, and 15 were
conducted face-to-face. Most of my interviewees are located in the three major
biotech clusters in California: San Diego (3), San Francisco/Bay area (1), and Los
Angeles (17). Table III-1 gives a belief summary of the demographic information
of interviewees.
84

Demographic profiles of interviewees
Gender Number Percentage
Male 21 87.5%
Female 3 12.5%

Job function
Academic faculty 4 16.7%
Academic technology transfer specialists 7 29.1%
Biotech company executives 8 33.3%
Biotech company scientists 5 20.8%

Education Level
Bachelor’s degree 1 4.2%
Master’s degree (including MBA) 4 16.7%
Doctorate degree (including PhD, MD, and
JD)
19 79.2%

Types of academic institutes
Public university 4 36.7%
Private university 5 45.5%
Private research institutes 2 18.2%
Table 3-1. Demographic information of interviewees
One interesting characteristic of my interviewees is that most of them have
extensive experience in both academia and industry and in both the business and
the research sides. For instance, faculties often have extensive experience with
industry. One professor used to work in several very famous biotech companies for
almost 10 years before he returned to academia. Another professor was the director
of a major industry-oriented research center for 15 years. A third professor was also
the CEO of a newly founded biotech company. The technology transfer officers
interviewed often have extensive research experiences: two of them used to be
assistant professors at research universities and another one used to work as a
postdoc in a leading research university. Almost all the business executives had
85

doctoral degrees in life science disciplines and had extensive research or teaching
experience at universities. Finally, all the company scientists interviewed were
postdoc researchers at universities for several years before they joined industry.
One of scientist, at the time of interview, has just accepted an offer to become an
assistant professor at a medical school. This crossover of experience is, in a way, a
demonstration of the close ties between academia and biotech industry.
All the interviewees were aware that I was a graduate student conducting
dissertation research, and gave explicit consent to be part of this project. Every
interviewee was informed of his/her rights as a human subject and signed the
Informed Consent Form before the interview started. Interviews generally lasted
between 40 minutes to an hour, with most taking around 45-50 minutes.
Among the 24 interviews conducted, 21 interviews were recorded (with
interviewee’s written consent) with a digital voice recorder and uploaded to a
password-protected personal computer for transcription and further analysis. Three
interviewees declined to be recorded and I took notes of the interview by hands
instead. However, one recorded interview was incomprehensible due to the
technical failure of the recorder. In the end, a total of 20 interviews were
transcribed for content analysis.
Of all the interviews, 21 were conducted in English. For the rest 3
interviews, even though the interview questions were read in English, the
interviewees chose to use a combination of English and Chinese Mandarin in
86

answering the questions. Their answers were later transcribed and translated into
English by me for data analysis.
Data analysis
Both thematic analysis and content analysis were conducted to make sense of the
data collected through interviews. For thematic analysis, themes and sub-themes
were identified and discussed (Aronson, 1994). Special attention was paid to
comparing and contrasting the responses from different panels of interviewees, for
instance, how academic scientists and industry scientists engage in and make sense
of academic-industry knowledge sharing similarly or differently.
I used Atlas-ti, a hermeneutic qualitative data analysis software to assist the
management of the interview data and the content analysis. To that end, each
transcript was organized into paragraphs, with each paragraph represents the whole
or a portion of interviewee’s response to one question. The transcripts were
transformed into rich text format recognized by the Atlas-ti program. I coded each
paragraph in transcripts to identify common themes and subthemes. Some codes
were created based on previous literature on academic-industry knowledge sharing,
while others emerged from the texts. Usually a paragraph was assigned one, or in
most cases, multiple codes. The codes were constantly modified: new codes were
added, repetitive codes were merged, and unnecessary codes were deleted. I coded
all texts three times to ensure a consistent and comprehensive coding. During the
process, comments and memos were added to explain codes and the relationship
87

among them. In the end, a total of 158 codes were identified and used for data
analysis. Appendix C contains a complete list of all codes and frequencies with
which they occur. Appendix D includes sample codes and representative quotations
that those codes are associated with.
Atlas-ti program enables the building of cognitive maps that explicate the
qualitative relationship among codes. Cognitive maps are representation of people's
knowledge with regard to a particular issue (Ambrosini & Bowman, 2001). One
type of cognitive map is causal map, which is a graphic representation with nodes
and ties that link them. The nodes are often constructs important to a particular
issue (in this case, codes), and the links represent the causal relationship among
nodes. The relationships include: “is associated with,” “leads to,” “prevents,”
“facilitates,” “in part of” etc. I built and revised these relationships while coding the
transcripts as they emerged in my interviewees’ narratives. In the end, related codes
were put into different cognitive networks that addressed different aspects of
academic-industry knowledge sharing. Four conceptual networks were created. For
instance, the network of “Informal Knowledge Sharing” spelled out the
determinants, characteristics, and consequences of informal knowledge sharing
between academia and industry (See Appendix E for a list of networks).
In the next chatter, I will present and discuss the findings of the interview
study.

88

Chapter 4. Research Findings
This chapter presents the findings of the interview study. The chapter will be
organized according to the research questions discussed in Chapter 3.
RQ1: Why do academic research institutes and commercial biotech companies
share knowledge?
The study of academic-industry knowledge sharing naturally starts from an
examination of the goals or motivations of different players involved. While
existing literature mainly focuses on the comparing the goals of academic
institutions and commercial companies, this study proposes that it is important to
understand the motivations of academic-industry knowledge sharing on both the
organizational level and the individual/personal level. For instance, as the special
nature of the work of academic scientists gives them much independence and
freedom, they sometimes can be quite detached from the universities and research
institutes they work for and have their own individual goals and agendas in sharing
knowledge with industry. In the following section, I will discuss the goals of
biotech companies, academic institutes, and academic scientists in academic-
industry knowledge sharing. Figure 4.1 presents a conceptual network of the goals
of academic-industry knowledge sharing.
89

Figure 4-1. Goals of academic-industry knowledge sharing
Biotech companies
The primary goal of biotech companies in engaging in academic-industry
knowledge sharing is to identify new technologies that can be translated into
products. Secondly, biotech companies engage in knowledge sharing with
academic research community so that they can utilize the knowledge and expertise
of academic researchers to solve practical problems in their own R&D and
manufacturing process. Even though this kind of knowledge sharing is more
focused on the day-to-day R&D work of biotech companies and more detail
oriented, many university faculties and company scientists identify technical
problem-solving as the most important reason for which biotech companies engage
90

in collaboration with academic institutes and labs and one of the most frequent
forms of academic-industry collaboration. For instance, a senior scientist at a major
biotech companies said,
From time to time, there could be knowledge sharing or collaboration done
for more problem solving [purpose]. It may not be with a product in mind,
but with maybe certain technical challenge.

Next, acquiring knowledge from academia instead of developing it in-house
is often associated with the goal of lowering the cost of research infrastructure
investment and getting the results needed faster, which, in turn, lead to higher and
sooner profit for the company. Furthermore, biotech companies collaborate with
academic institutions to create a positive social image related to high quality
science, an orientation towards public good, and honesty. Very often academic
institutions and academic scientists carry with them not only cutting edge science
and technology in life sciences, but also an enormous amount of social capital. By
partnering with academic institutions, biotech companies can increase their
credibility and appeal to their investors and the general public. One technology
transfer officer at a major research university pointed this out by saying
Sometimes some of those [academic] researchers become very well known
and to have those people on your board or your scientific board makes you
more credible with the investors.

Finally, collaborating with academic institutions may help companies identify their
future employees. This is especially important for biotech companies who need
employees who are highly specialized scientists and are often hard to find.
91

In summary, biotech companies engage in knowledge sharing mainly to
acquire knowledge, be it basic knowledge, technology, technical expertise, or
people who possess the knowledge or expertise. Because of biotech companies’
dependence on academia, some people consider academic-industry knowledge
sharing as a one-way process, i.e. the transfer of knowledge from academia to
industry. The Director of Licensing from a private medical research institutes said,
[Companies] acquire knowledge. They don't want to share. They come and
they use it for their ends. Sharing implies a two way street. Sometimes they
pretty much come with a pre-defined agenda about what they are going to
get out of their investment.

While existing research on academic-industry knowledge sharing often compares
academia and industry in terms of their motivations, behaviors, and outcomes,
assuming that academia can be considered as an integrated entity, my interviews
shed lights on the subtle internal differentiations within the academic community
and discovered that academic institutions, whose interests are represented by
technology transfer officers and academic scientists often have different goals in
collaborating with industry. Sometimes, these differences cause internal conflicts
that prevent successful academic-industry knowledge sharing.
Academic institutions
Academic institutions such as universities, public and private research institutes,
and research hospitals, engage in academic-industry knowledge sharing for a
number of reasons. Most fundamentally, they want to share knowledge with
92

industry is to fulfill university’s mission to benefit society and the general public.
Both interviewees from academia and industry agreed upon this point.
A more contested goal of academic institutions’ engagement with industry
is to gain financial reward. All interviewees from industry and some interviewees
from academia identified monetary reward as one of the major incentives for
academic institutions to share knowledge with industry. However, interviewees
from academia, especially technology transfer specialists, are sometimes reluctant
to admit this, claiming that most of the patent licensing deals they have do not pay
off. The Assistant Chancellor of Intellectual Property at a major public research
university in California said,
Very few universities actually make a lot of money [from the licensing of
intellectual property]. If you look at how much money they spend on
research, licensing of intellectual property will generate no more than lower
single digit percentage of the expenditure. [As a result], … I don’t think that
(monetary gains) is really a goal. It is kind of like a fringe benefit. It isn’t a
goal at all.

Even though university’s failure to cash out does not validate the claim that
academia is not looking for financial profit, the ambivalent attitude of technology
transfer officers is a good testimony of university’s uneasiness in negotiating its
traditional identity as a non-profit institution for public good and its newly formed
identity as owners of intellectual properties and resources with enormous market
value.
93

In addition, academic institutions share knowledge with industry and
society at large in order to build reputation and social capital. As Owen and Powell
(2001a) pointed out, today the “institutional prestige for research universities is
increasingly defined in terms of both academic and commercial science” (p. 99).
This goal is becoming increasingly important because today’s public is less likely
to frown upon universities’ partnership with industry as “going to the dark side”
and is becoming more positive towards such relationships. In fact, academic
institutions are often under increasing peer pressure to join this game. One senior
technology transfer officer from a major public university commented on the
building of social capital by saying,
I don’t think it is the primary goal, but every university has to admit that
universities are often times ranked by citation and things like that. It is same
with companies. The more knowledge you share, the more reputable your
university may become.

Finally, academic institutions, as grantees of federal or state research grants and
investments, are often required to share knowledge with industry as a condition of
their funding.
7

7
For instance, the intellectual property policy of the newly established California Institute of
Regenerative Medicine (CIRM) requires that the grantee organization must disclose its inventions
produced by CIRM-funded research to both CIRM and the general public. For more information on
CIRM intellectual property policy, see http://www.cirm.ca.gov/policies/pdf/IPPNPO.pdf.
94

Academic scientists
Even though they are employed by universities or research institutes, academic
scientists have considerable autonomy and freedom, and consequently their own
goals and agendas in sharing knowledge with industry. The most important
incentive for academic scientists to collaborate with industry and to share their
knowledge and expertise with industry is to acquire extra research funding. While
academic scientists in medicine and life sciences get most of their funding from
federal funding agencies such as NIH, they often need extra funding to make ends
meet at a time when federal funding is no longer adequate and the competition for
federal grants becomes white-hot. All academic scientists interviewed were very
frank about it. A professor at a private research university said,
Exactly, because it is not so easy to get funding from the government. In
fact, I have some colleagues who have large research groups and they worry
that all the money is going to Afghanistan, and there is no money for
research.

The second goal of academic scientists in their engagement with industry is
to get involved in world-class research conducted in biotech companies, especially
large established biotech/pharmaceutical companies. While in the past people tend
to think of academic-industry knowledge sharing as the one-way transfer of
knowledge from academia to industry, today the sharing of knowledge between
academia and biotech companies is by no means a one-way street. Very often
biotech companies are world leaders in their own areas of research and as a result,
95

by collaborating with such companies, academic scientists can get access to
cutting-edge research in those areas that benefits their own research. A senior
scientist at a leading biotech company said,
I think nowadays, a lot of academic scientists collaborate with industry [for]
two major reasons. First is that a lot of leading sciences nowadays are done
at industry anyway. So for very successful companies, especially science-
based companies, they do very good science. For scientists in academia,
they feel that collaborating with company at the cutting edge of science,
they can benefit [from] the knowledge as well.

Furthermore, by collaborating with industry, academic scientists can gain access to
proprietary information, data, materials, and equipments from industry for their
own research. These proprietary materials and data are often very expensive and
not available from other sources. One assistant professor at a public research
university commented,
So one thing about biology is that good data are hard to come by and
expensive to produce. So in the few collaborations that we have with the
industry, the model of collaboration is that we get some data, in exchange
for our software and ideas. I think that is one of the primary reasons.

Next, some academic scientists engage in knowledge sharing with
commercial companies and commercialize their scientific discoveries for personal
financial and egoistic reasons. Working with industry is financially very profitable.
At the same time, some academic scientists simply want to see their discoveries
lead to real products. One senior technology transfer officer at a private research
university told the investigator,
96

They [faculties] like to see their own research come to some kind of actual
tangible benefit beyond just the knowledge itself. Not everybody has that
desire, but many of them do. And I think that they look to companies to be
able to accomplish some sort of public good.

Finally, one common goal shared by academic institutions and their
scientists is to build a good working relationship with commercial companies,
which will help them place their graduates, PhDs, and postdocs.
An analysis of the goals of academic institutes and academic scientists shows that
while the two share certain goals, there is also considerably discrepancy and even
conflicts in their motivations in collaborating and sharing knowledge with industry.
The internal differentiation within academia sometimes can cause problems, which
will be discussed in later part of the chapter.
RQ2: How is knowledge shared between academia and industry?
After examining why different players want to engage in academic-industry
knowledge sharing, I will look at how knowledge is shared between the two
communities. The understanding of academic-industry knowledge sharing consists
of the answering of the following three questions: (1) what types of knowledge are
shared between academia and industry? (2) how is knowledge protected as
intellectual property? and (3) what are the channels adopted by academicians and
industry in sharing knowledge with each other? These questions will be answered
next.
97

RO2a: What knowledge do academia and industry share?
While the technology transfer literature reduces knowledge shared between
academia and industry to that codified in patents and publication, this simplistic
view of knowledge was clearly rejected by the interviewees. Instead, many
interviewees emphasized the complex nature of the knowledge shared between
academia and industry and discussed several major categories of knowledge shared
between academia and industry, including new scientific discoveries, technologies,
understanding of scientific backgrounds, technical know-hows, research equipment
and materials, understanding of industry and market, and knowledge about “who is
doing what” in the disciple.
New scientific discoveries represent a significant part of the knowledge
shared between academia and industry. When academic scientists make new
discoveries, they share them through publications, academic/industry conferences,
and other channels of knowledge diffusion. New scientific discoveries are often
more basic and far away from market. They are often protected by copyright.
Technologies are another type of knowledge shared. Examples of technologies
shared include new materials invented that can be used for therapeutic purposes,
new molecules created to make new drugs, and new devices designed for the
diagnosis of diseases. Compared to scientific discoveries, these technologies are
often more practical and closer to market. They are often protected by patents and
shared through patent licensing agreements.
98

Understanding of scientific backgrounds is another type of knowledge
pursued in academic-industry knowledge sharing, especially by commercial biotech
companies. Biotech companies often hire academic scientists as their consultants to
take advantage of their understanding of the scientific background of a discipline
and get advices and feedbacks on companies’ new research and development
projects.
Technical know-hows are often less articulated and is often shared when
scientists from a university and a company work together on a project. Examples of
technical know-hows include how to make an equipment work, how to replicate the
result of an experiment, etc. The CSO of a biotech company gave one example of
the sharing of such tech know-how
For example, one scientist from the university can tell another scientist from
the private industry when you do a searching work, what kind of technique
or what types of commercial product are better for certain purpose. Those
are helpful.

Technical know-hows are often shared through the day-to-day personal interaction
among scientists instead of formal technology transfer agreements.
Research equipment and materials are also shared between academia and
industry. For instance, companies and universities can share the use of expensive
research equipments. Or they can use each other’s proprietary materials or data for
their own purposes. They are often shared through research collaboration and
material or data transfer agreements.
99

Knowledge about industry and market is another type of knowledge shared
between academia and industry. An understanding of the industry and market can
help both biotech companies and academic scientists identify promising directions
of investigation and have a realistic evaluation of the commercial prospects of their
current projects. A professor from a major private research university gave the
following example:
Very recently we had a heated discussion with a company that is developing
a product. And I know about the product and I said, "Josh, the window of
your product is about three years because there is something else coming
along, and your product is going to be eclipsed and totally useless after
three years."

Knowledge about who-is-doing-what is another useful information shared
between academia and industry. It resembles a transactive memory system in which
it is more important to locate the knowledge needed rather than actually possessing
the knowledge (Hollingshead, 1998). The Chairman of a biotech consulting said,
It would be related to the scientific disciplines that you are familiar with or
the questions going to be posed. You know. Whether you are aware of
something else that looks interesting at your school or whether you are
aware of someone doing something.

Traditionally this knowledge has been shared through informal channels such as
personal communication, conferences, or consortiums. However in recently years
many biotech-consulting firms have played the role of knowledge brokers. They
often have extensive connections and knowledge about in both industry and
100

academia and they make profit by helping biotech companies to locate the
technology or information they need.
The characteristics of the knowledge shared often affect the channels used
in knowledge sharing, people’s attitude and practice in knowledge sharing, and the
outcome of knowledge sharing. Interviewees’ narratives about academic-industry
knowledge sharing concentrated on two dimensions of knowledge: articulatedness
and market readiness.
Articulatedness of knowledge
Articulatedness of knowledge refers to the extent to which knowledge is codified.
The more articulated a piece of knowledge is, the easier it can be shared. In term of
its articulatedness, three types of knowledge shared between academia and industry
are identified: embedded knowledge, codified knowledge, and tacit knowledge.
Embedded knowledge: Embedded knowledge is encoded in the materials,
data, or equipments a person or an organization possesses. One example is the
genetic information embedded in genetically modified mouse used for medical
research. Another example is the gene database used by bioinformaticians.
Embedded knowledge comprises a considerable part of the knowledge shared
between academia and industry. The director of Intellectual Property Office of a
major public university said to the investigator, “I think material transfer agreement
is really a huge area of knowledge sharing. We do probably more than 600 hundred
incoming material transfer agreements each year. That’s just from industry.” The
101

sharing of embedded knowledge is usually achieved through the transfer of the
material or database from one organization to another protected by transfer
agreements, and, thus, can be done relatively straightforwardly with little need of
further interaction or communication.
Codified knowledge: Codified knowledge refers to the knowledge clearly
articulated in publications, conference proceedings, patent documents, or other
written or spoken formats. This kind of knowledge can "be communicated from its
possessor to another person in symbolic form and the recipient of the
communication becomes as much 'in the know' as the originator" (Winter, 1987, p.
171). Compared to knowledge embedded in materials, equipments, or data,
codified knowledge is harder to share. The successful utilization or
commercialization of the codified knowledge often requires the simultaneous
sharing of tacit knowledge.
Tacit knowledge: Compared to embedded knowledge and codified
knowledge, tacit knowledge is the least articulated and least systematized, and is
consequently most difficult to share. In academic-industry knowledge sharing,
many types of tacit knowledge are involved, for instance, technical know-how,
understanding of science and technology (not codified in patent or publication), and
research techniques. One interviewee discussed the sharing of tacit knowledge,
saying,
Well, it's not always information we would say was patentable intellectual
property, but it's know-how. It's idea sharing. It's a sense of a way of
102

working, an alternative methodology. It really happens in conversations that
go alone the line, "Well, have you tried this?" That might not be knowledge
per se, but it's an alternative that someone has come up with.

Tacit knowledge is inherently more difficult to share for a number of reasons. First
as tacit knowledge is often embedded in the social, cultural, and cognitive contexts,
people are often unaware of the tacit knowledge they possess and unable to
articulate it (Choo, 1998). Furthermore, individuals do not need to externalize their
tacit knowledge in order to benefit from it, and as a result, they usually do not have
the incentive to codify their tacit knowledge (Stenmark, 2000-2001). Finally the
articulation and sharing of tacit knowledge might cost people their competitive
advantage and power, which will further hinder tacit knowledge sharing (Bordum,
2002; Leonard & Sensiper, 1998).
Market readiness of knowledge
Market readiness is another dimension of the knowledge shared between academia
and industry. Some knowledge is highly market ready, while others are very far
away from market. Market readiness of knowledge includes two parts: (1) the value
of the knowledge once it is developed into products, or the potential financial value
of the knowledge, and (2) how developed the knowledge is, or, in other words, how
far away the knowledge is from market. A piece of knowledge is first developed in
research labs as proof of concepts, which is later developed into lab scale
prototypes. Later this piece of knowledge undergoes pre-clinical trials and clinical
trials until it becomes product in the market.
103

The market readiness of a particular piece of a particular piece of
knowledge decides the channel adopted in knowledge sharing, the attitudes of the
two parties involved, and how strongly they claim their intellectual property rights.
This is discussed in the next two sub research questions.
RQ2b: How is the knowledge shared between academia and industry protected as
intellectual properties?
In the process of academic-industry knowledge sharing, various types of
knowledge are protected as different kinds of intellectual properties. There are three
general approaches to protect one’s intellectual properties: patent, trade secret, or
copyright.
Patent
Patent represents an agreement between an individual person/organization with the
government(s). The individual discloses all the contents of a new invention to the
government in exchange of the exclusive rights to utilize the invention and to
prevent others from using it for a fixed period of time (usually 20 years). For an
invention to become a patent, it has to be new, nonobvious, and useful. Any illegal
use of the patented information is subject to litigation. Patents are commonly used
in protecting biological, pharmaceutical, and biotechnological products and tools
(Californian Institute of Regenerative Medicine, 2006).
104

Trade secret
Instead of disclosing one’s valuable invention or information in exchange for
patent, one can protect it by keeping it confidential, or, in other words, a trade
secret. While a patent is only protected for a limited period of time, a trade secret
can last forever. However, the downside of protecting one’s IP as trade secret is
that the rightful owner of the trade secret cannot prevent others from reverse
engineering the same piece of information (Daizadeh et al., 2002).
Copyright
Beside patenting or keeping knowledge as trade secret, one can publish it and get
protection through copyright. Copyright is protected during the lifetime of the
author and seventy years afterwards. However, after being published, the
knowledge becomes “public.” In other words, everyone can use it for his/her own
end. As a result, by publishing the information, the inventor virtually gives up any
possibility to harvest financial rewards from it. Following the same logic, people
are free to share all knowledge that has been published in journals or presented in
conference, because this kind of the knowledge is considered to be already in the
public domain.
RQ2c: What are the channels of academic-industry knowledge sharing?
Different types of knowledge are shared through different channels of knowledge
sharing. Table 4.1 presents a list of the channels of academic-industry knowledge
sharing and the frequencies with which they were mentioned by interviewees.
105

Channels of knowledge sharing Frequencies
Patent licensing 61
Research collaboration 45
Publication 27
Conferences and seminars 26
Consulting 23
Personal communication 21
Spin-off 20
Research contract 13
Research grant 12
Material and data transfer 11
Invited visits and talks 10
Education and training program 7
Special research centers 7
Knowledge brokers 4
Invention assignment 3
Hiring 3
Strategic alliances 3
Table 4-1. Channels of academic-industry knowledge sharing
Patent licensing
Among all channels of academic-industry knowledge sharing, patent licensing is
the most frequently mentioned and it is not surprising at all. According to the
annual survey conducted by Association of University Technology Managers
(AUTM), the licensing of university patents has increased dramatically in the past
decade (Thursby & Thursby, 2000). Number of university licenses increased by
75% from 1991-1996 (Thursby, Jensen, & Thursby, 2001).
Patent licensing is a highly formal channel of interorganizational knowledge
sharing. A company licenses a patent from a university or another company by
paying a license fee as well as other royalty payments, and in exchange, acquires
106

the exclusive or non-exclusive rights to utilize and commercialize the technology
protected by the patent. Interestingly, some interviewees considered patent
licensing as almost a “ceremonial” channel of knowledge sharing. They argue that
because in submitting a patent application, the inventor has to disclose all related
information to the government and make it available to the public, whatever
knowledge shared through patent licensing is, in a sense, already public. In this
way, what is shared through patent licensing is not the knowledge per se, but the
right to use and commercialize the knowledge.
Furthermore, many interviewees pointed out that patent licensing only
transfers a very small amount of knowledge, as IP laws often prevent the sharing of
more knowledge, especially tacit knowledge between the two parties. As a result,
Nelson (2001) discredited the belief that university patenting and licensing leads to
effective knowledge sharing, and called it a “myth” (p. 16).
In addition, interviewees indicated that majority of the patents biotech
companies, especially large established biotech/pharma companies, licensed in are
from other biotech companies, instead of from academic institutions. This is mainly
due to the fact that academic research is often basic research with low market
readiness. According to Thursby et al (2001), of all patents licensed from
universities, 45.1% are proof of concepts without prototypes, 37.2% are lab scale
prototypes that need further development, and only 12.3% are ready for practical
and commercial use. As a result, small dedicated firms often license in these early
107

technologies from academia at a relatively low price, develop them until the
technologies are close to market, and license them out to large established firms.
University spin-off
University spin-off is another important channel of academic-industry knowledge
sharing. Several interviewees identified their own companies as spin-offs from
universities. Even though many different definitions of spin-offs have been
proposed in the literature
8
, most of my interviewees defined university spin-offs as
companies meeting the following criteria: (1) it is founded by faculties, researchers,
or students in a university; and (2) it is based on the intellectual properties
developed in the university, and (3) this intellectual property is already patented.
Spin-offs
Spin-offs are especially prominent in academic-industry knowledge sharing in the
biotech sector and in the commercialization of scientific discoveries made in life
science disciplines. For instance, according to the data of Shane (2004), among all
spin-offs of MIT between 1980 to 1996, 41% are biotech companies (including
biotech and medical device), which is only followed by software industry (23%),
computer hardware companies (6%) and semi-conductor industry (4%). Similar
patterns were found Lowe (2002)’ study of spin-offs from the University of
California system, Golub (2003)’s study of spin-offs originating from both

8
Scott Shane (2004) defined university spin-offs as start-up companies that exploits the intellectual
property of a university, no matter the intellectual property is patented or not. Edward Roberts
(1991), on the other hand, considered all companies started by people who used to study or work at
a university spin-offs.
108

Columbia University and New York University, and Sobocinski (1999)’s study of
University of Wisconsin system.
Material and data transfer
Material and data transfer refers to the sharing of proprietary materials and data
between two organizations. These materials and data are often protected by patents.
Universities often acquire these materials and data for their faculties to conduct
research with, and often in exchange, give the companies the right to utilize the
discoveries made based on those materials or data. The Director of the Intellectual
Property Office of a major public research university said,
We do a lot of material transfer agreements where we take in propriety
materials. They are often proprietary drugs or materials that companies have
that our researchers want to work with. It’s difficult to get access to them,
and the company is the only source if the materials are proprietary. So we
will do material transfer agreements to bring in those materials, but what is
required is that we share our data with the company, and also give them
access to the intellectual property that arises from the use of their materials.

Material and data transfer is a highly formal and straightforward channel of
knowledge sharing. It is almost entirely done through legal contracts and
agreements without much need for further explanation and interaction.
Publication
Publishing their research in academic journals is another basic channel through
with academic scientists and industry scientists share knowledge. Once a piece of
knowledge is published, it becomes public and can be used by anyone. Compared
to other channels of knowledge sharing, such as patent licensing or material
109

transfer, publishing is an untargeted channel of knowledge sharing. In other words,
the source or the creator of the knowledge does not know who will access and
utilize the knowledge. One professor referred to it as “throwing it over the wall,”
because “you don’t talk to the other side, and you have no idea who will get it and
in which way they are going to utilize it.”
Conference, seminars, and consortiums
Conferences, seminars, and consortiums belong to another category of academic-
industry knowledge sharing channels. Most of the scientific and technical
knowledge shared through these channels are often public knowledge, information
that has been published or disclosed to the public in other way. At the same time,
conferences, seminars, and consortiums provide an ideal venue for the sharing of
knowledge about industry, disciple, and knowledge about who-is-doing-what. Such
interactions often lead to further and more in-depth sharing of private knowledge
through other channels, such as patent licensing or research collaboration.
Consulting
Consulting is another prominent channel of academic-industry knowledge sharing.
In this case, companies pay academic scientists a fee in exchange for their time and
expertise to advice companies R&D. All academic scientists interviewed for these
study have consulting relationship with biotech companies. Consulting is usually
the personal behavior of university faculties, and it usually doesn’t involve the
sharing of any patented knowledge. As a result, it is highly informal.
110

Research collaboration
One important channel identified by the interviewees yet rarely discussed in
literature is research collaboration. Research collaboration includes a variety of
ways in which academic institutes and commercial companies pool their financial,
intellectual, or material resources together for the discovery of new knowledge and
the development of new products. Research contract and research grant are two
common types of research collaboration.
Research contract refers to the agreement between a company and an
academic lab by which the academic lab uses its own equipment and intellectual
capital to help the company solve a certain problem in exchange for financial
support or other returns. In this case, the results of this collaboration will remain a
trade secret possessed by the company, and the academic researcher has no claim
on the intellectual property and no right to publish the results. Research contract is
biotech companies’ favorite way of collaborating with academics because it enables
companies to “outsource” a particular scientific or technical problem to academic
researchers who have more expertise on the subject for a reasonable cost. One
example of contracted research is clinical trials, in which a biotech company
commissions a university to test the safety and effectiveness of a new therapeutic
product, be it a new drug or medical device, to meet the requirement of FDA
approval at the university’s hospital.
111

Sometimes, a company offers academic scientists a grant to pursue a subject
that is to the interest of the company. In this case, the academic researchers have
considerable intellectual freedom and often the right to publish their results. Or
sometimes, academic scientists and commercial companies apply for federal grants
together and once they get the grant, they will collaborate and contribute to
different stages of the grant. One example is the Small Business Innovation
Research (SBIR) grant.
Almost all the academic and company scientists interviewed identify
research collaboration, both funded and unfunded, as the most important
knowledge-sharing channel and the primary way through which they work with
industry. Compared to transaction-based channels such as patent licensing or
material transfer, research collaboration often involves long-term interaction
between academia and industry on both the organizational level and the individual
level. Furthermore, academic scientists argued that while most of the other
channels are merely the venue for the transferring of knowledge from one
organization to another, different forms of research collaboration not only
facilitates the sharing of knowledge but also promotes the innovation and the
creation of new knowledge.
The ownership of new knowledge created through research collaboration is
sometimes problematic if two parties involved failed to reach a clear agreement
from the very beginning. The CSO of a biotech company commented,
112

However, once it comes to the so-called creation, the new discovery type of
discussion, it would cause problem. For example, if you have a professor in
discussion with a scientist from private industry, talking about some new
discovery idea, and eventually each one contribute to part of it. There is a
new idea eventually coming out. And [the] university files a patent, or the
company files a patent application. I am very sure that down the road, there
will be legal battle that will cause trouble. A friendly collaboration may turn
out to be an adversarial type of problem. So personal interaction, nobody
should prevent that, because personal interaction is good. However, because
of our environment for the past twenty years, the condition, especially the
financial incentive is huge, so a lot of times we do see everybody becomes
very cautious.

Personal communication
Personal communication is a highly informal channel of interorganizational
knowledge sharing. Knowledge sharing through personal communication usually is
not protected or regulated by any legal contracts but guided by personal beliefs and
ethical codes. Personal communication as a channel of academic-industry
knowledge sharing is mainly on the individual/person level instead of between
organizations.
Knowledge sharing through knowledge brokers
The fact that academic research is often too theoretically oriented and too far away
from market, and that discoveries made in university labs often need further
development before they become attractive targets of commercial companies leads
to the emergence of knowledge brokers. The manager of an industry-oriented
research center commented on the role of her organization, saying,
The issue today for biomedical sector is the development cycle is so long,
because you have to get FDA approval. Venture firms will not be inclined
113

to invest in very early phase inventions. However there is a world full of
idea in universities … that deserve to be groomed for commercialization.
But there is no money. The only thing is grant money, and it's quite
competitive. So [we] wanted to be in between there to allow those early
phase inventions to be developed and accelerated through the process, and
bring them out to the industry when they are much more mature and can be
picked up either by venture firms or large companies.

These knowledge brokers can take many different forms, such as companies
specialized in helping biotech companies identify and locate the technology they
need and industry-oriented research centers
9
. Knowledge brokers often have an in-
depth understanding of the biotech industry and its market. They also have
extensive ties with universities and research institutions that enable them to
facilitate the transfer of knowledge from academia to industry and to profit from it.
Strategic alliances
Even though strategic alliance has been identified as an important channel of
academic-industry knowledge sharing in literature (e.g. Eisenhardt &
Schoomhoven, 1996; Hagedoorn, 1993a; Hagedoorn & Schankenraad, 1994; Han,
2004; Lerner & Merges, 1998; Monge et al., 1998; Oxley, 2004), few interviewees
mentioned it. To many of them, the interactions between academic institutes and
industry are often more transaction-based rather than strategically based long-term
collaborations. The Director of Technology Licensing of a major medical research
institute said,

9
One example of such industry-oriented research center is the Al Mann Institute at University of
Southern California.
114

Strategic alliances actually are a big word. We don’t do a strategic alliance,
but we do transaction-based partnerships. And again, there is not a case
where [we] will become a GlaxoSmithKline partner or Norvatis partner.
There is just case in which companies come to us to make certain that we
can …and we offer benefits for them as well. So strategic alliances may
work for other institutions, but it doesn’t work for us. Just we do limited-
time interactions and collaborations

Which channel to use?
Interviewees from academia have very different preference for channels due to
their different motivations in academic-industry knowledge sharing. Technology
transfer officers, representing the interest of the institution, prefer patent licensing
as licenses can potentially bring huge financial return to universities and research
institutes. However, because academic scientists have different goals in academic-
industry knowledge sharing, they often prefer research collaboration and consulting
as the channels of knowledge sharing. Research collaboration can get them
immediate funding and other resources for their own research and allow them to get
access to the cutting edge research conducted in companies. Consulting, on the
other hand, can get them quick individual financial return.
Formality of knowledge sharing channels
All the channels of academic-industry knowledge sharing discussed above can be
evaluated in terms of formality. Formality refers to the extent to which the
knowledge sharing through a particular channel can be accomplished solely
through legal/formal contracts. For instance, the transfer of material and data, the
licensing of patent, and publication are highly formal, and knowledge sharing
115

through these channels is often relatively straightforward and transaction-based. On
the other hand, the sharing of knowledge through consulting, research
collaboration, or personal communication can be highly informal.
An analysis of interviewees’ comments on channels of knowledge sharing
identify a positive relationship between the articulatedness of the knowledge and
the formality of the channel(s) through it is shared (See Figure 4.2). In other words,
formal knowledge sharing channels are used when the knowledge shared is highly
articulated, or when the knowledge is embedded in concrete objects, such as
proprietary materials and data. On the other hand, when knowledge is less
articulated, and more tacit, people tend to choose more informal channels.

Figure 4-2. Relationship between the formality of the channel and the
articulatedness of knowledge
116

Furthermore, a positive relationship can be found between the formality of the
channel and the market readiness of the knowledge (as illustrated in Figure 4.3).
Highly formal channels are likely to be used in sharing knowledge with high
market readiness, while informal channels are likely to be used in sharing
knowledge low in market readiness. For instance, companies and academic labs
often share early stage knowledge through research collaboration. However, once
the knowledge becomes more developed and the market potential of the knowledge
becomes more obvious, companies are very likely to license this piece of
knowledge for their own use.

Figure 4-3. The relationship between the formality of channel and maturity of the
knowledge

117

Use of multiple channels
While each channel discussed above enables the sharing of knowledge between
academia and industry, another insight emerging from the interviews is that these
channels are often used in combination. For instance, companies not only have
university faculties on board as advisors or consultants, but also license patents
from universities. Furthermore, they could engage in research collaboration with
universities, and hire graduates and scientists from universities. Similarly,
university labs often have multiple ties with commercial companies. The using of
multiple ties enables the sharing of embedded, codified, and tacit knowledge
simultaneously, and, as a result, often leads to more successful collaborations and
commercialization.
RQ3: How does professional culture (of academia and industry) affect the
knowledge sharing between the two communities?
Professional culture has been identified as an important factor influencing the
sharing of knowledge between academia and industry (e.g. Siegel, Thursby,
Thursby, & Ziedonis, 2001b). Interviews with professionals from the two
communities identified both divergence and convergence of the two cultures and
how the characteristics of the two professional cultures influence the motivation,
process, and outcome of academic-industry knowledge sharing.
In making sense of the culture of academia and the culture of industry, I
will mainly rely on the integration and the differentiation perspectives reviewed in
118

Chapter 2. More specifically, while the general and shared characteristics of each
culture can be identified, the fine differentiations within each culture will also be
discussed. At the same time, even though I don’t interpret these two cultures as
“fragmented,” I do recognize the complexities and inconsistencies in interviewees’
narratives regarding the two cultures. Finally, in discussing the cultural
characteristics of academia and industry, I will try to refrain from using
representational language to characterize culture, and let themes and subthemes
emerge from the narratives. The findings discussed next are only my interpretation
of the narratives of my interviewees.
The divergence of academic and industrial culture
The culture of the academia and the culture of the industry are very different.
Interviewees tend to describe these cultural differences through pairs of dialectics.
Table 4.3 presents a list of the dialectics interviewees used in discussing the
characteristics of the two cultures and the frequencies with which they were
mentioned. These dialectics will be discussed in detail next.
119

Characteristics Academia Industry Frequencies
Institutional incentive Science driven Bottom line
driven
26
Personal reward system Publication,
funding
Milestone 15
Characteristics of research
Discipline vs. freedom Academic
freedom
Discipline 18
Timeline Long Short 15
Concentration vs. distraction Distraction Concentration 18
Teamwork or independence Independence Teamwork 6
Characteristics of knowledge management
Emphasis on IP protection No Yes 11
Openness vs. secrecy Openness Secrecy 30
Others 3
Table 4-2. Characteristics of academic culture and industry culture
1. Institutional incentive and personal reward system
Institutional incentive and personal reward system are at the root of the many
differences between the culture of academia and the culture of industry. On the
institutional level, industry is bottom-line-driven, while academia is science-driven.
As a result, biotech companies often invest in projects with foreseeable economic
gains. On the contrary, even though sometimes the work of academic researchers
leads to enormous economic gains, the most important goal of their research is
scientific discovery while monetary benefits are only the by-products of their work.
On the individual level, company employees are driven by the need to delivery
milestones and products in a timely manner and they are rewarded accordingly in
terms of promotion, salary raise, or company stocks. Academia follows a different
set of reward system in which faculties want to get their work published and
120

recognized. The Assistant Vice Chancellor of Intellectual Properties at a public
research university commented on this difference, saying,
The culture of industry is very simple. They are held responsible to the
stockholders, so bottom line is very important to them. That doesn't mean
they are bad. People invest in companies because they want to generate
financial return. So to them, doing well financially is important. On the
university's side, we certainly want to do well (financially), but doing
(public) good is important. …We cannot just simply tell a professor
“because there is big money here, drop everything you are doing regardless
of your interest. Let's just do this.” That just doesn't happen.

The differences in the institutional incentive and personal reward system lead to the
different characteristics of academic research and industry research and to the
different manners with which new knowledge is created and managed in these two
communities.
2. Characteristics of academic and industry research
Several pairs of dialectics emerged in interviewees’ discussion of the characteristics
of academic research and industry research, including: curiosity-driven vs. market-
driven, distraction vs. concentration, independence vs. teamwork, and, academic
freedom vs. discipline. These dialectics illustrate the distinctive natures of
academic research and industry research.
Curiosity-driven vs. market-driven: Even though many biotech companies
engage in cutting-edge research that is comparable in quality to any leading
academic labs in the world, industry and academia are very different in how they
choose the subject of their research. Industrial research is often more project-based
and market-driven. Biotech companies choose a particular research topic because
121

they believe that this line of research will eventually lead to the development of
products with good market prospect. On the contrary, the innovation process in
academia is often curiosity-driven. Academic scientists’ choice of direction of
research is based solely on their desire to explore a new area of scientific inquiry or
on what the discipline considers as important and promising. The same Assistant
Vice Chancellor quoted above said,
Many people in the university are curiosity-driven. We want to find out why
this happens this way. To the extent that whether somebody could benefit
from what we find out, or whether we ourselves could benefit from it is
more secondary. So the culture in industry is really bottom line driven. We
are more curiosity driven.

Distraction vs. Concentration: The second major difference between
academic research and industry research is that between distraction and
concentration. Because academic research is curiosity-driven, it is legitimately full
of distraction. Since academic researchers enjoy greater freedom, they can explore
new questions and new areas out of their intellectual curiosity. This is one
important way in which groundbreaking discoveries are made in the academic
world. A professor said,
Basically industries are objective driven, and particularly very small biotech
industries are very immediately driven. One of the big tricks of
biotechnology industry is not to attempt too much and discover too much.
Take one idea, commercialize it, and don’t get distracted. The academic
culture is driven in totally different way and must be. It is rich in
distractions. If you lose sight of your original objective, it doesn’t matter if
you replace it with a better objective. So it is not objective-driven at all. It is
curiosity driven. They are very very very different.

122

On the other hand, the goal of industry research dictates that it should be very
concentrated and free of distractions. Industry research is highly focused and
organized, and company scientists cannot invest time and effort to explore new
questions simply because these questions are “interesting.”
However, some interviewees pointed out that even though academicians
tend to think of any form of discipline as against their sacred right of academic
freedom, very often discipline is essential in pushing an academic discovery
through the process of development until it becomes a marketable product.
Independence vs. teamwork: Another important difference between the
culture of academia and that of industry is that academia values independence,
while industry values teamwork. In companies, research work is often
compartmentalized and distributed. Different divisions, departments, and teams
work in collaboration for a shared goal: to develop product and to make profit. The
situation is quite different in the academic world. Even though there are indeed
much collaboration among different labs, departments, and universities, the
interdependence among different Principle Investigators (PIs) is by no means
essential to the survival and development of their individual labs. Every PI is
independent in managing his/her own projects, people, and funding. The CSO of a
dedicated biotech firm said,
One very distinct difference between the two cultures is in the company,
like my company, … we work as a team. But in universities, each lab works
as an individual party. Sharing knowledge sometimes turns out to be
123

negative in the university because another researcher may get the idea and
jump ahead. But in a company, everybody has one goal in moving forward
in one direction. In the university, each lab is one entity. So although
university has more professors and more research scientists, but it turned
out becoming many small entities, but in a company, although there is a
smaller number of people, but it’s only one entity. Therefore, culture-wise,
that’s a big distinctive difference.

This tradition of independence is cultivated and passed on from one generation of
academics to the next through the training of graduate students and junior scientists
in academic labs. In academic institutions, young scientists are often given a lot of
freedom in discovering their own interests and establishing their own intellectual
development paths. One senior scientist of a very large and successful biotech
company recalled,
When I was at graduate school, or doing a postdoc, although the PI was an
expert in certain area, usually [he gave] the students or postdocs a lot of
freedom to do different things. [He didn’t] really worry about how you get
things done. [As a result,] usually… graduate students or postdocs work on
their own project. There is not whole lot of collaboration with other
scientists within the lab, coz each graduate student has his own dissertation
research, and postdocs have their own research and they want to publish
papers. …So within a particular lab, teamwork isn’t that much emphasized.
You are pretty much on your own [if you are a] graduate student or postdoc.
But that’s not to say that there is not a lot of collaboration among academic
scientists. Of course, there are. But I think in terms of emphasis on
teamwork, there isn’t that much as in industry. In fact, if you go to a job
interview in academia, you will not see a lot of people [who] want to know
how well you can work in a team. [laugh]

If you compare [academia] to industry, teamwork is very important in
industry, because to get things done more efficiently, industry is like the
assembly line. You need collaborations between different parts or functions.
So in company, you also emphasize teamwork, especially across different
functions so that you can push the product as fast as possible.

124

Academic freedom vs. discipline: Related to the dialectic of independence
and teamwork, academic researchers often enjoy academic freedom, while industry
scientists are often subject to much discipline. In general, companies exert much
more control over their employees than universities over their faculties. Companies
have complete control over their scientists in terms of what they work on and how
they conduct their research. Furthermore, companies also have sole ownership of
any intellectual properties generated and complete control on how these IPs should
be shared, hoarded, or utilized. Universities and research institutes, on the other
hand, traditionally give their faculties considerable autonomy and freedom. One
intellectual property manager at a public research university commented on this
difference, saying,
The difference is that in academia, you don’t have the employees the way
you have employees in industry. So industry comes to us and expects that
we are going to be able to control what our investigator does, but we can’t
control our investigators the way companies control their employees. Our
investigators have to be able to collaborate openly with others, have the
ability to freely publish, to go to conferences, and to disseminate
information freely.

Similarly, a senior scientist at a large biotech company said,
Of course the company is for profit. So you would expect quite reasonably
that all your work contribute to the success of the company. So in that
regard, you cannot do any free-will science at the company. You cannot be
working for a couple of years, thinking you are just doing basic research,
but not contributing to the company's bottom line. So in that regard, if you
compare scientists working in industry and scientists working in academia, I
guess the scientists in academia probably have more freedom. They could
pretty much work on whatever they are interested in and want.

125

Long vs. short timeline: Industry research often has a much shorter timeline
than academic research. A professor at the medical school of a private research
university applauded the speed with which innovation moves forward in biotech
companies, saying,
I have been impressed with the quality of the scientists, the technical
capabilities, and the rapidity with which they focus on a project and move it
forward in industrial setting, much more rapidly than the university
professor could in his laboratory, and that’s one of the things that impressed
me tremendously. I mean, we could spend 10 years, trying to move from
point A to point B with a limited manpower in my laboratory, compared to
a good pharmaceutical company. They can move from point A to point B in
less a year, maybe 6 months. It would take me 10 years to do that.

This difference is result of the two cultural differences discussed above.
First, industry research is much more focused than academic research, and is
naturally faster. Second, companies have more control over their scientists and their
resources, and, as a result, they can assemble a critical mass of resource and
brainpower in a short period of time to concentrate on a project the management
perceives as having market potentials. Academic labs, on the other hand, are at a
disadvantage in this respect. Academic labs usually have less resource than
companies. Most of them also have fewer scientists than companies. Furthermore,
the scientists in an academic lab often have their own individual projects and it is
hard to make them all concentrate on one project. To make it worse, because
academic labs have high mobility, graduate students and postdocs come and go
very frequently. As a result, research in academic labs often lacks continuity and a
126

lot of time is spent on training the new scientists so that they can continue with the
current research in the lab.
Sometimes, this difference in timelines can cause problems when
companies and academic labs are collaborating with each other. Companies are
often frustrated with the slow progress of academic research. A senior scientist at a
major biotech company said,
In a lot of the projects collaborating with academic scientists, you provide
them with 1 or 2 year's funding, and usually they will work at their own
pace in their way. Company usually wants update in different stages, but
I'm not sure it always happens though. Sometimes, in certain cases you
could have collaboration going you, you fund them for a year, and you think
for every 3 or 4 months, you will have an update. The professor, for
whatever reason, may have a graduate student or postdoc working on this
for a while, and then all of a sudden, that graduate student is gone or
postdoc is gone, and he has to find another person to do this. So it may not
be as smooth as you think. But for scientists in company, we operate under
very well defined timeline, so usually we want to know after three months,
where we are, so we can plan for the next step, right? So that is very
different from academia, where" if I cannot solve this project, I will just
work on it for another year, and eventually I will solve it." For company,
they just don't have that much time usually [laugh].

The convergence of academic and business culture
Despite the distinctive differences between academic and industry culture discussed
above, a tendency towards convergence has been observed. This is especially
salient between academia and knowledge-intensive industries such as
biotechnology industry. When being asked to comment on the differences between
academic culture and industry culture, a professor answered,
127

Well, I think if you had asked me this question 20 years ago, the answer
would be considerably different than the answer I would give you now. I
think whereas in the past, the academic feeling was that we were in the
white towers doing research for research’s sake, whereas now the
academicians see the alliance with industry as a way of getting their
technologies that they developed into commercial products. As I have
already discussed that … the culture has changed and there is sort of a
coming-together of industry and academicians.

The convergence of the two cultures is demonstrated in a few ways. First, both
academic institutions and biotech companies have a thirst for knowledge and put
great emphasis on scientific discoveries. Compared to other industries, biotech
industry is extremely new and highly knowledge intensive. As a result, some
biotech companies conduct world-class research that is comparable in quality to
any leading academic institutions. Sometimes biotech companies will even invest
in basic research that is related to their R&D. According to a technology transfer
officer,
The similarity lies in the striving for scientific excellence and discovery.
Both scientists from academia and scientists from industry have the similar
goal.

Similarly, a company scientist said,
Well, there are a lot of similarities between the academia and industry
because even [though] biotech is a 30-year old industry, it is still pretty new
compared to some other traditional industries. If you go to a biotech
company, you will see people dress very casually and engage in very active
and casual talking all the time. So there [are] a lot of similarities… So even
[though] we are developing some real commercial products, these products
are actually frontier research products. So we have to do it in the way [in
which] top scientists are doing it. We have to focus on the real intellectual
128

researches, not in human, financial rules or doctrines of commercial
industry development. So we are doing similar things, basically.

In the past, the research conducted in companies was often considered inferior, and
company scientists were, as a result, not as highly regarded as their academic
counterparts. This is also changing. A professor pointed this out by saying,
The culture has changed and there is sort of a coming-together of industry
and academicians, also in terms of the people now. So postdocs coming out
of the basic science department at the medical school previously would aim
for an academic position to become an assistant professor and academician,
but now there seems to be equal interest in either going to the academic
world or the industrial world, because the capabilities of industry are very
significant now. They have great technologies. They’ve got great
equipments. They have money. And so whereas previously if you went into
industry as postdoc, it's sort of your career was ended in terms of any
academic situation. You are shuttled off to the industry. But now, that is not
so much a fear, although in general if one does a postdoc in industry, one
tends to stay in the industry. There are by and large excellent scientists
equivalent in many cases to the scientists in the academic world. So I think
there has been a coming-together of both the industrial people and the
academicians in terms of their qualifications and understanding of one
another.

Second, while company is moving toward academia in terms of the research they
conduct, today more and more academicians are also moving from the ivory tower
to the office tower. The traditional notion of academics as conducting pure
scientific research is giving way to a new professional culture and institutional
environment that encourages applied research and promotes the commercialization
of academic discoveries. A senior professor who have spent his entire life
conducting research at universities described the old tradition, saying,
129

There is a tradition that came originally from European academics. It was
the resistance to sharing information. In the British and German tradition,
for example, professors really didn't want to share his time and information
with industry, with the students, fine, with other professors, fine. But [now]
an agency that supports science said, "We want you to work with industry.
We will give you money if you work with industry." And one old English
professor of science actually cried when he addressed the group, saying, "I
will not prostitute my science." So it will never be spoken of, but there is
this secret opinion that many scientists have the idea that talking to industry
or commercializing their work is a prostitution of their work. That's it.
That's disappearing. That's with the older group. But it is nonetheless the
feeling.

However, a new tradition that more is industry-friendly and entrepreneurially
minded is gaining momentum in the American academia. Today, universities and
academic institutes encourage their faculties to conduct applied research and to
collaborate with industry. Academic institutions often provide faculties with
administrative assistance and other incentives. For instance, some universities have
policies that grant their faculties time and resources to work with industry. A senior
scientist at a biotech company commented on this new tradition by saying,
Well, I think in the US, researchers in universities not only are interested in
publishing papers. They also want to apply their intellectual properties to
real world to produce real products for the society. So that's a very good
advantage of the American research system. A lot of universities are
actually very active in developing collaboration with industry to apply their
research products into industry. They don't want to stay in the ivory tower.
They would also be glad to see their products being used for society.
Professional culture and knowledge sharing
The difference in the institutional incentive and personal reward system, and the
dissimilar ways in which research is carried out in academia and industry causes
130

academia and commercial companies to manage their intellectual properties and
collaborate with the other side in very different manners. In the next section of this
chapter, I am going to discuss the differences in the ways in which academia and
industry share knowledge with each other, and argue that some of the common
beliefs about academic-industry knowledge sharing are sometimes “myths.”
One common conception or misconception regarding academic-industry
knowledge sharing is that academia is more open in spreading its knowledge, while
industry often guards its knowledge with great caution. For example, a manager of
a biotech company specialized in developing university inventions said,
There is a major difference that has to do with the sharing of information.
The faculty or the doctor has a mission to spread the information. They are
very open. They publish a lot. In fact they are measured on publishing.
That's how they get their promotion to tenure. For us, because we need to
protect the intellectual property during the first few weeks, we want to see
that there is no sharing at all, there is no publishing done, and there is no
public speech on it. Otherwise it becomes prior art, and we don't have a
patent position anymore. So it's a little bit of opposite to the culture of a
university.

In fact, academics take great pride in their openness and in their rights to be open.
Many of the academic scientists believed that being open is essential for their
research and for the development of science.
However, an analysis of the different attitudes of academic scientists,
technology transfer officers, business executives, and company scientists
demonstrate that it is too simplistic to label academia as “open” and companies as
“secretive.” My interviews suggest that both academic institutions and companies
131

could be very aggressive in claiming and protecting their intellectual properties.
This is especially true today when the potential financial reward of a particular
technology is astronomical.
IP ownership: both academic institutions and commercial companies claim
ownership over any profitable intellectual properties generated by their employees
such as patents. How strongly universities and companies exercise their ownership
depends on the market value of a particular technology. One technology transfer
officer from a major private medical research center said,
We know and we kind of support academic-academic interactions should be
freer than academic-industry interactions, because the immediate file you
lose or gain is less in theory in general.

A piece of IP co-developed by university and industry can become problematic
with or without prior agreement on its ownership.
IP protection: a comparison of how academic institutions and companies
deal with three major types of intellectual properties: copyright, patent, and trade
secrets discloses more similarities than differences. First, in terms of publication
protected by copyright, even though publication is the most predominant channel
through which knowledge is shard within academia, within industry, and between
the two communities, both academic institutions and biotech companies try to
make sure that the knowledge they share through publication is either of no market
value or the knowledge is already patented before publication. The difference
between the two is that companies can often reinforce this rule more effectively
132

given their better understanding of market and their tighter control over their
scientists.
Second, the patents owned by universities and companies are protected by
the same patent laws. How strongly universities and companies defend their patents
against infringement depends on the market value of the technology and on how
much resources they have to protect patents. Patent litigation tends to be very time
consuming and expensive. Due to their lack of resources, academic institutions
often cannot protect their patents as well as commercial biotech companies.
Finally, most complicated and difficult of all, the protection of trade secrets
is determined by three factors: employee’s desire to share knowledge, the
ownership of the proprietary information, and organization’s control over their
employees. First, academic scientists, especially university faculties have the
inborn need and urge to publicize their research achievements through publication,
through lectures, seminars, conferences, and even through casual talks. On the
other hand, company scientists are rewarded for reaching their individual milestone
goals, and as a result, they do not have the need to publicize their knowledge and
their discoveries outside of their immediate group of co-workers. Furthermore, in
industry, trade secrets are the properties of the company, while in academia, trade
secrets such as technical know-hows, reside in the mind of individual faculties and
researchers, and, if they are not patented, they are not the properties of the
university. Finally, in terms of organizational control, commercial companies can
133

more or less decide and control what knowledge their employees, especially their
scientists, share, whom they share knowledge with, and how they share it. In
companies, whenever their scientists want to give a presentation at academic or
industry conferences or give a seminar at the academic labs they are collaborating
with, they need to have their materials and PowerPoint slides examined and
approved by company’s legal department. Consequently, companies are often very
good at keeping secrets to themselves. Academic institutions, on the other hand,
can only “train” their scientists to share knowledge without endangering their
intellectual property, but cannot demand or force faculties to behave in a certain
way. The intellectual property manager of a public research university said,
We are not good at keeping secrets, so when things are being said to
companies and we are calling them confidential information, we do try to
put confidential marks on the information we are giving to companies, and
we try to make it known to the company that things are confidential, but it's
hard to get researchers to be diligent about maintaining good records on
what in fact they have shared with companies.

As a result, academic institutions are often considered more “open” in
knowledge sharing, or, from another point of view, are notorious for leaking
information, much to the annoyance of their technology transfer officers and IP
managers.
Subcultures of knowledge sharing
Despite the general dichotomy between openness and secrecy, one needs to pay
attention to the subtle differences within academic community as well as that
134

within the business community on people’s attitudes towards how knowledge
should be shared and how IP should be managed in the process of academic-
industry knowledge sharing. Academic scientists, technology transfer specialists,
business executives, and company scientists often have very different approaches to
knowledge sharing and IP management. In a way, they form their own subcultures
in their own communities when it comes to knowledge sharing and knowledge
management.
Technology transfer officers: Of all the major stakeholders involved in
academic-industry knowledge sharing, technology transfer officers are probably the
group most concerned with intellectual properties, since protecting and capitalizing
on university’s intellectual property is their primary responsibility. Even though
most technology transfer officers support open knowledge sharing in principle, they
often complain about the fact that many academic scientists are often either too
“carefree” to protect their own IPs or too “ignorant” about how to protect IPs.
Academic scientists: Compared to technology transfer officers, academic
scientists often have a much more relaxed attitude towards intellectual property and
knowledge sharing. Most of the academic scientists reported that they would share
“whatever needed.” An assistant professor of bioinformatics said,
If somebody asks me a question and I am consulting for them, I will just
give them the answer. I wouldn’t worry too much about who owns the IP,
because I think sometimes there is too much damage if you (want to restrict
the IP) and there is no way really to restrict the part.

135

Furthermore, academic scientists are often more trusting. They usually do not share
the hostile or defensive attitudes held by the technology transfer officers. While
being asked how much he trusted his industrial partners, a professor and former
director of a major NSF research center answered proudly, “Actually I trust them
completely.”
Business executives: Business executives fully understand the importance of
academic-industry knowledge sharing, At the same time, they tend to be very
careful about, and also skillful in, protecting the IP of their companies, given the
fierce competitions among biotech firms. When being asked about how much he is
concerned with IP when engaging in knowledge sharing with academia, the CSO of
a biotech company said,
So how much are we concerned? Yes, we are very much concerned. It [IP
infringement] always happens, because [for] the university professors, their
job is to perform research. There is no clear cut says, “here is the boundary
[and] you cannot cross over.” There is no clear cut, because research is an
overlapping effort by everyone. Now to us, we are definitely concerned. We
do have experience with certain situations. Once we agreed on certain
things, and later on, we find out that the university professors later on do
some research, which turns out to become a competitor to the original
agreement we had with the university. That happens all the time. It’s not
because the professor is unethical. [It is] because [for] university professors,
their job is to continue the research. Like I said, it’s an overlapping effort.
So that’s another concern we have, but we have to live with it. That’s the
fact. That’s life.

Company scientists: Company scientists are more constrained in knowledge
sharing than their academic counterparts because of company IP control. However,
IP protection is often something low on their priority list. Generally, company
136

scientists are very well trained in following company rules and regulations with
regard to IP protection. They tend to believe that as long as they follow companies’
rules, IP protection is something for other people to worry about. In other words,
company scientists are “passive” in protecting their IPs, or rather, company’s IPs.
One senior scientist from a major medical device company told the investigator that
he believed “internally the company needs to decide what is the portion that we can
share, and what is the portion that cannot be shared” (emphasis added). Another
biotech company scientist said,
Usually all these agreements have been written by lawyers, right? They
have thought about all kinds of possibilities already. As a scientist, I am not
a lawyer. So if the company wants me to collaborate with another scientist
from academia, I will only think in the scientific term. The company will
have legal experts and lawyers to take care of all the other legal aspects.
That’s, I think, the best way to have a successful collaboration.

At the same time, company scientists are often quite willing to share knowledge.
The same scientist quoted above told the investigator,
So in general, yeah, you don’t want to hold off any information unless the
information is not much related to your collaboration. If that is the case,
generally you don’t want to disclose. But with respect to particular
collaboration, you share as much scientific data as possible.

Variations among different academic institutions on IP management
Furthermore, to understand how knowledge is shared between academia and
industry, one needs to look at the finer distinction among different types of
academic institutions. In general, non-teaching research institutes are more careful
and more skillful in protecting their intellectual properties. They often provide
137

more training to their researchers regarding IP protection and technology
commercialization. They often have a stronger team of IP attorneys and a better
track record in commercializing and protecting their IPs. The Director of
Technology Licensing of a major medical research center said,
They know us. We are [name of the institute]. People tend not to try and
misappropriate our intellectual property, because we have a history of legal
action in that respect. We are lucky because we have an institutional history
of having some extremely successful piece of intellectual property. So
almost as an institutional rule rather than disclose something, most of our
investigators will give us a call and ask, “is it something we should be
protecting,” and then after that “should we disclose it?” We are trying to
increase their reach that they should be thinking about patent protection
before any disclosure. And I think we’ve pretty much got that message
delivered to all of our leading investigators and inventors on campus. So it
is just part of our institution, part of our policy, and part of our practice that
our investigators understand that sometimes there are things they shouldn’t
say without talking to us at first. By and large, that works extremely well.

On the other hand, not all academic institutions have this tradition. When asked to
comment on the IP management situation of her university, a technology transfer
officer from a major research university complained,
Faculties are all over the map in terms of that. There is faculty who don’t
want to talk to anybody until we file the patent application. There are other
faculties who are a lot more relaxed about that. They are not as concerned
and are more focused on how the collaboration might be of benefits.

There is also the common belief that private universities are usually more flexible
and skillful in dealing with intellectual properties and technology
commercialization. However, most of the interviewees believe that there is no
systematic difference between the ways in which public and private universities
138

deal with knowledge sharing and intellectual properties. One technology transfer
officer with extensive experience with both private and public university said,
I don’t see a lot of difference in the way that both work with universities or
with industry. I think partnering practices are very similar, and I think the
goals and expectations of the university when they participate in this kind of
relationships is relatively consistent. So I agree that the goals are very
similar, and the outcomes are very similar.

Instead, interviewees attributed the variations observed in different universities’
handling of IPs in academic-industry knowledge sharing to the characteristics and
institutional histories of individual universities.
Intraorganizational sharing vs. interorganizational sharing
Another subtle distinction emerged from the interviews is related the difference
between the ways in which academic and company scientists deal with
intraorganizational knowledge sharing and interorganizational knowledge sharing.
Given that academics may be more open, while companies tend to be more
secretive in knowledge sharing in general when they are sharing knowledge with
people from other organizations, this pattern is no longer true when these scientists
are engaged in intraorganizational knowledge sharing.
While company is often labeled as “secretive,” the sharing of knowledge
within a company is usually very open or even more open than the knowledge
sharing within a university. Because of company’s emphasis on teamwork, and
because in most of the cases, company scientists do not have any claim over the
139

intellectual property they develop, knowledge sharing is often very intensive
among scientists within a company or within a research department in the
company. However, in academic institutions, because each scientist is a separate
enterprise, the sharing of knowledge is more guarded and knowledge hoarding is
sometimes more frequent. The CEO of a major private medical research institute
commented,
Well, interesting enough, academic culture ideally is supposed to be a lot of
sharing of ideas and approaches. For whatever reason, I think there are a lot
of reasons for this, the reality is that there is less information sharing
culturally in academic organizations than there needs to be to move science
along. And I've talked to people both in industry and academics about this.
In industry internally within a particular company, there is much more
information sharing than there is back at the academic organization because
everyone has a common interest in seeing a good product developed. There
isn't that common interest in academics. It's much more fragmented. The
self- interest is much more fragmented academics than it is in a company.
So the information sharing, and it's just the reverse of what you might think
it would be in some sense. Now clearly companies are extremely careful
about sharing their own proprietary information with other companies, but
within the company itself, there is more of an openness.
Summary
Professional culture is indeed an important factor in academic-industry knowledge
sharing. Interviews with four panels of professionals who are heavily involved in
academic-industry knowledge sharing, i.e. company executives, company
scientists, academic scientists, and university technology transfer specialists, were
employed to examine how the culture of academia and the culture of industry affect
academic-industry knowledge sharing.
140

First, despite of the traditional difference between the academic culture and
the industry culture, a tendency towards convergence has been observed.
Commercial companies are putting increasingly more emphasis on basic research
and scientific excellence, while academic institutions such as universities are
becoming more encouraging towards applied research and commercialization of
academic discoveries.
Second, while in the past, academia is often celebrated for its openness in
spreading knowledge and industry is often labeled as being secretive, the general
dichotomy of “open” vs. “secretive” has been proven to be too simplistic and even
outdated in the rapidly changing world. Academic institutions and commercial
companies share many commonalities in their management of knowledge and
intellectual properties. Furthermore, a comparison of the different opinions and
behaviors of company executives, company scientists, academic scientists, and
university technology transfer specialists reveal that the so-called academic culture
and industry culture is actually differentiated among themselves, respectively. This
is especially true within the academic world where academic scientists and
technology transfer specialists have very different attitudes towards how academic
knowledge should be managed and how academic-industry knowledge sharing
should be conducted. Finally, interviews suggest that the openness of academia and
the secrecy of industry are only relative. Interestingly, industry is more open than
academia in intraorganizational knowledge sharing. Scientists within a company
141

share knowledge openly and freely because of the institutional incentive and
personal reward systems in the industry and because of the teamwork spirit, which
is very prominent in companies. On the other hand, the sharing of knowledge
within an academic institution sometimes turns out to be less frequent and less open
because PIs are individual entities and sometimes they are in a competitive
relationship with each other.
RQ4: What is the role of informal knowledge sharing in academic-industry
knowledge sharing in biotech industry?
Existing literature has pointed out the importance of informal knowledge sharing in
the development of biotech industry (e.g., Zucker, Darby, & Armstrong, 1998;
Zucker, Darby, & Brewer, 1998). However, most of the current studies on
academic-industry knowledge have been focused on the knowledge sharing through
formal channels by studying the patterns of patent licensing (e.g. Bercovitz et al.,
2001; Jensen & Thursby, 1998; Ratliff, 2003; Thursby et al., 2001; Thursby &
Thursby, 2000, 2001), publication citation, patent citation (Gittelman & Kogut,
2003; Jaffe & Trajtenberg, 1999; Jaffe, Trajtenberg, & Henderson, 1993) and
strategic alliances (Eisenhardt & Schoomhoven, 1996). There is a lack of empirical
research on knowledge sharing through informal channels. This could be partly
attributed to the lack of empirical data on the subject, since it is often hard to
operationalize and measure informal knowledge sharing using the traditional social
scientific quantitative method.
142

As discussed in RQ2, the knowledge shared through informal channels
tends to be less articulated and of no immediate market value, including technical
know-hows, research techniques, general understanding of the science and market,
etc. Even though knowledge shared through informal channels do not lead to
immediate benefits, it is nevertheless essential to the establishment of more formal
contractual relationships and to the development of science in general. The CEO of
a biotech company said,
My guess is that the predominant channel [of academic-industry knowledge
sharing] is through personal or informal communication. I think technology
transfer offices or patent licensing offices are probably not the most
effective setting for alliance, because it takes a more broad approach setting
up alliances than sending out huge list of available technologies out to
companies and to investors, which are really hard to deal with. I think it is
most effective when either the scientist or the company or the investor
knows what they are looking for or knows who would be interested in it and
has relationships that help to facilitate that.

Figure 4.4 presents the conceptual network of informal knowledge shared derived
from the content analysis.
143

Figure 4-4. Informal knowledge sharing

Goals of informal knowledge sharing
Different stakeholders involved in academic-industry knowledge sharing have
widely different ideas about informal knowledge sharing. Generally speaking,
academic scientists and biotech company executives emphasize the importance of
informal knowledge sharing and interaction, while university technology transfer
people sometimes are not very enthusiastic about informal knowledge sharing due
to their concerns of IP protection. Finally, company scientists are generally not for
or against informal knowledge sharing.
144

Informal knowledge sharing serves two purposes: a technical one and a
relational one. First, informal knowledge sharing promotes both basic research and
product development by enabling academia and industry to utilize each other’s
technical expertise. Secondly, informal knowledge sharing facilitates the building
of relationship and social networks.
Interviewees, especially industry executives, emphasize the importance of
having informal interactions before going into formal relationships of knowledge
transfer. They predominantly see informal knowledge sharing (through channels
such as personal communication) as a precursor to formal contractual technology
transfer, through channels such as patent licensing. Informal knowledge sharing
enables companies to identify the knowledge they want to acquire and potential
partners they want to collaborate with. A serial entrepreneur who had launched
several successful biotech companies argued that one of the major causes of failed
technology transfer is
A failure to spend time at the beginning of the relationship, listening and
hearing what each of the parties want out of the relationship. I think
sometimes people go into a relationship from both sides, assuming that
they’ve already known what the other party want, and depending on the
company, depending on the university, their needs can be very different.
They may be looking for money. They may be looking for fame. They may
be looking to run a spin-off business. There are all kinds of things that press
people’s buttons ... They need more listening.

Technology transfer officers support informal knowledge sharing for the
same reason. One technology transfer officer said, “the biggest foundation of
145

relationship is personal or informal communication. It all comes down to the people
in the end.”
Interviewees use some very interesting metaphors to capture the importance
of informal interaction and knowledge sharing prior to entering formal
relationships. A senior technology transfer officer used the metaphor of dating to
illustrate the relationship between informal knowledge and formal contractual
knowledge sharing by saying,
So for a lack of better term, the informal sharing mechanism is more like
“dating,” until you find somebody you say, oh Jesus, this is a relationship I
like to formalize. Then you go into a formal channel (of knowledge sharing).

Another serial entrepreneur who used to head a knowledge broking firm called this
kind of informal knowledge sharing “kissing the frog.” He said,
I call it “kissing frog.” It sounds rude, but you have to kiss a lot of frogs to
find the prince or princess. This is all about kissing frogs. You’ve got to be
polite, form close relationship, network as much as you can, that will make
the deal doing much easier in the end.

Both of these metaphors emphasize the importance of informal communication
before two parties entering into a formal contractual relationship. They also imply a
trial-and-error process in identifying the right partner to share knowledge with.
Academic scientists value informal knowledge sharing for different reasons. They
often see it as an opportunity to discuss the latest research and ideas with their
colleagues in industry. A professor said,

Well, just investigator to investigator, the biotech investigator to university
investigator. When I went out to the company, we sat down around the table
146

and they showed their slides, I showed our slides, we discuss our
technologies, and just investigator to investigator from the biotech to
academic person.

However, technology transfer officers often see informal knowledge sharing as
useful yet potentially dangerous because it can lead to information leaking and IP
infringement. When asked to rate the importance of informal knowledge sharing on
a scale of 1-5, the CEO of a private research institute, who was trained as a IP
lawyer, only gave it a 2. He explained his decision by saying,
I think informal sharing has some value, but I think that formalizing some
of these relationships from a legal perspective would actually be a good
idea. I don't think that inhibits knowledge. I think that actually creates a
better environment for the sharing of knowledge. So that people know there
is contractual bond that keeps things confidential, people know that there is
a patent application is being applied for in a certain area. I would say that
more formalized relationships actually promote dialogue.
Trust and informal knowledge sharing
Trust has been both theoretically and empirically proved to be a determinant of the
effectiveness of knowledge sharing (Szulanski, Cappetta, & Jensen, 2004). On the
positive side, trust is found to increase the effectiveness of knowledge sharing by
increasing the amount of information shared (Carley, 1991; Tsai & Ghoshal, 1998),
decreasing the cost of exchange (Curall & Judge, 1995), increasing the “voluntary
deference to authority” of the source of knowledge (Kramer, 1999; Tyler &
Degoey, 1996), and facilitating the institutionalization of innovations (Kostova &
Roth, 2002). On the negative side, trust can decrease the effectiveness of
knowledge sharing by distracting attention from the actual content of the
147

knowledge (Allen & Stiff, 1989), decreasing the attentiveness in communication,
and lessoning the diversity of thoughts and actions (Webb, 1996).
The interviews revealed a deeply engrained distrust between academia and
industry and the paramount importance of building trust in successful academic-
industry knowledge sharing. In most cases, distrust is a result of unpleasant past
experiences and IP infringement. Interviewees from both academia and industry tell
stories of betrayal and infringement. The CEO of a biotech-consulting firm said,
There is outright thievery. I can give you a list that we can spend the rest of
the day on the phone with. There is a famous story about a company where
somebody spent some time in a laboratory. The scientist thought the
discussions that he had with the other scientist were in a public domain and
the second scientist went back and start a company, and the company goes
worth several hundred million dollars in a few years. What he did was he
spent a summer in somebody’s lab and took the stuff and patented it, never
giving the first guy a share of it. It goes on everyday.

A technology transfer officer said half jokingly, “We trust them [companies] as far
as we can throw them.”
The distrust between the two communities was identified by interviewees as
one of the major barriers to academic-industry knowledge sharing. It leads to
knowledge hoarding and prevents the effective communication of expectations in
academic-industry knowledge sharing. A technology transfer officer pointed out:
I think the barrier [in academic-industry knowledge sharing] is more like
the level of trust, whether the company trusts us enough. As I mentioned, if
knowledge sharing is one directional, it never satisfies anybody. There is an
old saying, “you don’t get the right answer unless you ask the right
question.” Sometimes companies worry that just by asking questions, they
may let their competitors know what they are up to. So there is some
148

concern, and it will negatively affect the two-way information sharing. If
we don’t get a right question, we can’t give a good answer.

Despite of the pervasive distrust between academia and industry, interviewees all
believed that trust is key to successful academic industry knowledge sharing. One
CEO said,
There …needs to be a trust, a belief that this particular company or this
particular institution can deliver on the promise so that the value that's
placed on it should be one that is realistic.

Two types of trust in the process of academic-industry knowledge sharing are
identified by the interviewees: trust based on formal contracts protected by legal
means and trust based on relationship and past experience. Trust based on formal
contracts is built on the belief that the other party will refrain from behaving
unethically due to the constraints of the contracts. One such contract is non-
disclosure agreement or confidentiality agreement. When interviewees, especially
those from technology transfer offices, talked about trust, they often discussed
about how these legal contracts could protect the legal rights of both universities
and companies. One technology transfer officer said,
So there needs to be a certain amount of trust amongst universities and
industry, but that is often times secured in written documents, binding
people not to discuss their information outside of their agreed-to channels.

The second type of trust is more intuitive and built based on past experience of
interacting with a person or an organization. This is more important when people
149

share knowledge informally in the absence of any legal agreement. One senior
scientist at a biotech company said,
So trust is very important here. Trust is built on long-time relationship. You
start your collaboration with some common projects, and trust is built
gradually in the course of collaboration. Later, you start to collaborate with
them on more important projects. The company is unlikely to give an
important to a totally strange lab.

Interestingly this way of categorizing trust resembles theory of trust proposed by
Paul Adler (2001) in his influential paper “Market, hierarchy, and trust: The
knowledge economy and the future of capitalism.” According to Adler, there are
three sources of trust: past experience, calculation of interest, and norms. First,
repeated experience in the past can lead to trust or distrust. Secondly, trust can be
built upon an estimation of the risk of being taken advantage by the other party.
Finally, trust is a result of professional norms and the belief that other will behave
ethically according to these norms (Adler, 2001; Liebeskind et al., 1998). It is clear
that the two types of trust identified in this study correspond to the first two sources
of trust discussed by Adler (2000).
However, the trust identified in the process of academic-industry
knowledge sharing lacks the third type of trust identified by Adler, one that is based
on professional norms and values. One technology transfer officer identified this
absence as a barrier to academic-industry knowledge sharing,
I wish there were some generally accepted principles that we could pass
things back and forth. In the academic community, we basically do that
150

already. [However] that interface between the company and the
university … isn’t working very well yet.
Communication technology and informal knowledge sharing
New communication technologies have made communication between the
academic and business communities easier than ever. Most of the interviewees
identify that they use email, Internet, and telephone extensively. Other
communication technologies used include teleconferencing and video conferencing.
All interviewees applauded the advancement of new communication technologies
and how it has transformed the ways in which they work and interact with others.
Despite of the convenience brought by these communication technologies,
most of the interviewees argue that no matter how comfortable they are with these
technologies, these new communication technologies cannot replace the need for
face-to-face communication. They point that face-to-face communication is
essential in the building of a trusting relationship among different parties involved
in academic industry knowledge sharing. Even a brief personal contact will be
much more effective than the exchange of many emails. However, once the
relationship has been established, emails and telephone will be very cost-effective
tools in interacting and communicating amongst people. There would not be much
difference between face-to-face interaction and mediated communication. The CEO
of a biotech company said,
But that being said, we would probably never make an investment without
having met people in person. So even through all of those things are very
efficient, and we use them all the time, we still will never rely completely
151

on them. We’ll always want to meet people in person, and have the
opportunity to spend a lot of time getting to know people and talk to people,
and that always requires meeting in person.

At the same time, when two parties have had previous interaction in the
past, there is already a trusting bond between them. When they want to start
a new collaboration, the need for face-to-face interaction is no longer that
paramount.
Conclusion and limitation
Summary
This chapter presented the results of an interview study of academic-industry
knowledge sharing in the biotech sector. Through semi-structured interview,
supplemented by ethnographic observation, this chapter provided insights into the
attitudes, behaviors, and perceptions of academic scientists, technology transfer
specialists, biotech business executives, and company scientists on academic-
industry knowledge sharing.
The knowledge shared between academia and industry includes a wide
variety of information, including, but not limited to, scientific discoveries and
inventions, technical know-hows, understanding of the industry, research
techniques, etc. The interview study discovered that the knowledge shared between
academia and industry should be understood along two dimensions: articulatedness
and market readiness. The knowledge shared between academia and biotech
companies are protected as different types of intellectual properties: copyright,
patent, and trade secret.
152

Interviewees reported using a wide variety of channels in academic industry
knowledge sharing, including patent licensing, research collaboration, publication,
conferences and seminars, consulting, spin-off, research contracts, research grants,
material and data transfer agreements, invited talks and visits, educational and
training programs, specially industry-oriented research centers, knowledge brokers,
hiring, and personal communication. While existing literature mainly focuses on
publication, patent licensing, and spin-off, this study emphasizes the importance of
research collaboration as a channel of knowledge sharing. Furthermore, the study
identified the important role of knowledge broker as a third party that bridges the
gap between academia and industry.
Interviews further identified that these academic-industry knowledge-
sharing channels could be understood on a continuum from highly formal to high
informal. The adoption of a certain channel depends on the two characteristics of
the knowledge: articulatedness and market readiness. Interviewees also reported
that the simultaneous use of multiple channels often leads to better outcomes:
success knowledge sharing, successful product development, and
commercialization.
The interview study compared the similarities and differences between
academic culture and industry culture and discussed how these cultural
characteristics affect the motivation, process, and outcome of academic industry
knowledge sharing. Interviewees often use pairs of dialectics to describe the
153

differences between academic culture and industry culture. In terms of incentives,
academia is often science driven while industry is often bottom line driven. The
culture of academic research is characterized as curiosity-driven, rich in distraction,
enjoying academic freedom, and typically of low speed. The other of industry, on
the other hand, is characterized by a set of contrasting characteristics: object-
driven, highly concentrated, highly disciplined, and very fast. In terms of the
culture of knowledge sharing, academia is often considered open while industry is
often labeled as secretive.
Even though these dichotomies offer a shortcut in understanding the
difference of academic culture and industry, and how it affects academic-industry
knowledge sharing, they are often too simplistic to offer a comprehensive picture.
The past two decades have seen a convergence between the academic culture and
the industry culture. On the one hand, biotech industry is becoming increasingly
science-driven and interested in basic research. On the other hand, academia is
becoming more and more industry-friendly and proactive in establishing
collaborative ties with industry.
The changing characteristics of the academic and industry further
challenges the simplistic dichotomy of “open vs. secretive” in academic-industry
knowledge sharing. First, the interviews suggested that it is problematic to consider
academia or industry has each having an integral culture. Among academics,
faculties and technology transfer specialists are completely against each other on
154

certain issues of knowledge sharing and IP management. For instance, for faculties,
success in academic-industry knowledge sharing is measured by whether a useful
product is developed and how much research funding the faculty can get for his lab.
On the contrary, the success of technology transfer offices is measured by the
number of patent licensed and the amount of royalty income generated (Thursby et
al., 2001). As a result, the two have very different agenda and strategies in their
interaction with industry. Similarly, significant differences can be observed
between the attitudes and the practices of business executives and company
scientists. Consequently, a comparison of the narratives given by academic
technology transfer officers and biotech business executives reveals more
similarities than differences in their attitudes towards IP ownership, IP
management, and knowledge sharing.
Secondly, there is great internal variation among academic institutions in
their handling of intellectual property and academic-industry knowledge sharing.
Some universities are more open, while others are more careful in guarding their
intellectual properties. In general, non-teaching research institutes are more
cautious and effective in protecting their IPs in academic-industry knowledge
sharing, and thus often considered more secretive than universities. However, no
systematic difference between public and private universities was identified.
Thirdly, the dichotomy of openness and secrecy is even more dubious when
one considers the differences in their attitudes towards knowledge sharing or their
155

openness between interorganizational and intraorganizational knowledge sharing.
More specifically, while companies are often times labeled as secretive in sharing
knowledge with academia or with other companies, the intraorganizational
knowledge sharing within one company is amazingly open. Knowledge sharing is
rather free and intensive within a company, because of the shared goal of the
company, and because of the company ownership of the IP. On the contrary, the
intraorganizational knowledge sharing within a university might not be as free as
one might think, because of the independence of each academic labs, and because
of the sometimes conflicting agendas of faculties.
Finally, informal knowledge sharing was identified to be one indispensable
part of academic-industry knowledge sharing. Informal knowledge sharing fulfills
two important goals: a social goal and a technical goal. First, informal knowledge
sharing facilitates the building of relationship and social network. Secondly,
informal knowledge sharing often facilitates the sharing of tacit knowledge that is
not very articulated such as technical know-hows. Trust is a central feature in
informal knowledge sharing and relationship building.
Reliability and validity
Because the analysis of data collected through interviews is often interpretative in
nature, the reliability of interview studies largely base on the existence of tapes and
transcripts. In this study, all interviews were recorded whenever permitted by the
156

interviewees, and all recorded interviews were thoroughly transcribed, which,
according to Perakyla (1997), can ensure the reliability of interview studies.
In evaluating the validity of interview studies, we assume that there is a gap
between what is observed and what these observations stand for (Perakyla, 1997).
To ensure the validity of interview studies, I compared the results of my interview
studies to existing literature and industry documents in the process of data analysis.
To further ensure the validity of the results of the interview study, triangulation will
be used. Triangulation refers to the simultaneous use of several research methods
and to compare the respective results originating from those methods to see if they
reach a coherent conclusion. Martin (1992) proposed the use of "hybrid
methodology" in the study of organizational communication. To that end, the next
chapter will be devoted to a network analysis of the academic-industry knowledge
sharing networks.
157

Chapter 5. A theory of knowledge networks
The previous two chapters provided an in-depth interpretative analysis of
academic-industry knowledge sharing as a special form of interorganizational
knowledge sharing. The interview study identified two very different modes in
which knowledge is shared across organizational borders: interaction-based
knowledge sharing and transaction-based knowledge sharing. Informal knowledge
sharing, characterized by long-term informal interaction and communication, plays
a prominent role in the interaction-based knowledge sharing.
This chapter will start with a brief review of current research on
interorganizational knowledge networks. It will then present a theory of knowledge
networks and hypotheses based on the findings of both the interview study and the
existing literature. Then it will introduce the data and statistical methods used in
testing these hypotheses. The next chapter will be devoted to presenting and
discussing the results of the network analysis.
Knowledge sharing in interorganizational networks
An open system view of organizations suggests that in order to survive and succeed,
an organization needs to effectively manage its relationship with its environment
and the other organizations in it (Scott, 1992; Thompson, 1967). While in the past
researchers have examined the exchange of financial resources and products
between organizations and their environments, lately increasing attention has been
paid to understanding organizations’ exchange of ideas, practices, information, and
158

technologies. Or in other words, knowledge of the environment (e.g. Baum &
Ingram, 2002). Researchers and managers alike have come to the realization that
organizations need to effectively manage their knowledge-sharing ties with other
organizations in the environment in order to increase their competitiveness in
today’s knowledge economy.
Interorganizational networks not only allow the sharing of existing
knowledge but also facilitate the creation of new knowledge. In knowledge-
intensive industries such as the biotech industry or the semiconductor industry,
whose knowledge base is expanding and changing very quickly and in a way that is
described as “competence-destroying,” the locus of innovation is often the network
of organizations instead of any individual organization (Podolny, Hannen, & Stuart,
1996; Powell et al., 1996).
Network analysis has been used to study the relationships among
organizations and the flow of knowledge across organizational borders since the
1970s. The concept of a network could be a metaphor, a method, and/or a theory
(Barnes, 1979; Monge & Contractor, 2003). Looking at interorganizational
knowledge sharing as networks allows for an understanding of the knowledge flow
structure of a particular industry on both micro and macro levels. A variety of
interorganizational networks that allow the sharing of knowledge have been studied
in the literature including: strategic alliance networks, interlocking directorate
networks, supply networks, citation networks, and co-authorship networks.
159

Strategic alliance network
A strategic alliance network is probably one of the most frequently studied
interorganizational networks (e.g. Arora & Gambardella, 1990; 2002; Hagedoorn &
Schankenraad, 1994; Shan, Walker, & Kogut, 1994; Stuart, 2000). Organizations
build strategic alliances for both strategic and social purposes (Eisenhardt &
Schoomhoven, 1996). First, strategic alliances can help companies acquire new
technology and knowledge, which in turn, gives the company a competitive
advantage and enhances its economic performance. Strategic alliances among firms
also increase the likelihood of informal knowledge sharing and knowledge
spillovers (Saxenian, 1994). Second, establishing strategic alliances gives
companies more social capital (Stuart, 2000).
Many empirical studies confirmed that strategic alliances have a positive
effect on companies’ innovative output
10
. For instance, Shan, Walker, and Kogut
(1994) found a positive relationship between a company’s embeddedness in the
strategic alliance network and the number of patents it has. Stuart (2000) studied
interorganizational technology alliances in the semiconductor industry and found
that building technology alliances with large and innovative partners will increase
the innovation rate and economic performance of a company. Such technological
alliances are especially important to young and small companies. Powell et al.

10
However, Hagedoorn and Schankenraad (1994) found that it is the utilization of the technology
acquired or developed in technology alliances, rather than a simple involvement in the network that
has a positive effect on the economic performance of a company.
160

(1996) studied the relationship between company’s centrality and its performance
and found that prior centrality in a strategic alliance network is a good predictor of
subsequent R&D collaboration and company growth. Building on the work of Shan
et al. (1994) and Powell et al. (1996), Ahuja (2000a) conducted a longitudinal study
of how the characteristics of a firm’s ego collaboration network affects its
innovation output and found that direct ties lead to higher innovation outputs.
Interlocking directorate network
Interlocking directorate networks are created by “linkages among people who serve
on multiple corporate boards” (Monge & Contractor, 2003, p. 216).An interlocking
directorate network is another type of interorganizational network—one built to
manage interorganizational dependency, legitimacy and resource sharing (Mizruchi,
1996). It enables the sharing of knowledge, corporate practices, and strategies
across organizational borders (Haunschild, 1993). Interlocking directorate networks
have been heavily researched due to the availability of highly reliable data since the
1970s (Mizruchi, 1996).
The structure of an interlocking directorate network is important because it
shapes the direction of interorganizational knowledge flow. Mizruchi (1982)
studied American interlock networks between 1904 and 1974 and found that a
corporate elite network is a small world in that virtually all nodes are reachable
with less than 4 steps. Davis, Yoo, and Baker (2003) conducted a similar study on
an interlocking network between 1982-2001 and discovered that despite the major
161

corporate restructuring in the past few decades, the structure of the interlocking
network has remained constant. The high connectivity of corporate elite networks
allows the quick diffusion of new practices and strategies among organizations.
Supply-chain network
A supply-chain network consists of one focal organization and a group of other
organizations that provided products, materials, and other resources to the focal
organization. Noticing that Toyota’s production network enables faster knowledge
diffusion than those of other major auto makers, Dyer and Nobeoka (2000)
conducted an interpretive study of the supply-chain network of Toyota through
interviews. They found that Toyota successfully created a “network identity” and
propagated the notion that knowledge does not belong to any single company but
the network as one entity. This decreases the transaction costs of knowledge
sharing, which partly explains the success of the Toyota network.
Citation networks
Even though citation networks and co-authorship networks are not
interorganizational networks per se, they provide a useful way of mapping the flow
of knowledge in the scientific and business communities across organizational
borders.
Paper citation networks represent the flow of knowledge in the scientific
community and illustrate the developmental trajectory of scientific ideas, while
162

patent citation networks map the flow of knowledge in industry communication and
demonstrate the progress of technologies. Patterns of patent citation show clear
signs of knowledge spillover in that patents from the same geographic location are
more likely to cite each other, controlling for the existing concentration of patent
activity (Jaffe & Trajtenberg, 1999; Jaffe et al., 1993).
Podolny et al. (1996) proposed that innovation does not develop within
individual organizations, as previously stated by Schumpeter (1947). Instead,
innovation develops in its own technological niche. Whether an innovation is going
to lead to a technological breakthrough or a dead end is a function of the
crowdedness of the niche and the status of the company, which they define in terms
of the company’s position in the citation networks.
In comparing patent citation and paper citation networks, Gittelman and
Kogut (2003) found that these two networks follow different evolutionary logics:
papers written by famous scientists are most likely to get cited but patents with high
technical richness and market impact are more likely to get cited. Highly
influential papers and highly influential patents do not overlap.
Similarly, asking the question, “does science and technology co-evolve,”
Murray (2002) compared the citation patterns of papers and patents regarding a
particular technology: tissue engineering. She discovered that the science network
(operationalized as paper citation network) and the technology network
(operationalized as patent citation network) are distinct from each other. The
163

overlap between the science and technology is through spillover, spin-off, patent
licensing, and consulting instead of through co-authoring or citation.
Co-authorship network
Co-authorship is often used as an indicator of an organization’s embeddedness in
the scientific community as well as the developmental trajectory of a particular
discipline or industry. For instance, Newman (2001) studied the co-authorship
network of several scientific disciplines, including biomedical, physics, and
computer science between 1995-1999 and found that the scientific co-authorship
network is a small world and follows the power law distribution (Barabási & Albert,
1999). In other words, any pair of scientists can reach each other through less than
6 steps and the network is characterized a small number of highly connected nodes
and a very large number of lowly connected nodes. Cross-disciplinary differences
are also discovered. For instance, nodes in the experimental high-energy physics
network have a higher average degree than nodes in other networks, because the
former requires the collaboration of a large number of people. Another interesting
find is that the biomedical network is not as centralized as other networks. In other
words, the biomedical network is characterized by many nodes with a low degree
of connectedness while other scientific networks are characterized by few nodes
with higher degree of connectedness. Finally, the biomedical network has lower
transitivity than other scientific networks such as the networks of physics and
164

computer science, i.e., two scientists who both collaborate with a third scientist are
not likely to co-author with each other.
Critique
The existing literature on interorganizational knowledge networks is flawed by a
split between network structure and network content. In reviewing the network
paradigm in organizational research, Borgatti and Foster (2003) categorized
network studies into two broad groups: studies of network “topologies” and studies
of network “flows” (p. 1002). The former focuses on the structural patterns of
interconnection, while the latter considers network content, or the resources that
flow through networks. Current network studies have overemphasized network
structure without paying enough attention to what is actually flowing in the
network (Smith-Doerr & Powell, 2005; Stinchcombe, 1990). As a result, they tend
to treat all ties as comparable.
While researchers generally acknowledge the complex nature of knowledge
shared in interorganizational networks, most of the empirical studies on
interorganizational knowledge sharing fail to specify the characteristics of the
knowledge flowing in different types of networks. A few studies that did identify
the characteristics of knowledge in a specific network have a rather restrictive
understanding of knowledge. This dissertation investigates all of these elements,
including network structure, network content, and node attributes, as all of them are
expected to affect network evolution and knowledge development.
165

One way to combine the study of network content with that of network
structure is to look at different types of content flowing through different networks.
Organizations participate in different networks to acquire different types of
resources and different kinds of knowledge. These differences in motivation,
resources, and knowledge sharing process create dissimilar structural signatures at
the dyad, triad, and global levels. For instance, Laumann, Galaskiewicz, and
Marsden (1978) emphasized “relation-specific structures” (p. 463), where links of
one type between organizations do not necessarily imply another type of link.
Lazega and Pattison (1999) also suggested that “the precise form of these
interdependencies is likely to depend on the types of resources involved, since
transfers or exchanges of each type are likely to be subject to different constraints.”
(p. 75). More recently, Contractor, Wasserman, and Faust (2006) argued that if
resource dimensions in multiple networks differ, corresponding tie formation
mechanisms and network structures will differ.
On the other hand, some researchers argue that organizations are situated in
an organizational environment comprised of multidimensional ties overlaid on each
other (e.g. Baum & Ingram, 2002; Gulati & Garguilo, 1999). When the same sets of
nodes are connected by multiple relations, one network of interaction becomes the
context for the set of other networks. A central proposition from this view is that
multiple types of network ties are interdependent on each other (Robins & Pattison,
2006), and ties in one network influence the formation or dissolution of ties in other
166

networks (Galaskiewicz & Bielefeld, 1998; Lomi, 1997; Lorrain & White, 1971).
An earlier account of the multiplexity of relations has been raised by Aldrich
(1979), who suggested that multiplexity or the redundancy of relations increases
the stability of relationships. The interdependency of multiple networks has been
articulated in the theory of embeddedness (Granovetter, 1973, 1985), which asserts
that economic relations are influenced by social structural and personal relations.
More specifically, structural embeddedness focuses on the way the structure of an
organizational network and an organization’s structural position influences
economic outcomes via its influences on opportunities for new social alliances and
networks (Gulati & Garguilo, 1999; Uzzi, 1996, 1997). For instance, Uzzi (1996;
Uzzi, 1997) identified three prototypes of network structures: underembedded
arm’s-length network, integrated network and overembedded network. A company
is underembedded when all of its first-order and second-order ties are arm’s length
ties, and it is overembedded when both its first-order and second-order ties are
embedded ties. The ideal situation is when the company has an integrated network
in which its first-order ties are mainly embedded ties and its second-order ties
consist of both arm’s-length and embedded ties. Empirical studies have supported
such an inverted U shape relationship between a company’s embeddedness and its
economic performance (Uzzi, 1996, 1997).
These two approaches do not necessarily contradict each other. Instead, they
complement each other by explaining the development of multiple
167

interorganizational networks. More specifically, organizations engage in multiple
networks to acquire a variety of resources flowing in these networks. As a result,
these networks have different patterns of growth. At the same time, because
organizations are embedded in multiple relations, two organizations that have one
type of ties are more likely to have other types of ties.
Theoretical development: Transactional vs. interactive knowledge networks
According to resource dependency theory, proposed by Pfeffer and Salancik
(1978), organizations participate in interorganizational relations to acquire
resources they need in order to survive and to manage their dependency on the
environment. Resource dependency theory considers the emergence of two major
categories of interorganizational relations: those based on resource exchange and
those based on the interpretation of organizational boundaries (Laumann et al.,
1978).
Two forces drive the formation of interorganizational relations: competition
and cooperation. Competition is the logic of a perfect market, while cooperation is
based on collective purpose. As discussed in Chapter 2, organizations engage in
market exchange with each other. During this process, each individual organization
is assumed to be independent, self-interested, and in pursuit of autonomy. The
relationship between organizations is characterized by distrust and opportunism.
These organizations share little, if any, collective goals. However, this competition
logic does not explain the formation of all interorganizational relations. Sometimes,
168

organizations form interorganizational relations driven by the logic of cooperation,
when they have a high degree of collective orientation.
The forces of competition and cooperation lead to the formation of two
distinctive types of interorganizational relations: those based on resource exchange
and those based on the interpretation of organizational boundaries (Laumann et al.,
1978). Resource-based links allow the transfer of highly specific resources such as
funds, personnel, and technology as well as more general resources such as power
and influence (Laumann et al., 1978). However, beyond the goal of resource
transfer, organizations build ties with each other when they have shared goals and
when there is a need to reduce the high degree of uncertainty found in transactional
relations (Williamson, 1979). This leads to the formation of interorganizational ties
based on “the interpretation of organizational boundaries” (Laumann et al, 1978, p.
465). In building such ties, organizations give up certain level of autonomy in
exchange for reduced uncertainty and the achievement of collective goals.
Laumann et al. (1978) categorized such relations into three groups based on the
different levels of involvement among organizations. First, organizations can have
common membership. For instance, by joining a trade association, an organization
can participate in an interorganizational community with minimum level of
involvement and loss of autonomy. Or organizations can have shared membership,
for instance, through interlocking directorates, which implies a higher degree of
169

interdependence. Finally, organizations can engage in joint programs, which imply
an even higher level of involvement and the surrender of more autonomy.
Applying the distinction between exchange-based interorganizational ties
and solidarity-based ties to the problem of interorganizational knowledge sharing
leads to the identification of modes of interorganizational knowledge sharing:
knowledge transfer and knowledge sharing; and two different types of
interorganizational knowledge networks: transactional knowledge networks and
interactive knowledge networks.
The transaction knowledge network involves a one-way transfer of
knowledge from one organization to another through formal, contractual channels
such as patent licensing, material transfer, or publication citation. Such a network is
called “transactional knowledge network,” because knowledge sharing in this kind
of network is mainly based on short-term exchange-based transactions. Most of the
existing studies on interorganizational knowledge networks have been focused on
different types of transaction-based networks (Owen-Smith & Powell, 2004). The
interactive knowledge network, on the other hand, allows for the two-way sharing
of knowledge through informal, long-term interaction and communication. This
type of network is called an interactive knowledge network because it is based on
long-term interactive ties, such as strategic alliances, interlocking directorates, and
research collaborations.
170

These two modes of knowledge sharing complement each other (Powell,
1998). For instance, Liebeskind et al. (1996) examined the knowledge seeking
practices of two very successful biotech companies and discovered that company
scientists acquired knowledge from outside sources through both transactional and
interactive ties. What’s more, it was found that company scientists obtained a lot
more knowledge through informal collaborative ties than through formal
contractual market exchanges.
The transactional knowledge network and interactive knowledge network
differ in terms of the characteristics of the knowledge sharing process and network
structure. What’s more, they grow according to different evolutionary logics. The
subsequent part of this chapter will be devoted to a detailed discussion of the
characteristics of these two networks and several hypotheses derived from their
theoretical construction.
Characteristics of the knowledge sharing process
The knowledge sharing processes in these two knowledge networks differ in terms
of the nature of the knowledge shared, the nature of ties, and the outcome of the
knowledge sharing process. A summary of these differences is presented in Table
5-1.
171

Nature of the knowledge shared
Articulated and migratory knowledge is also shared in the transactional knowledge
network. Such knowledge is often well codified and individually packaged, and is
not embedded in a specific context. Such knowledge is often of low developmental
level, mainly data or information. Examples of this kind of knowledge shared
across organizational borders could include patents, publications, proprietary
materials, as well as data. This kind of knowledge is often low in “stickiness”
(Szulanski, 2003). In other words, such knowledge can be transferred with relative
ease on formal contractual basis, such as patent licensing, publication citation,
material and data transfer agreements.
An interactive knowledge network is more often related to the sharing of
embedded, less articulated, and highly contextualized knowledge—for instance,
technical know-how and market know-how. Very often knowledge that has a high
developmental level can be shared in the interactive knowledge network. This kind
of knowledge is often highly sticky and can only be shared through more
interactive channels. Such knowledge networks are interactive because in most
cases, knowledge sharing that takes place between two organizations is a long term,
two-way process that includes of the exchange of ideas, information, knowledge, as
well as on-going discussions and interactions. More importantly, an interactive
knowledge network not only enables the sharing of knowledge among different
172

organizations but also provides a fertile ground for knowledge creation and
innovation.
Characteristics of ties
The transactional knowledge network is based on transactional ties that are formal,
contractual, based on arm’s length exchange, and exploitative in nature. Examples
of such ties include patent licensing, data transfer, and material transfer. On the
other hand, the interorganizational knowledge sharing in the interactive network
occurs through interactive ties that are collaborative, exploratory, long-term, and
sometimes, informal. Examples of these interactive ties include: personal
communication, consulting, interlocking directorate, and research collaboration.
Arm’s length tie vs. embedded tie: Arm’s length ties are the venues of
exchanges in ideal, atomistic markets (Hirschman, 1970). According to classic
economics theory, the rational individual who is self-interested and seeking profit
engages in arm’s length transactions with others. Arm’s-length ties are impersonal
and loose (Uzzi, 1996). On the other hand, there is a different kind of tie: the
embedded tie. While an arm’s-length tie is motivated by the selfish pursuit of
interests, an embedded tie is characterized by trust and reciprocity (Powell, 1990;
Uzzi, 1996).
The transactional knowledge network is characterized by arm’s length ties.
Organizations engage in transactional knowledge networks to build such ties to
share knowledge in exchange for monetary reward. Their interactions are guided by
173

the rule of market exchange in that each party seeks to maximize its own interest.
In contrast, the interactive knowledge network features embedded ties that are
based on the logic of community instead of the market and characterized by
collaboration instead of competition.
Exploitative ties vs. exploratory ties: March (1991) identified two distinct
ways in which organizations learn: exploration and exploitation. According to his
theory, exploration means trying out with new alternatives and exploitation refers
to further developing and refining existing process. Exploration is often more risky
than exploitation. The exploration-exploitation model implies the order of
organizational learning and product development, i.e. the innovation process
usually starts with exploration the results of which then are then perfected and
refined through exploitation until the cycle has been completed and the
organizations starts the exploration process again. Most companies engage in both
kinds of learning simultaneously.
Recently March’s theory on exploration and exploitation has been applied
to the study of strategic alliances between organizations (Koza & Lewin, 1998;
Rothaermel & Deeds, 2004). It is proposed that companies enter into exploratory
alliances when they want to discover something new together. On the other hand,
companies enter exploitative alliances to combine their complementary assets to
concentrate on the development of products based on existing technology or patents
(Rothaermel & Deeds, 2004). Thus a company makes its decision on what kind of
174

alliance to build based on whether it wants to explore new opportunities or exploit
existing opportunities (Koza & Lewin, 1998).
The transactional knowledge network is characterized by exploitative ties,
while the interactive knowledge network is characterized by exploratory ties.
Engagement in the transactional knowledge network enables organizations,
including both commercial companies and academic institutions to access and
utilize the knowledge developed externally. However, they engage in interactive
ties to explore new possibilities and create new knowledge. This leads to another
difference in the knowledge sharing process in the two types of networks: the
outcome of knowledge sharing.
Outcome of the knowledge sharing process
In the transactional knowledge network, knowledge is transferred from one
organization to another. That is why R&D literature labels this process as
“knowledge transfer.” There is no collaboration in the process of these transactions.
However, in the interactive knowledge network, not only is knowledge transferred
but at the same time new knowledge is created in the process of this long-term
interaction and conversation. As a result, the interactive network is an innovation
network
175

Transactional knowledge
network
Interactive knowledge
network
Codified Tacit and codified
Migratory knowledge Embedded knowledge
High market readiness Low market readiness
Nature of the
knowledge shared
Low stickiness High stickiness
Short-term Long-term
Formal, contractual Informal
Resource exchange Knowledge spillover
Arm’s length Embedded
Characteristics of
the KS process
Exploitive Exploratory
Outcome of KS Knowledge shared Knowledge shared and
created
Table 5-1. Characteristics of the knowledge sharing process
Characteristics of the knowledge networks
With different resources flowing into them, the transactional knowledge network
and the interactive knowledge network are different not only in terms of nature of
the knowledge shared, characteristics of knowledge sharing process, and outcome
of the knowledge sharing process, but also in terms of the structures of the network
on the nodal, dyad, and global level and the evolutionary logics of the two
networks.
First, the transactional ties in the transactional knowledge network are
market exchange ties, and as a result, the relationship between dyads is an
exchange relationship. Resource exchange is the theoretical framework most
appropriate for the explanation of the formation of transactional ties (Blau, 1964;
Monge & Contractor, 2003).
176

The exchange approach to knowledge sharing in networks, as represented
by the network exchange theory (Bienenstock & Bonacich, 1992, 1997; Cook,
1977, 1982; Cook & Whitmeyer, 1992; Cook & Yamagishi, 1992; Markovsky,
Willer, & Patton, 1988; Willer & Skvoretz, 1997; Yamagishi, Gillmore, & Cook,
1988), emphasizes that the sharing of knowledge between any two individuals or
organizations should be understood in the context of the larger social network and
their positions within it. It holds that the power of a particular organization in the
network is decided by how vulnerable it is to exclusion. Power is a function of the
organization’s centrality in the network. In short, it is the not the individual
characteristics of a particular organization, but the characteristics of its position
within the network that gives it power and influence.
While resource exchange in the transactional knowledge network is based
on the assumption of self-interest, the interactive knowledge network is
characterized by collaborative ties based on the premise of mutual interest (Monge
& Contractor, 2003). This type of network often features a community structure
(Laumann et al., 1978). A community has a more or less cohesive goal and
organizations in such a community often are more likely to engage in long-term
collaborations with each other rather than short-term transactions. Such a
community-based network often has a centralized structure, which facilitates the
realization of collective goals and the diffusion of knowledge (Marwell, Oliver, &
Prahl, 1988).
177

Studies in interorganizational networks have yet to fully examine the
relationship among multiple and overlapping networks (Powell, White, Koput, &
Owen-Smith, 2005). Therefore, responding to the suggestions that there is
difference in the network structures that is “most conducive to proliferating ideas
and transforming them into commercially significant inventions” (Podolny, Stuart,
& Hannan, 1996, p. 660), this study explores the similarities as well as differences
of attachment mechanisms in the development of the transactional and the
interactive knowledge networks and the interplay between resource relationships
and differential structural evolution.
Centrality has a key effect in the interactive knowledge network. In the past,
researchers have examined the effects of three types of centrality on various output
variables. They are: degree centrality, closeness centrality, and betweenness
centrality. Degree centrality is defined as the number of links a node has
(Wasserman & Faust, 1994). It gives a node visibility. Closeness centrality is the
extent to which a node in the network can reach other nodes through the smallest
number of steps. It gives the node control over the access to resources, and thus
power (Wasserman & Faust, 1994). Betweenness centrality is the extent to which a
node sits in between two nodes that are not directly connected (Wasserman & Faust,
1994). It also gives the focal node the power to control resources.
Past empirical studies found various effects of these different measures of
centrality on organizational performance. Powell et al. (1996) studied the effects of
178

degree centrality and closeness centrality on the subsequent growth of companies
and the likelihood of future ties. They discovered that while degree centrality
positively predicts the subsequent size and growth of companies, closeness
centrality predicts the number of R&D ties organizations engage in over time.
Owen-Smith and Powell (2004) found that betweenness centrality in a
geographically dispersed network has a positive effect on a company’s innovative
output. But this is not the case in a geographically clustered network. Powell et al.
(2005) tested the accumulative advantage hypothesis, i.e., organizations that exhibit
a high degree of centrality are more likely to form interorganizational ties later, but
they found very week support for the model. I argue that the reason for the
complexity and inconsistencies in the findings of these empirical studies can be
attributed to the fact that these existing studies treat networks composed of
different kinds of ties as more or less comparable. For instance, Powell et al. (1996)
examines R&D collaboration networks, while Owen-Smith and Powell (2004) and
Powell et al. (2005) look at four types of ties at the same time: R&D collaboration,
patent licensing, commercialization and marketing arrangement, and investment.
While existing research offers interesting initial insight into how knowledge is
shared in interorganizational networks, it is necessary to recognize the different
structural characteristics and evolutionary logics of these different types of
networks as well as how knowledge flows differently in these distinctive networks.
179

Organizations participate in the interactive and transactional knowledge
networks out of different motivations. While organizations establish short-term
transactional ties to acquire the resources they need, including knowledge, data,
material, and capital, organizations enter interactive ties, such as a research
collaboration or an interlocking directorate not only to access knowledge but also
to take advantage of the social capital of their partners. As a result, when they are
considering forming new interactive ties, organizations are more likely to build ties
with previously highly centralized organizations in the network. On the contrary,
when organizations seek to build new transactional ties to acquire a specific piece
of knowledge or technology, they are more likely to turn to those partners that can
provide the knowledge needed at the lowest price. This leads to the following
hypothesis:
H1: The growth of an interactive knowledge network follows the logic of
preferential attachment, while the growth of a transactional knowledge
network does not show this pattern.
As discussed in H1, when organizations seek to form new interactive ties,
they are more likely to connect with those organizations with high degree
centrality. However, their previously high degree of centrality does not give an
organization an advantage in the transactional knowledge network. As a result, as
the two networks grow, the transactional network will become increasingly
decentralized. Thus there should be low within-cluster centralization and low
180

centralization at the global level in the transactional knowledge network, and high
within-cluster centralization and high centralization at the global level in the
interactive knowledge network. This leads to Hypothesis 2.
H2: An interactive knowledge network will demonstrate a higher (a) global-
level centralization and (b) within-cluster centralization than a
transactional knowledge network.
Structural inertia and embeddedness are two theoretical concepts useful in
understanding the evolutionary and coevolutionary dynamics of multiple
interorganizational networks. Structural inertia is a force that leads to the retention
of organizational forms and structure (Hannen & Freeman, 1984). When applied to
network changes, structural inertia can be defined as, “a persistent organizational
resistance to changing interorganizational dyadic ties or difficulties that an
organization faces when it attempts to dissolve old relationships and form new
network ties” (Kim, Oh, & Swaminathan, 2006, p. 704).
This leads to a discussion of economic behaviors that are neither completely
isolated from nor totally decided by the social relational contexts in which they
occur (Granovetter, 1985). The fact that economic organizations are embedded in
social relations prevents them from engaging in completely opportunistic behavior
and helps the building of trust, which in turn, lowers the transaction costs
(Granovetter, 1985). Organizations are subject to network inertia partly because
they are embedded in a set of more or less stable social relationships.
181

Past empirical studies found that when organizations build repeated ties
with the same partner, they can significantly reduce the uncertainty in their
relationships (Podolny, 1994) and the costs of transaction (Ahuja, 2000b; Gulati,
1995b). Prior collaboration also builds trust, which increases the likelihood of
repeated interaction in the future (Gulati, 1995a). As a result, two organizations that
have one type of tie in the past are more likely to build the same type of ties in the
future. This should be true for both the interactive and the transactional knowledge
networks, which leads to hypothesis 3a and hypothesis 3b.
H3a: In the interactive knowledge network, two nodes that have had ties in
the past are more likely to share ties in the future.
H3b: In the transactional knowledge network, two nodes that have had ties
in the past are more likely to share ties in the future.
Two organizations can engage in a multiplicity of ties. For instance, an
organization can have both transactional and interactive ties with another
organization by licensing a patent from the other organization and also establishing
research collaborations with it. Researchers have expressed differing opinions
regarding whether two organizations that have had one type of tie in the past are
also likely to have other types of ties in the future.
Some argue that as the social network ages and embeddedness increases,
the coevolution mechanism of network structure across multiplex networks will
182

increase. In other words, different networks will become increasingly similar as
they develop. Based on the alliance literature, interorganizational networks
facilitate new alliances by providing information about the capabilities and
reliability of potential partners, the partnership’s activities, and the community’s
activities as a whole, and by decreasing the hazards associated with new alliances
(Ahuja, 2000a; Baum & Ingram, 2002; Gulati & Garguilo, 1999). As a result, the
existence of one type of tie in the past is likely to predict the occurrences of other
types of ties in the future. Gulati and Gargiulo (1999) found that prior direct and
indirect alliances between organizations are likely to enhance further alliances
between organizations. Others argue that this cross-network prediction does not
need to happen. For instance, Laumann (1978) argued that, “links of one type
between organizations (e.g., transfer of funds or personnel) need not imply bonds
of another type between the same organizations” because the formation of different
ties is driven by different motivations (p. 463). This leads to the following
hypotheses.
H4a: The existence of past interactive ties between two nodes predicts the
occurrence of future transactional ties.
H4b: The existence of past transactional ties between two nodes predicts
the occurrence of future interactive ties.
183

Methodology
This section discusses the data used to test the hypotheses proposed above and the
methods used to analyze them.
Data
All data used in this study were derived from the Medtrack Database. Medtrack is
a private company producing information and analysis service to biotech
companies and investors. Their “deals and alliances” database contains a list of all
the deals and alliances in the biomedical field during the period between 1987 and
2007. Each entry in the database represents a deal or alliance. Each entry includes
the following information: date, types of deal, technological component, source of
the data, deal title, deal summary, parties involved in the deal, royalties/payments,
and therapeutic category.
All the deals and alliances were coded into eleven categories in the original
database: acquisition, co-development, commercialization, development,
distribution, license, manufacturing, marketing, research, sales, and supply. Table
5-3 presents the number of each type of deals. These categories are not mutually
exclusive, i.e. a particular deal can belong to more than one category. For instance,
a deal between two companies can involve both research and marketing. The
Medtrack database was then coded into 11 separate networks based on the
categorization of the deals, with each node representing a unique organization and
each tie representing a deal. When there are more than 2 organizations involved in
184

a deal, the data were coded as if there is a tie between any pair of organizations. For
instance, when organizations A, B, and C are all involved in a deal, this deal was
coded as: A-B, B-C, and A-C.
Type of deals Number of deals
Development 6768
Research 5248
Collaboration 5845
License 4907
Distribution 2612
Marketing 2342
Commercialization 1494
Supply 1443
Manufacturing 1308
Acquisition 1000
Sales 918
Co-development 632
Joint venture 450
Miscellaneous 3314
Table 5-2. Number of deals
In the original database, the “parties involved” are categorized as following:
“global company (any company that is not trading publicly in the US),” “US public
company,” “Academic organization,” “NGO,” “Defense organization,” and “Other
organization.” For this study, these organizations are recoded into three exclusive
categories: “Corporation,” “Academic organization,” and “Governmental agency.”
Operationalization
In preparing the data for network analysis, I operationalized interactive knowledge
network as the network of research collaboration deals and operationalized
transactional knowledge network as the network of patent licensing deals.
185

Sampling
To test hypotheses 1, 3a, 3b, 4a, and 4b, I sampled a subnetwork from the original
dataset. I used a subnetwork instead of the complete network derived from
Medtrack for two reasons. First, the original dataset contains deals of many
therapeutic categories. It would be meaningful to look at the evolutions of
interorganizational networks regarding one specific therapeutic category in order to
understand the development trajectory of one research area. Secondly, including a
subnetwork can significantly reduce the time for computer to analyze the data.
To reduce the size of network for network analysis, only those deals and
alliances that are related to HIV/AIDS research and commercialization were
selected. A total of 129 license deals and 115 research deals were identified.
Data analysis
Regression analysis
Simple linear regression was conducted to test hypothesis 1. First, I calculated the
degree centrality of all nodes in both transactional and interactive networks in years
2002-2006, using UCINET 6.0 (Borgatti, Everett, & Freeman, 2002). Second, all
the centrality data were exported into a spreadsheet to be analyzed in SPSS. Next, I
ran separate linear regression analyses in SPSS for both interactive network data
and transactional network data to examine whether previous centrality in a network
predicts subsequent centrality in the same type of network.
186

Network analysis
UCINET 6.0 was used in conducting descriptive statistical analysis and testing
hypotheses 2, 3 and 4 (Borgatti et al., 2002). Several network analysis techniques
were used.
The first pair of statistics measured was centrality and centralization.
Centrality is a measurement of how much power an organization has in the network
(Wasserman & Faust, 1994). Centrality can be measured in different ways,
including: degree centrality, closeness centrality, and betweenness centrality. For
the purpose of this study, Freeman degree centrality was used. While centrality is a
measure of nodal attributes, centralization is a measurement of the power inequality
of the network (Wasserman & Faust, 1994). It is calculated as the ratio between the
degrees of most centralized nodes and the degrees of the least centralized nodes in
the network. These two statistics were used in testing hypothesis 2.
Hypotheses 3a, 3b, 4a, and 4b were tested using MRQAP. MRQAP stands
for multiple regression quadratic assignment procedure. It is used to regress one
dependent matrix onto one or more independent matrices. It is particularly useful in
testing regression hypotheses on network data (Krackhardt, 1988). MRQAP
includes two steps. First it performs a standard multiple regression across
corresponding cells of the dependent and independent matrices. Second, it
randomly permutes rows and columns of the dependent matrix and recomputes the
187

regression model. The second step is repeated a large number of times (in this study
2000 times) to estimate the standard errors of the statistics of interest.
In order to test hypotheses 3a and 3b (two nodes sharing a tie in the past are
more likely to share the same type of tie in the future), I constructed two groups of
adjacency matrices, one for interactive network and one for transactional network.
In each group, I constructed matrices representing the network for each year
between 2002 and 2006. In each group, I regressed the 2002, 2003, and 2004
networks on the 2005 network, and regressed the 2003, 2004, and 2005 networks
on the 2006 network using MRQAP.
In order to test hypotheses 4a and 4b (that two nodes sharing one type of tie
in the past are more likely to share another type of tie in the future), I conduct
multiple regression analyses very similar to those used in testing 3a and 3b. The
only difference is that I used one type of network, for instance the interactive
network, to predict the other type of network, for instance, the transactional
network.
Visualization
Finally, I used Pajek to provide year-by-year graphic representations of both
interactive networks and transactional networks between 1995 and 2006. Pajek is a
network visualization software that is particularly suitable for the representation of
large and complex networks (Batagelj & Mrvar, 2007).
The results of the data analysis will be presented next in Chapter 6.
188

Chapter 6. Result of network analysis
The previous chapter presents a theory of knowledge networks that identifies two
distinctive yet related types of interorganizational knowledge networks: interactive
knowledge networks and transactional knowledge networks. A group of hypotheses
were proposed based on the theoretical understanding of the structural and
evolutionary characteristics of these two networks. This chapter presents and
discusses the results of the network analysis study.
Descriptive statistics
There are a total of 6644 license deals and 5797 research deals in seventeen-year
period between 1990 and 2006 in the database. Table 6-1 presents some of the
basic descriptive statistics of the two networks. During this period, a total of 4803
organizations participated in the two networks. Among these organizations, 3281
participated in the licensing network and 3054 participated in the research network.
A total of 1598 organizations participated in both networks. Table 6.2 presents the
breakdown of these organizations by category. These organizations are coded into
three mutually exclusive categories: corporations (4073), academic institutions
(658), and governmental agencies (72). The twenty most connected organizations
in each network are listed in Table 6.3
The data also show that both networks grow significantly during the period
between 1990 and 2006 (see Figure 6-1). Table 6-4 presents the number of ties in
the license network and the research network each year between 1990 and 2006. As
189

the data demonstrate, the number of nodes and ties in both networks increase
significantly through the years. The appendix includes the visualization of the two
networks in 1995, 2000, and 2006 generated by Pajek.
Another interesting finding from the longitudinal data is that as the
networks grow, the license network becomes increasingly decentralized while the
centralization of the research network remains relatively constant, as demonstrated
in Figure 6-2. This indicated that the two networks grow following different
evolutionary logics.
Size Degree Network Centralization
No. of
nodes
No. of
ties
Degre
e
Norm.
Degree
Centraliza
tion
Hetero
-
geneit
y
Norm.
License
network
3281 6444 3.928
(SD=
9.827)
0.006
(SD=0.0
15)
0.28 0.22 0.19
Research
network
3054 5647 3.697
(SD=1
4.321)
0.011
(SD=0.0
43)
2.01 0.52 0.49
Table 6-1. Descriptive statistics of two networks

Number Types of organizations
4073 Corporation
658 Academic, Research institutions, nongovernmental/nonprofit health
organizations and hospitals
72 Governmental agencies
Table 6-2. Number of nodes by category
190

Number of license and research deals (1990-2006)
0
200
400
600
800
1000
1990
1992
1994
1996
1998
2000
2002
2004
2006
License
deals
Research
deals

Figure 6-1. Number of license and research deals 1990-2006
License network Research network
Name Degree Norm.
degree
Name Degree Norm.
degree
1 GlaxoSmithKli
ne (2873)
185 0.282 The U.S. Department
of Health and Human
Services (6796)
677 2.016
2 Novartis (4886) 181 0.276 Pfizer Inc (5313) 144 0.429
3 Pfizer Inc
(5313)
165 0.252 Merck & Co Inc
(4328)
120 0.357
4 Merck & Co Inc
(4328)
151 0.230 GlaxoSmithKline
(2873)
120 0.357
5 Johnson &
Johnson (3617)
143 0.218 Novartis (4886) 117 0.348
6 Roche Holdings
Ltd (5885)
118 0.180 Johnson & Johnson
(3617)
95 0.283
7 Bayer AG (888) 112 0.171 AstraZeneca (710) 90 0.268
8 Sanofi-Aventis
(6045)
108 0.165 Bayer AG (888) 90 0.268
9 Crucell (1895) 97 0.148 Eli Lilly & Co (2328) 78 0.232
10 Bristol-Myers
Squibb
Company
(1242)
87 0.148 Roche Holdings Ltd
(5885)
74 0.220
Table 6-3. 20 most connected organizations in two networks
191

11 The U.S.
Department of
Health and
Human Services
(6796)
87 0.133 Sanofi-Aventis
(6045)
72 0.214
12 AstraZeneca
(710)
85 0.130 Bristol-Myers Squibb
Company (1242)
60 0.179
13 Abbott
Laboratories
(123)
78 0.110 University Health
Network (6067)
59 0.176
14 Wyeth (7488) 78 0.110 MERCK (4028) 55 0.164
15 Amgen Inc
(495)
70 0.107 Amgen Inc (495) 53 0.158
16 GlaxoSmithKli
ne (2873)
67 0.102 Boehringer Ingelheim
(1194)
51 0.152
17 University Of
California
(7073)
64 0.098 US Army (7002) 50 0.149
18 University
Health Network
(6067)
57 0.087 University Of
California (7073)
49 0.146
19 Genentech Inc
(2769)
57 0.087 Medarex Inc (4190) 42 0.125
20 MERCK (4028) 54 0.082 General Electric
Company (2775)
42 0.125
Table 6-3. Continued
192

License network Research network Year
Number
of ties
Number
of nodes
Centrali-
zation
Number
of ties
Number
of nodes
Centrali-
zation
2006 940 1060 0.43 878 973 3.73
2005 895 1007 0.31 775 858 2.99
2004 877 941 0.35 711 791 3.94
2003 656 775 0.96 711 760 3.70
2002 600 671 1.59 657 701 2.94
2001 621 672 1.88 583 618 4.89
2000 464 506 1.20 397 426 3.85
1999 339 391 1.91 281 313 4.07
1998 280 299 1.25 193 223 3.47
1997 213 243 1.72 162 164 5.35
1996 161 192 3.00 94 104 3.06
1995 121 168 5.19 69 89 2.59
1994 79 102 3.26 43 55 6.60
1993 78 98 2.23 33 50 7.82
1992 71 97 13.16 27 44 4.32
1991 36 45 10.47 21 32 5.81
1990 13 25 4.35 12 21 4.74
Table 6-4. Number of ties
0
2
4
6
8
10
12
14
1990
1992
1994
1996
1998
2000
2002
2004
2006
License
Research

Figure 6-2. Longitudinal centralization scores of two networks
11

11
The centralization scores of both research network and license network show unusual variation in
the early 1990s, especially those between 1990 and 1994. The dataset indicates that the numbers of
193

To test hypotheses 1, 3, and 4, I used a subnetwork: the AIDS network from 2002
to 2006. Some descriptive statistics of these two subnetworks are presented in
Table 6-5.

Interactive Transactional
Overall characteristics
Number of nodes 80 84
Number of ties 61 81
Centralization 11.34 2.50
Longitudinal statistics
2002 13 6
2003 10 14
2004 14 11
2005 12 25
Number
of ties
2006 12 28
Table 6-5. Descriptive statistics of AIDS networks, 2002-2006
Hypotheses testing
H1: The growth of interactive knowledge network follows the logic of preferential
attachment, while the growth of transactional knowledge network does not show
this pattern.
To test hypothesis 1, linear regression tests were conducted to evaluate the
prediction of the centrality of nodes in one particular year (t) from their centralities
in previous two years (t-2, and t-1) in the interactive network and the transactional
network. The results of these tests are presented in Table 6-6.

ties in these years in both networks are less than 100. Such abnormality is likely to be caused by the
incompleteness of the data in the early years.
194

In analyzing the interactive network data, it is found that previous centrality
is a significant predictor of subsequent centrality, which supports the first part of
H1. I regressed nodes centralities in 2002 ( β=0.32, p< 0.01) and 2003 ( β=0.21, p<
0.01) on their centrality in 2004, with R
2
=0.13, which means that 13% of the
variance in nodes centralities in 2004 were explained by nodes centralities in 2002
and 2004. I regressed nodes centralities in 2003 ( β=0.49, p< 0.01) and 2004
( β=0.41, p< 0.01) on their centrality in 2005, with R
2
=0.31, which means that 31%
of the variance in nodes centralities in 2005 were explained by nodes centralities in
2003 and 2004. I regressed nodes centralities in 2004 ( β=0.16, p< 0.01) and 2005
( β=0.01, p=0.77) on their centrality in 2006, with R
2
=0.13, which means that 13%
of the variance in nodes centralities in 2006 were explained by nodes centralities in
2004 and 2005.
In analyzing the transactional network data, it is found that previous
centrality is not a significant predictor of subsequent centrality, which supports the
second half of H1. I regressed nodes centralities in 2002 ( β=0.16, p=0.15) and 2003
( β=-0.08, p=0.90) on their centrality in 2004, with R
2
=0.01, which means that only
1% of the variance in nodes centralities in 2004 were explained by nodes
centralities in 2002 and 2004. I regressed nodes centralities in 2003 ( β=0.16, p=
0.21) and 2004 ( β=0.08, p=0.54) on their centrality in 2005, with R
2
=0.01, which
means that only 1% of the variance in nodes centralities in 2005 were explained by
nodes centralities in 2003 and 2004. I regressed nodes centralities in 2004 ( β=0.01,
195

p=0.91) and 2005 ( β=0.07, p=0.25) on their centrality in 2006, with R
2
=0.01, which
means that again only 1% of the variance in nodes centralities in 2006 were
explained by nodes centralities in 2004 and 2005.
Hypotheses 2004 2005 2006
2002 0.317**
(0.238)

2003 0.209**
(0.206)
0.488**
(0.418)

2004 0.409**
(0.263)
0.158**
(0.232)
2005 0.017
(0.021)
R
2
0.13 0.31 0.06
Interactive
network
Adjusted R
2
0.12 0.31 0.05
2002 0.165
(0.098)

2003 -0.079
(-0.008)
0.155
(0.084)

2004 0.079
(0.042)
0.014
(0.007)
2005 0.073
(0.076)
R
2
0.01 0.01 0.01
Transactional
network
Adjusted R
2
0.00 0.00 0.01
** p<0.01
Table 6-6. Regression analysis results for hypotheses 4a and 4b
H2: An interactive knowledge network will demonstrate higher (a) global-level
centralization and (b) within-cluster centralization than a transactional knowledge
network.
It is found that the interactive network is more centralized than the transactional
network. The interactive knowledge network has a freeman centralization score of
196

2.01, with heterogeneity = 0.52 and normalized centralization = 0.49. The
transactional knowledge network has a freeman centralization score of 0.28, with
heterogeneity = 0.22 and normalized centralization = 0.19.
To test the second part of the hypotheses, the clustering coefficient, i.e.,
density in the local neighborhood of an actor, measured by the proportion of
transactive to intransitive triads are calculated (Robins, Pattison, & Woolcock,
2005). The results are presented in Table 6-7. It is found that the interactive
network indeed has higher within cluster centralization than a transactional
network.
In short, the interactive network is found be more centralized than the
transactional network on both the global level and local/cluster level. As a result,
hypothesis 2 is supported.
H3a: In the interactive knowledge network, two nodes that have ties in the past are
more likely to share ties in the future.
H3b: In the transactional knowledge network, two nodes that have ties in the past
are more likely to share ties in the future.
197

Transactional network Interactive network
Year Neighborhood
density
(Clustering
coefficient)
Overall
density
Neighborhood
density
(Clustering
coefficient)
Overall
density
1990 0.000 0.022 0.000 0.029
1991 0.280 0.018 0.000 0.021
1992 0.168 0.017 0.214 0.014
1993 0.110 0.008 0.125 0.014
1994 0.095 0.004 0.150 0.015
1995 0.071 0.004 0.094 0.009
1996 0.137 0.004 0.162 0.009
1997 0.143 0.004 0.106 0.006
1998 0.088 0.003 0.073 0.004
1999 0.073 0.002 0.142 0.003
2000 0.093 0.002 0.123 0.002
2001 0.104 0.001 0.118 0.002
2002 0.054 0.001 0.142 0.001
2003 0.090 0.001 0.147 0.001
2004 0.124 0.001 0.158 0.001
2005 0.181 0.001 0.198 0.001
2006 0.127 0.001 0.157 0.001
Table 6-7. Clustering coefficient of two networks
Table 6-8 presents the MRQAP results for Hypotheses 3a, 3b, 4a, and 4b. To test
hypotheses 3a, I regressed the existence of ties in the interactive network in 2002
( β=0.30, p< 0.01), 2003 ( β=0.01, p< 0.01), and 2004 ( β=0.27, p< 0.01) on ties in
2005. I also regressed the existence of ties in 2003 ( β=0.16, p< 0.01), 2004
( β=0.50, p< 0.01), and 2005 ( β=0.02, p< 0.01) on ties in 2006. Both regressions
provided significant results, indicating that the existence of previous interactive ties
is an important determinant of the formation of future interactive ties in each dyad.
The total variance explained by the models is respectively R
2
=0.238

and R
2
=0.312,
198

indicating that the existence of previous relationship explain 28% and 31% of the
future ties in each dyad.
To test hypothesis 3b, I regressed the existence of ties in 2002 ( β=0.05, p<
0.01), 2003 ( β=0.17, p< 0.01), and 2004 ( β=0.68, p< 0.01) in the transactional
knowledge network on ties in 2005. I also regressed the existence of ties in 2003
( β=0.14, p< 0.01), 2004 ( β=0.23, p< 0.01), 2005 ( β=0.15, p< 0.01) on ties in 2006.
Both regressions provided significant results, indicating that the existence of
previous transactional ties is an important predictor of the formation of future
transactional ties in each dyad. The total variance explained by the models is
respectively R
2
=0.305

and R
2
=0.096, indicating that the existence of previous
relationship explain 31% and 10% of the future ties in each dyad. As a result, both
hypothesis 2a and hypothesis 2b are supported.

199

2005 2006
2002 0.295**
(0.317)

2003 0.012**
(0.011)
0.156**
(0.143)
2004 0.270**
(0.292)
0.460**
(0.500)
2005 0.019**
(0.019)
R
2
0.238 0.312
H3a: Interactive
knowledge
network

Adjusted R
2
0.238 0.312
2002 0.047**
(0.023)

2003 0.174**
(0.130)
0.139**
(0.095)
2004 0.682**
(0.452)
0.228**
(0.138)
2005 0.153 **
(0.139)
R
2
0.305 0.096
H3b: Transactional
knowledge
network

Adjusted R
2
0.305 0.096
2002 0.071**
(0.050)

2003 0.095**
(0.103)
0.270**
(0.291)
2004 0.190**
(0.182)
-0.142**
(-0.136)
2005 0.045**
(0.064)
R
2
0.077 0.063
H4a: Transactional
network predicts
interactive network
Adjusted R
2
0.077 0.063
** p<0.01, * p<0.05
Table 6-8. MRQAP results for hypotheses 3 and 4
200

2002 0.310**
(0.224)

2003 0.285**
(0.180)
0.263**
(0.152)
2004 -0.007*
(-0.005)
0.178**
(0.121)
2005 -0.057**
(-0.036)
R
2
0.131 0.042
H4b: Interactive network
predicts transactional
network
Adjusted R
2
0.131 0.042
Table 6-8. Continued
H4a: When two nodes have transactional ties in the past, they are more likely to
have interactive ties in the future.
H4b: When two nodes have interactive types in the past, they are more likely to
have transactional ties in the future.
To test hypotheses 4a, I regressed the existence of transactional ties in 2002
( β=0.07, p< 0.01), 2003 ( β=0.09, p< 0.01), and 2004 ( β=0.19, p< 0.01) on
interactive ties in 2005. I also regressed the existence of transactional ties in 2003
( β=0.27, p< 0.01), 2004 ( β=-0.14, p< 0.01), and 2005 ( β=0.04, p< 0.01) on
interactive ties in 2006. Both regressions provided significant results, indicating
that the existence of previous interactive ties is an important determinant of the
formation of future interactive ties in each dyad. The total variance explained by
the models is respectively R
2
=0.08

and R
2
=0.06, indicating that the existence of
previous transactional ties explain 8% and 6% of the future interactive ties in each
dyad.
201

To test hypothesis 4b, I regressed the existence of interactive ties in 2002
( β=0.31, p< 0.01), 2003 ( β=0.29, p< 0.01), and 2004 ( β=-0.01, p< 0.05) on
transactional ties in 2005. I also regressed the existence of interactive ties in 2003
( β=0.26, p< 0.01), 2004 ( β=0.18, p< 0.01), and 2005 ( β=-0.06, p< 0.01) on
transactional ties in 2006. Both regressions provided significant results, indicating
that the existence of previous transactional ties is an important determinant of the
formation of future transactional ties in each dyad. The total variance explained by
the models is respectively R
2
=0.131

and R
2
=0.042, indicating that the existence of
previous relationship explain 13% and 4% of the future ties in each dyad.
However, comparing the R
2
values in the regressions above, it is found that the R
2

values in hypotheses 4a and 4b are very small, while the R
2
values in hypotheses 3a
and 3b are significantly larger. It shows that interactive ties are good predictors of
future interactive ties and transactional ties are good predictors of future
transactional ties. However, these two types of ties do not predict each other very
well.
Discussion
The data analyses presented above demonstrate that the interactive knowledge
network and the transactional knowledge network do follow some similar and some
different evolutionary and co-evolutionary logics.
First, it is found that in the interactive knowledge network, organizations
are more likely to build ties with previously highly centralized nodes, and thus, the
202

rich gets richer. For instance, a node with high degree centrality at time t is more
likely to have a high degree centrality at a later time. However, this preferential
attachment pattern is not found in the transactional knowledge network. In other
words, a measure of previously high centrality does not give an organization an
advantage in forming subsequent transactional ties. This also explains why the
transactional knowledge network becomes increasingly decentralized across time,
while the centralization score of the interactive knowledge network remains
relatively constant.
12
Because of the differences discussed above, even though both networks
grow significantly during the period between 1990 and 2006, the transactional
knowledge network becomes increasingly decentralized. As a result, the interactive
network shows greater centralization on both the global level and the local level
than the transactional knowledge network. In other words, there is more inequality
in terms of power in the interactive knowledge network.
In both networks, two dyads that have ties between each other in the past
are more likely to have the same type of ties in the future, as discussed in
hypotheses 3a and 3b. This is consistent with findings of the interview study
presented in Chapter 4. However, when examining whether two nodes with one
type of tie in the past are more likely to have the other type of tie in the future, I
discovered that even though this effect does exist, it is not as strong as the

12
In testing this hypothesis, I only looked at several years of data (2002-2006) so it is possible that
over a longer period of time other patterns would emerge.
203

prediction of the same type of tie, with very small R
2
values in the regressions used
to test hypotheses 4a and 4b.
The results suggest several things. First, it is clear that different networks
have their own courses of development, following different evolutionary logics.
Both the interactive knowledge network and the transactional knowledge network
show sign of structural inertia. However, only the interactive knowledge network
shows sign of preferential attachment. This offers a possible explanation of the
inconsistency in current research in terms of how interorganizational knowledge
networks grow. Second, it is possible that these two types of network co-evolve. In
other words, existing tie between two nodes in one network might increase the
chances by which the same nodes will have ties in another network at a later time.

204

Chapter 7. Discussion and conclusion
This chapter will summarize the purpose, theoretical framework, research context
and methodologies, and main findings of this dissertation and discuss its
conclusions, implications, limitations and further research directions.
Theoretical framework
This dissertation applies the theory and methodology of social network analysis to
the study of interorganizational knowledge sharing. Several literatures, including
the interorganizational relations literature, the knowledge management literature,
the academic-industry knowledge sharing literature, and the social network analysis
literature inform the current study of interorganizational knowledge sharing as
interorganizational knowledge networks.
Dealing with the challenge to bridge the gap between the study of network
content and that of network structure (Contractor et al., 2006), I proposed a new
theory of interorganizational knowledge networks, which identifies two distinctive,
yet related types of interorganizational knowledge networks: interactive knowledge
network and transactional knowledge network.
An interactive knowledge network is based on long-term, interactive
collaborative relationships among different organizations. One example of
interactive knowledge networks is the research collaboration network. The
interactive knowledge network allows the sharing of a broad range of knowledge.
On the other hand, a transactional knowledge network is based on short-term,
205

transactional relationships among organizations, such as patent licensing, material
transfer, and data transfer. Only codified knowledge is shared in such a network.
While the transactional knowledge network allows the transfer of knowledge
among organizations, the interactive knowledge network enables the creation of
new knowledge through long-term communication and interaction.
Research contexts and methods
The biotechnology industry is an ideal ground for the study of academic-industry
knowledge sharing and interorganizational knowledge networks. Intensive
knowledge sharing and interaction between academic institutions and commercial
biotech companies is a distinctive feature of the industry (Mowery, Nelson, Sampat,
& Ziedonis, 1999). Furthermore, due to the fast expanding knowledge base of the
biotech industry, no single company can possess all the resources and capacities to
succeed alone. Innovations in the biotech industry take place in the
interorganizational networks instead of individual organizations (Powell et al.,
1996).
A two-study design was used to help us understand the sharing of
knowledge different types of organizations in the biotech sector, including
commercial companies, academic institutions, governmental agencies and NGOs.
Study 1 focused on understanding academic-industry knowledge sharing as a
special type of interorganizational tie. It is an interview study supplemented by 13
months’ of ethnography at a biotech business research center at a major private
206

research university. Interviews with several panels of biotech professionals,
including biotech company executives, company scientists, trade association
leaders, academic scientists, and technology transfer officers offered insights into
the goals, channels, and characteristics of academic-industry knowledge sharing.
Furthermore, this study allows us to understand academic-industry knowledge
sharing and interorganizational knowledge sharing not only as the interaction
among organizations but also as practiced by individual organizational members on
an on-going basis.
Study 2 is a quantitative study of interorganizational knowledge networks.
It is based on longitudinal network data derived from Medtrack, a private
biomedical industry database that contains data on deals and alliances in the
biomedical industry from 1990 to 2007. Nodal level, dyad level, and network level
measurements and statistics are used to test a group of hypotheses derived from
theoretical construction. The main findings of each research question and
hypothesis will be discussed next.
Discussion of main findings
The first four research questions were answered by interview and ethnography data.
RQ1: what are the goals of academic-industry knowledge sharing?
While existing literature on academic-industry knowledge sharing usually
distinguish the goals of academic institutions from those of commercial companies
207

on the organizational level, this study emphasizes the importance of understanding
the motivations of academic-industry knowledge sharing on both the organizational
level and the individual level. This is especially true with regard to academic
institutions in which individual researchers have considerable autonomy and
freedom from institutional constraints.
Biotech companies engage in academic-industry knowledge sharing to
identify new technologies that they can further develop and commercialize, to
outsource part of their in-house R&D at a lower cost, solve their own technical
problems, and to identify potential employees.
Within the academic community, academic scientists and academic
institutions often have different and even conflicting goals in their engagement in
academic-industry knowledge sharing. While academic institutions share
knowledge with industry to raise the institutional profile and social capital and to
gain financial reward, academic scientists are more likely collaborate with industry
to get extra research funding quickly and to involve themselves in the advanced
research conduct in companies. Consequently, academic institutions prefer patent
licensing as the channel of knowledge sharing, as it brings enormous financial
return in the long run, while academic scientists often favor research collaboration,
which will give them a limited amount of research money and access to company
research quickly.
208

RQ2: How is knowledge shared between academia and industry?
Answering RQ2 requires the understanding of the following three questions: What
knowledge do academia and industry share, how is the knowledge shared between
academia and industry protected as intellectual properties, and what are the
channels of academic-industry knowledge sharing?
The knowledge shared between academia and industry includes but is not
limited to, scientific discoveries and inventions, technologies, technical know-
hows, understanding of science or industry, research equipment and materials,
research techniques, and research data. The interview study discovered that the
knowledge shared between academia and industry could be understood along two
dimensions: articulatedness and market readiness. The knowledge shared between
academia and biotech companies are protected as different types of intellectual
properties: copyright, patent, and trade secret.
Interviewees reported using a wide variety of channels in academic-industry
knowledge sharing, including patent licensing, research collaboration (including
research contracts and research grants), publication, conferences and seminars,
consulting, spin-offs, material and data transfer agreements, invited talks and visits,
educational and training programs, special industry-oriented research centers,
knowledge brokers, hiring, and personal communication. While existing literature
mainly focuses on publication, patent licensing, and spin-off, this study emphasizes
the importance of research collaboration as a unique channel that allows long-term
209

interaction and knowledge sharing as well as the creation of new knowledge.
Furthermore, the study identified the important role of knowledge broker as a third
party that bridges the gap between academia and industry.
The interview study further identified that these academic-industry
knowledge-sharing channels could be understood on a continuum from highly
formal (e.g. patent licensing) to high informal (e.g. personal communication).
Highly formal channels are likely to be used when the knowledge is very well
articulated and highly ready for market. Interviewees also reported that the
simultaneous use of multiple channels often leads to better outcomes: effective
knowledge sharing, successful product development, and commercialization.
RQ3: How do the professional cultures of academia and industry affect the
knowledge sharing between the two communities?
Academic culture and industry culture have both similarities and differences
Academia is often science driven while industry is often bottom line driven. This
difference in the basic incentive of the two communities leads to the many
difference between academic and industry research and the ways in which the two
communities manage knowledge and engage in knowledge sharing. In terms of the
characteristics of research, academic research is characterized as curiosity-driven,
rich in distraction, enjoying academic freedom, and typically slow. The culture of
industry, on the other hand, is characterized by a set of contrasting features:
product-driven, highly focused, highly disciplined, and very fast. In terms of the
210

culture of knowledge sharing, academia is often considered open while industry is
often labeled as secretive.
Even though the pairs of dialectics discussed above offer a shortcut in
understanding the difference between academic culture and industry and how
professional cultures affect academic-industry knowledge sharing, they are often
over-simplified. Furthermore, the past two decades have witnessed a convergence
between the academic culture and the industry culture. On the one hand, biotech
industry is becoming increasingly science-driven and interested in basic research.
On the other hand, academia is becoming more and more industry-driven and
proactive in commercialization.
The changing characteristics of the academic and industry further challenge
the simplistic notion that academic culture is open and industry is secretive. First,
the interviews suggested that it is problematic to consider academia or industry as
each having an integral culture. Within academia, faculty scientists and technology
transfer specialists are often completely against each other on many issues of
knowledge sharing and IP management. For instance, for faculties, success in
academic-industry knowledge sharing is measured by whether a useful product is
developed and how much research funding the faculty can get for his lab. On the
contrary, the success of technology transfer offices is measured by the number of
patent licensed and the amount of royalty income generated for the university
(Thursby et al., 2001). As a result, the two have very different agenda and strategies
211

in sharing knowledge with industry. Similarly, significant differences can be
observed between the attitudes and the practices of business executives and
company scientists. Interestingly, a comparison of the narratives given by academic
technology transfer officers and biotech business executives reveals more
similarities than differences in terms of their attitudes towards IP ownership, IP
management, and knowledge sharing.
Second, there is great internal variation among academic institutions in their
handling of intellectual properties and academic-industry knowledge sharing. Some
universities are more open, while others are more careful and more skillful in
guarding their intellectual properties. In general, non-teaching research institutes
are more cautious and effective in protecting their IPs in academic-industry
knowledge sharing, and thus often more secretive than universities. However, no
systematic difference between public and private universities was identified.
Third, the dichotomy between openness and secrecy is even more dubious
when one considers the differences in people’s attitudes towards knowledge
sharing or their openness when they are engaged in interorganizational knowledge
sharing and intraorganizational knowledge sharing. More specifically, while
companies are often labeled as secretive in sharing knowledge with academia or
with other companies, the intraorganizational knowledge sharing within one
company is amazingly open. Knowledge sharing is very free and intensive within a
company due to the shared goal of the company and the company ownership of the
212

IP. On the contrary, the intraorganizational knowledge sharing within a university
is sometimes not as free as one might think, because of the independence of each
academic labs and because of faculties’ conflicts of interests.
RQ4: What is the role of informal knowledge sharing in academic-industry
knowledge sharing in biotech industry?
Informal knowledge sharing was found to be one crucial component of academic-
industry knowledge sharing. Informal knowledge sharing fulfills two important
goals: a social goal and a technical goal. First, informal knowledge sharing
facilitates the building of relationships and social networks. Second, informal
knowledge sharing often facilitates the sharing of tacit knowledge that is not very
articulated such as technical know-hows. Trust is a central feature in informal
knowledge sharing and relationship building.
The second study tested a group of hypotheses regarding the structural and
evolutionary characteristics of the interactive knowledge and the transactional
knowledge network.
H1: The growth of interactive knowledge network follows the logic of preferential
attachment, while the growth of transactional knowledge network does not show
this pattern.
A preliminary look at the change in centralization in the two networks shows a
clear difference. While the transactional knowledge network becomes increasingly
213

decentralized over time, the centralization of the interactive knowledge network
remains constant. H1 provides one possible explanation for this difference.
It is found that in the interactive knowledge network, nodes that have higher
degree centrality at a previous time are more likely to have higher degree centrality
at a later time. In other words, organizations are more likely to form ties with
highly centralized organizations in the network. However, this pattern of
preferential attachment is not found in the transactional knowledge network, which
means that previous higher degree centrality does not give an organization any
advantage in forming later transactional ties.
H2: Interactive knowledge network will demonstrate higher (a) global-level
centralization and (b) within-cluster centralization than transactional knowledge
network.
An analysis of the Medtrack data shows that the interactive knowledge
network does demonstrate higher global-level centralization as well as within-
cluster centralization than the transactional knowledge network.
214

H3a: In the interactive knowledge network, two nodes that have ties in the past are
more likely to share ties in the future.
H3b: In the transactional knowledge network, two nodes that have ties in the past
are more likely to share ties in the future.
In both the interactive and the transactional knowledge networks, dyads that
have ties at a previous time are more likely to have the same type of ties in the
future. This is consistent with the findings of the interview study.
H4a: When two nodes have transactional ties in the past, they are more likely to
have interactive ties in the future.
H4b: When two nodes have interactive types in the past, they are more likely to
have transactional ties in the future.
In both the interactive and the transactional knowledge networks, dyads that have
one type of ties at a previous time are more likely to have the other type of ties in
the future. This is also consistent with the findings of the interview study.
Theoretical implications
This dissertation advances current research on interorganizational knowledge
sharing in three major aspects.
First, organizations are embedded in a web of social and business relations.
To understand interorganizational knowledge sharing between two organizations,
215

we need to situate it in the large interorganizational networks in which these
organizations are embedded. In other words, whether two organizations are going
to build knowledge-sharing ties is not only a function of the characteristics of
individual organizations, but also a function of the interorganizational networks.
Second, interorganizational knowledge sharing is not only the transaction
among individual organizations, but also the interaction and communication among
individual members of these organizations. As a result, it is essential to study
interorganizational knowledge sharing as perceived, practiced and interpreted by
individuals. For instance, the interview study shed lights on how different groups of
people involved in interorganizational knowledge sharing, company executives,
company scientists, technology transfer specialists, and academic scientists,
perceive and participate in academic-industry knowledge sharing differently and
how these differences causes conflicts and frustrations on both sides.
Third, in understanding interorganizational knowledge sharing as networks,
it is essential to connect network content to network structure (Contractor et al.,
2006). While the majority of the current studies on interorganizational network
focuses on either what is the content flowing in the network or what are the
structural characteristics of the network (Contractor et al., 2006), this study bridges
the gap in the existing literature by combining the study of network content to the
study of network structure. It pointed out that when the contents of networks differ,
the structures of those networks would be different too. More specifically, this
216

dissertation identifies two distinctive yet interdependent interorganizational
knowledge networks: the transactional knowledge network and the interactive
knowledge network. Theoretical constructions as well as actual data analysis
concluded that these two types of interorganizational networks are different in
terms of the nature of knowledge shared, characteristics of ties, and the outcome of
knowledge sharing. These differences, in turn, lead to the difference in the network
structures and evolutionary logics of the two networks. For instance, while in both
networks, two organizations having a previous tie are more likely to have other ties
in the future, only the interactive knowledge network follows the logic of
preferential attachment.
Practical implications
For those participating in academic-industry knowledge sharing
Knowledge sharing between academic institutions and commercial companies is a
tricky question. The communication between academia and industry has been
plagued with distrust, disappointments, and disillusionment. The interview study
sheds light on a few things people should keep in mind when they engage in
academic-industry knowledge sharing. Each side enters such relationship for
different purposes. It is essential to understand the motivations and values of each
other before entering into a knowledge-sharing relationship. Because of the
different modes and dynamics of academic and industry research, academic labs
217

and commercial companies in research collaborations need to have a clear
understanding of each other’s expectations and a timeline that is approved by both
parties to avoid misunderstanding and frustration.
For those making policies regarding knowledge-intensive industries
Science policy makers in the US as well as around the world have long been trying
to extend the development model of biotech industry to other science-based
industries to increase the social benefits of public and private funding in knowledge
infrastructure and basic research (Dalpe, 2003).
This study clearly shows that academic institutions and scientists are
actively participating in the commercialization of scientific discoveries. The Bayh-
Dole Act has clearly encouraged this trend. Allowing universities and others
organization to possess the intellectual properties derived from federally funded
research gives academic institutions the rights and incentives to patent and license
their knowledge. However, at the same time, such arrangement often hinders the
sharing of knowledge through other channels. For instance, to maximize the
financial return of their patents, academic institution often require their scientists to
patent their discoveries before they can publish it, which significantly slows down
the diffusion of knowledge not only to the industry but also in the academic
community as well. Furthermore, even though universities allow their scientists to
consult for companies, universities want to have strict control over what knowledge
can be shared through consultation and what knowledge can be shared only through
218

patent licensing, as patent licensing brings significant financial reward to the
university while consulting only benefits individual scientists.
Limitations and directions for future research
In the interview study, I identified the interviews based on quota sampling and
convenience sampling. Even though attempts have been made to provide a variety
of points of view, the representativeness of the sample cannot be ensured. One way
solve this problem is to conduct case studies on specific interorganizational or
academic-industry knowledge sharing deals and interviewed all people involved in
each projects. Such case studies can ensure that the investigator can gain access to
the stories, thoughts, and attitudes of all parties involved and construct a relatively
complete picture for each case. What’s more, such case studies can gain insights
into how different parties involved in a particular deal or alliance make sense of the
causes of the success or failure of the project.
In the network study, all organizations (companies, academic institutions
and governmental agencies) are treated as equal. However, it can be logically
hypothesized that these organizations play different roles in the interactive and the
transactional knowledge networks. Moreover, different types of companies, such as
large generalist companies and small specialists companies, may also participate in
the networks differently because of the different amount of resources they possess
and seek in these networks (Howard Aldrich, 1999). As a result, “the analysis of
interorganizational resource flows in multiple networks becomes more complicated
219

when nodes as well as linkages are differentiated by type” (Laumann et al., 1978, p.
465). Future research could be done to explore what kinds of network structural
positions different types of organizations are likely to occupy.
Furthermore, while the network study has been explicating the structural
characteristics of the interactive and the transactional knowledge networks, there is
no attempt to predict or test the relationship between the structural characteristics of
the two networks and their outcome variables, in terms of the amount and quality of
knowledge shared, the amount and quality of knowledge created, and the financial
performance of the commercial companies in the network. Further research could
be done to look at the correlation between the structures of interorganizational
knowledge networks and the outcome variables discussed above.
Another major limitation of the network study is that in the current dataset,
the direction of patent licensing is not specified. According to the theoretical
construction, the flow of knowledge in the interactive knowledge network is bi-
directional, while there is a one-way flow of knowledge in the transactional
knowledge network, i.e. from the holder of the patent to the licenser of the patent.
A failure in specifying the direction of transaction in the transactional knowledge
network prevents us from understanding the flow of knowledge.
Finally, while this study has been mainly focused on exploring and testing
the differences between the interactive and the transactional knowledge network,
only one hypothesis (H4) deals with the relationship between the two networks.
220

There is a need to further examine the relationship/interaction between these two
types of knowledge-sharing networks and its effect on the outcome of knowledge
sharing and on organizational performance. For instance, would the existence of
multiplex ties increase the chance of future ties and facilitates the effectiveness of
knowledge sharing? Furthermore, it would be important to examine whether the
multiplexity of ties increase the effectiveness of knowledge sharing, as suggested
by Almeida et al (2002).

221

References.
Adler, P. S. (2001). Market, hierarchy, and trust: The knowledge economy and the
future of capitalism. Organization Science, 12(2), 215-234.
Agrawal, A. (2000). Importing scientific inventions: Direct interaction, geography,
and economic performance. Manuscript.
Agrawal, A. (2001). University-to-industry knowledge transfer: Literature review
and unanswered questions. International Journal of Management Reviews,
3(4), 285-302.
Ahuja, G. (2000a). Collaboration networks, structural holes, and innovation: A
longitudinal study. Administrative Science Quarterly, 45, 425-455.
Ahuja, G. (2000b). The duality of collaboration: Inducements and opportunities in
the formation of interfirm alliances. Strategic Management Journal, 21,
317-343.
Aldrich, H. (1979). Organizations and environments. Englewood Cliffs, NJ:
Prentice-Hall.
Aldrich, H. (1999). Organizations evolving. Thousand Oaks, CA: Sage.
Aldrich, H., & Mueller, S. (1982). The evolution of organization forms:
Technology, coordination, and control. Research in Organizational
Behavior, 4, 33-37.
Allen, M., & Stiff, J. (1989). Testing three models for the sleeper effect. Western
Journal of Speech Communication, 53(411-426).
Almeida, P., Song, J., & Grant, R. M. (2002). Are firms superior to alliances and
markets? An empirical test of cross-border knowledge building.
Organization Science, 13(2), 147-161.
222

Alvesson, M. (1995). Management of knowledge intensive companies. Berlin/New
York: de Gruyter.
Alvesson, M. (2000). Social identity and the problem of loyalty in knowledge-
intensive companies. Journal of Management Studies, 37(8), 1101-1123.
Ambrosini, V., & Bowman, C. (2001). Tacit knowledge: Some suggestions for
operationalization. Journal of Management Studies, 38(6), 811-829.
Ancori, B., Bureth, A., & Cohendet, P. (2000). The economics of knowledge: The
debate about codification and tacit knowledge. Industrial and Corporate
Change, 9(2), 255-287.
Appleyard, M. M. (1996). How does knowledge flow? Interfirm patterns in the
semiconductor industry. Strategic Management Journal, 17(Winter special
issue), 137-154.
Ardichivili, A., Page, V., & Wentling, T. (2003). Motivation and barriers to
participation in virtual knowledge-sharing communities of practice. Journal
of Knowledge Management, 7(1), 64-77.
Argote, L. (1999). Organizational learning: Creating, retaining, and transferring
knowledge. Boston: Kluwer Academic Publishers.
Argote, L., & Ingram, P. (2000). Knowledge transfer: A basis for competitive
advantage in firms. Organizational Behavior and Human Decision
Processes, 82(1), 150-169.
Argote, L., Ingram, P., Levine, J. M., & Moreland, R. L. (2000). Introduction:
Knowledge transfer in organizations: Learning from the experience of
others. Organizational Behavior and Human Decision Processes, 82(1), 1-
8.
Armbrecht, F. M. R., Jr., Chapas, R. B., Chappelow, C. C., Farris, F. G., Friga, P.
N., Hartz, C. A., et al. (2001). Knowledge management in research and
development. Research Technology Management, 44(4), 28-48.
223

Aronson, J. (1994). A pragmatic view of thematic analysis. The Qualitative Report,
2(1), http://www.nova.edu/ssss/QR/BackIssues/QR2-1/aronson.html.
Arora, A., & Gambardella, A. (1990). Complementary and external linkages: The
strategies of the large firms in biotechnology. The Journal of Industrial
Economics, 38(4), 361-379.
Arora, A., & Gambardella, A. (1994). Evaluating technological information and
utilizing it-Scientific knowledge, technological capacity, and external
linkages in biotechnology. Journal of Economic Behavior and
Organization, 24, 91-114.
Atler, C., & Hage, J. (1993). Organizations working together. Newbury Park, CA:
Sage.
Audretsch, D. B. (2000). Is university entrepreneurship different? Manuscript.
Audretsch, D. B. (2001). The role of small firms in U.S. biotechnology clusters.
Small Business Economics, 17, 3-15.
Audretsch, D. B., & Stephan, P. E. (1996). Company-scientist locational links: The
case of biotechnology. The American Economic Review, 86(3), 641-652.
Badaracco, J. (1991). Knowledge Link: How firms compete through strategic
alliances. Boston, MA: Harvard Business School Press.
Baptista, R. (1998). Clusters, innovation, and growth: A survey of the literature. In
G. M. P. Swann, M. Prevezer & D. Stout (Eds.), The dynamics of industrial
clustering (pp. 13-51). New York: Oxford University Press.
Barabási, A. L., & Albert, R. (1999). Emergence of scaling in random networks.
Science, 286, 509-512.
224

Barnes, J. A. (1979). Network analysis: Orienting notion, rigorous technique, or
substantive field of study? In P. W. Holland & S. Leinhart (Eds.),
Perspectives on social network analysis (pp. 403-423). New York:
Academic.
Barringer, B. R., & Harrison, J. S. (2000). Walking a tightrope: Creatng value
through interorganizational relationships. Journal of Management, 2(3),
367-403.
Bartholomew, S. (1997). National systems of biotechnology innovation: Complex
interdependence in the global system. Journal of International Business
Studies, 28(2), 241-266.
Batagelj, V., & Mrvar, A. (2007). Pajek: Program for analysis and visualization of
large networks: Reference Manual. Retrieved 2007, May 1, from
http://vlado.fmf.uni-lj.si/pub/networks/pajek/doc/pajekman.pdf
Baum, J. A. C., & Ingram, P. (2002). Interorganizational learning and network
organizations: Toward a behavioral theory of the interfirm. In M. Augier &
J. G. March (Eds.), Economics of choice, change, and organizations. Essays
in the memory of Richard M. Cyert. Cheltenham, UK: Edward Elgar.
Bell, D. (1973). The coming of post-industrial society. New York: Basic Books.
Bell, D. G., Giordano, R., & Putz, P. (2002). Inter-firm sharing of process
knowledge: Exploring knowledge markets. Knowledge and Process
Management, 9(1), 12-22.
Bennet, A., & Bennet, D. (2003). Teh partnership between organizational learning
and knowledge management. In C. W. Holsapple (Ed.), Handbook on
knowledge management 1: Knowledge matters. Germany: Springer.
Benson, J. K. (1975). The interorganizational network as a political economy.
Administrative Science Quarterly, 20, 229-249.
225

Bercovitz, J., Feldman, M., Feller, I., & Burton, R. (2001). Organizational structure
as a determinant of academic patent and licensing behavior: An exploratory
study of Duke, Johns Hopskins, and Pennsylvania State Universities.
Journal of Technology Transfer, 26, 21-35.
Bergeron, B., & Chan, P. (2004). Biotech industry: A global, economic, and
financing overview. Hoboken, NJ: John Wiley & Sons, Inc.
Bienenstock, E. J., & Bonacich, P. (1992). The core as solution to exclusionary
networks. Social Networks, 14, 231-244.
Bienenstock, E. J., & Bonacich, P. (1997). Network exchange as a cooperative
game. Rationality and society, 9, 37-65.
Bierly, P. E., & Chakrabarti, A. (1996). Generic knowledge strategies in the US
pharmaceutical industry. Strategic Management Journal, 17(Winter special
issue), 123-135.
Bierly, P. E., Kessler, E. H., & Christensen, E. W. (2000). Organizational learning,
knowledge, and wisdom. Journal of Organizational Change and
Management, 13(6), 595-618.
Blau, P. M. (1964). Exchange and power in social life. New York: John Wiley.
Blodgett, L. L. (1991). Partner contributions as predictors of equity share in
international joint ventures. Journal of International Business Studies, 22,
63-78.
Boisot, M. H. (1998). Knowledge assets: Securing competitive advantage in the
information economy. Oxford: Oxford University Press.
Bordum, A. (2002). From tacit knowing to tacit knowledge - Emancipation or
ideology? Critical Quarterly, 44(3), 50-54.
226

Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). UCINET for Windows:
Software for Social Network Analysis (Version 6) [Program]. Harvard:
Analytic Technologies.
Borgatti, S. P., & Foster, P. C. (2003). The network paradigm in organizational
research: A review and typology. Journal of Management, 29(6), 991-1013.
Bower, D. J. (1993). Successful joint ventures in science parks. Long Range
Planning, 26(6), 114-120.
Brown, J. S., & Duguid, P. (1998). Organizing knowledge. California Management
Review, 40(3), 90-111.
Burt, R. S. (1983). Corporate profits and cooperation. New York: Academic Press.
Californian Institute of Regenerative Medicine. (2006). Intellectual property policy
for non-profit organizations. Retrieved August 1st, 2006, from
http://www.cirm.ca.gov/policies/pdf/IPPNPO.pdf
Campbell, D. (1974). Evolutionary epistemology. In P. A. Schillp (Ed.), The
philosophy of Karl Popper (Vol. 14-I). LaSalle, IL: Open Court.
Carley, K. (1991). A theory of group stability. American Sociological Review,
56(3), 331-354.
Castells, M. (1996). The rise of the network society. Malden, Mass: Blackwell
Publishers Inc.
Chandler, A. (1962). Strategy and structure: Chapters in the history of the
American industrial enterprise. Cambridge, MA: The MIT Press.
Chi, T. (1994). Trading in strategic resources: Necessary conditions, transaction
cost problems, and choice of exchange structure. Strategic Management
Journal, 15, 271-290.
227

Child, J., & Faulkner, D. O. (1998). Strategies of cooperation: Managing alliances,
networks, and joint ventures. Oxford, England: Oxford University Press.
Choo, C. W. (1998). The knowing organization. Oxford: Oxford University Press.
Chow, C. W., Deng, F. J., & Ho, J. L. (2000). The openness of knowledge sharing
within organizations: A comparative study of the United States and the
People's Republic of China. Journal of Management Accounting Research,
12, 65-95.
Chow, C. W., Harrison, G., McKinnnon, J., & Wu, A. (1999). Cultural influences
on informal information sharing in Chinese and Anglo-American
organizations: An exploratory study. Accounting, Organizations, and
Society, 24, 561-582.
Chow, C. W., Hwang, R., Liao, W., & Wu, A. (1998). National culture and
subordinates' upward communication of private information. International
Journal of Accounting, 33(3), 293-311.
Cockburn, I., & Henderson, R. (1998). Absorptive capacity, coauthoring behavior,
and the organization of research in drug discovery. Journal of Industrial
Economics, 46(2), 157-182.
Cohen, W. M., & Levinthal, D. A. (1989). Innovation and learning: The two faces
of R&D. The Economic Journal, 99, 569-596.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective
on learning and innovation. Administrative Science Quarterly, 35, 128-152.
Contractor, F. J., & Lorange, P. (1988). Why should firms cooperate: The strategy
and economics basis for cooperative ventures. In F. J. Contractor & P.
Lorange (Eds.), Cooperative strategies in international business (pp. 3-30).
New York: Lexington Books.
228

Contractor, F. J., Wasserman, S., & Faust, K. (2006). Testing multi-theoretical
multilevel hypotheses about organizational networks: An analytic
framework and empirical example. Academy of Management Review, 31(3),
681-703.
Cook, K. S. (1977). Exchange and power in networks of interorganizational
relations. Sociological Quarterly, 18, 62-82.
Cook, K. S. (1982). Network structures from an exchange perspective. In P. V.
Marsden & N. Lin (Eds.), Social structure and network analysis (pp. 177-
218). Beverley Hills, CA: Sage.
Cook, K. S., & Whitmeyer, J. M. (1992). Two approaches to social structure:
Exchange theory and network analysis. Annual Review of Sociology, 18,
109-127.
Cook, K. S., & Yamagishi, T. (1992). Power in exchange networks: A power
dependence formulation. Social Networks, 14, 245.
Cooke, P. (2002). Regional innovation systems: General findings and some new
evidence from biotechnology clusters. Journal of Technology Transfer.,
27(1), 133-145.
Cowan, R., David, P. A., & Foray, D. (2000). The explicit economics of knowledge
codification and tacitness. Industrial and Corporate Change, 9(2), 211-253.
Curall, S., & Judge, T. (1995). Measuring trust between organizational boundary
role persons. Organizational Behavior and Human Decision Processes,
64(2), 151-170.
Daizadeh, I., Miller, D., Glowalla, A., Leamer, M., Nandi, R., & Numark, C.
(2002). A general approach for determining when to patent, publish, or
protect information as a trade secret. Nature Biotechnology, 20, 1053-1054.
Dalpe, R. (2003). Interaction between public research organizations and industry in
biotechnology. Managerial and Decision Economics, 23(2,3), 171-185.
229

Das, S., Sen, P. K., & Sengupta, S. (1998). Impact of strategic alliances on firm
evaluation. Academy of Management Journal, 41, 27-41.
Das, S., & Teng, B. (1998). Resource and risk management in the strategic alliance
making process. Journal of Management, 24(21-42).
Davis, G. B., & Olson, M. H. (1985). Management information systems:
Conceptual foundations, structure and development. New York: McGraw
Hill.
Davis, G. F., Yoo, M., & Baker, W. E. (2003). The small world of the American
corporate elite, 1991-2001. Strategic Organization, 1, 301-336.
De Long, D. W., & Fahey, L. (2000). Diagnosing cultural barriers to knowledge
management. Academy of Management Executive, 14(4), 113-127.
Demarest, M. (1997). Understanding knowledge management. Long Range Plan,
30, 374-384.
Denison, D. R. (1996). What is the Difference between Organizational Culture and
Organizational Climate? A Native's Point of View on a Decade of Paradigm
Wars. The Academy of Management Review, 21(3), 619-654.
Di Gregorio, D., & Shane, S. (2000). Why do some universities generate more
start-ups than others? Manuscript.
Dickson, P. H., & Weaver, K. M. (1997). Environmental determinants and
individual-level moderators of alliance use. Academy of Management
Journal, 40, 404-425.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional
isomorphism and collective rationality in organizational fields. American
Sociological Review, 48, 147-160.
230

Dixon, N. M. (2000). Common knowledge: How companies thrive by sharing what
they know. Boston, MA: Harvard Business School Press.
Donaldson, L. (1987). Strategy and structural adjustment to regain fit and
performance in defense of contingency theory. Journal of Management
Studies, 24, 1-24.
Donaldson, L. (1995). American anti-management theories of organization.
Cambridge: Cambridge University Press.
Donaldson, L. (2001). Reflections on knowledge and knowledge-intensive firms.
Human Relations, 54(7), 955-963.
Doz, Y. L. (1996). The evolution of cooperation in strategic alliances: Initial
conditions of learning process. Strategic Management Journal, 17, 55-83.
Doz, Y. L., & Hamel, G. (1998). Alliance advantage. Boston: Harvard Business
Press.
Drotner, K. (2001). Global media through youthful eyes.
Dyer, J. H., & Nobeoka, K. (2000). Creating and maintaining a high-performance
knowledge-sharing network: The Toyota Case. Strategic Management
Journal, 21, 345-367.
Eisenberg, E. M., & Riley, P. (2001). Organizational culture. In F. M. Jablin & L.
L. Putnam (Eds.), The new handbook of organizational communication:
Advances in theory, research, and methods. Thousand Oaks, CA: Sage.
Eisenhardt, K., & Schoomhoven, C. B. (1996). Resource-based view of strategic
alliance formation: Strategic and social effects in entrepreneurial firms.
Organization Science, 7, 136-150.
Faulkner, D. O. (1995). International strategic alliances: Co-operating to compete.
Maidenhead, UK: McGraw-Hill.
231

Feldman, M., Feller, I., Bercovitz, J., & Burton, R. (2000). Equity and the
technology transfer strategies of major American research universities.
Manuscript.
Fisher, L. M. (1996). How strategic alliances work in biotech. Strategy and
Business, First quarter, 1-7.
Flanagan, C. J. (1954). The critical incident technique. Psychological Bulletin, 51,
327-358.
Fontana, A., & Frey, J. (2000). The interview: From structured questions to
negotiated text. In N. K. Denzin & Y. Lincoln (Eds.), Handbook of
qualitative research (pp. 645-672). Thousand Oaks: CA: Sage.
Frost, P. J., Moore, L. F., Louis, M. R., Lundberg, G. C., & Martin, J. (Eds.).
(1991). Reframing Organizational Culture. Thousand Oaks: CA: Sage.
Galaskiewicz, J., & Bielefeld, W. (1998). Nonprofit organizations in an age of
uncertainty: A study of organizational change. New York: Aldine de
Gruyter.
Gambardella, A. (1995). Science and innovation: The US pharmaceutical industry
during the 1980s. New York: Cambridge University Press.
Geiger, R. (2004). Research & relevant knowledge : American research
universities since World War II. New Brunswick, N.J.: Transaction
Publishers.
George, G., Zahra, S. H., & Wood, D. R. (2002). The effects of business-university
alliances on innovative outputs and financial performance: A study of
publicly traded biotechnology companies. Journal of Business Venturing,
17, 577-609.
Gertler, M. S., & Levitte, Y. M. (2005). Local nodes in global networks: The
geography of knowledge flows in biotechnology innovation. Industry and
Innovation, 12(4), 487-507.
232

Gibson, C. B., & Zellmer-Bruhn, M. E. (2001). Metaphors and meaning: An
intercultural analysis of the concept of teamwork. Administrative Science
Quarterly, 46, 274-303.
Gittelman, M., & Kogut, B. (2003). Does good science lead to valuable
knowledge? Biotechnology firms and evolutionary logic of citation patterns.
Management Science, 49(4), 366-382.
Goffe, R., & Jones, G. (1996). What holds the modern company together? Harvard
Business Review, 74(6).
Golub, E. (2003). Generating spin-offs from university-based research: The
potential of technology transfer. Unpublished PhD dissertation, Columbia
University, New York, NY.
Grabowski, H., & Vernon, J. (1994). Innovation and structural change in
pharmaceutical and biotechnology. Industrial and Corporate Change, 3,
435-439.
Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology,
78(6), 1360-1380.
Granovetter, M. (1985). Economic action and social structure: A theory of
embeddedness. American Journal of Sociology, 91, 481-510.
Gulati, R. (1995a). Does familiarity breeds trust? The implications of repeated ties
for contractual choice in alliances. Academy of Management Journal, 38(1),
85-112.
Gulati, R. (1995b). Social structure and alliance formation patterns: A longitudinal
study. Administrative Science Quarterly, 40(4), 619-652.
Gulati, R., & Garguilo, M. (1999). Where do interorganizational networks come
from? The American Journal of Sociology, 104(5), 1439-1493.
233

Gupta, A. K., & Lad, L. J. (1983). Industry self-regulation: An economic,
organizational, and political analysis. Academy of Management Review, 8,
416-425.
Gupta, B., Iyer, L. S., & Aronson, J. (2000). Knowledge management: Practices
and challenges. Industrial Management and Data System, 100(1/2), 17-21.
Hagedoorn, J. (1993a). Strategic technology alliances and modes of cooperation in
high-technology industries. In G. Grabher (Ed.), The embedded firm: On the
socioeconomics of industrial network (pp. 116-137). London and New
York: Routledge.
Hagedoorn, J. (1993b). Understanding the rationale of strategic partnering:
Interorganizational modes of cooperation and sectoral differences. Strategic
Management Journal, 14, 371-385.
Hagedoorn, J., & Schankenraad, J. (1994). The effect of strategic technology
alliances on company performance. Strategic Management Journal, 15,
291-309.
Hamel, G. (1991). Competition for competence and inter-partner learning within
international strategic alliances. Strategic Management Journal, 12, 83-103.
Hamel, G., Doz, Y. L., & Prahaland, C. K. (1989). Collaborate with your
competitors and win. Harvard Business Review, 89(1), 133-139.
Han, Y. S. (2004). Is technology preemption a motive for R&D alliances? Evidence
from the biotechnology industry. Journal of American Academy of
Business, 5(1/2), 215-219.
Hannen, M., & Freeman, L. C. (1984). Structural inertia and organizational change.
American Sociological Review, 49, 149-164.
Hansen, C. D., & Kahnweiler, W. M. (1993). Storytelling: an instrument for
understanding the dynamics of corporate relationships. Human Relations,
46(12), 1391-1409.
234

Hardt, M., & Negri, A. (2000). Empire. Cambridge, MA: Harvard University Press.
Harrigan, K. R. (1986). Managing for joint venture success. Lexington, MA:
Lexington Books.
Haunschild, P. R. (1993). Interorganizational imitation: The impact of interlocks on
corporate acquisition activity. Administrative Science Quarterly, 38, 564-
592.
Henderson, R., Jaffe, A., & Trajtenberg, M. (1998). Universities as a source of
commercial technology: A detailed analysis of university patenting, 1956-
1998. Review of Economics and Statistics, 80(1), 119-127.
Hennart, J. F. (1988). A transaction costs theory of equity joint ventures. Strategic
Management Journal, 9(361-374).
Hirschman, A. O. (1970). Exit, voice, and loyalty. Cambridge, MA: Harvard
University Press.
Hoffman, N., & Klepper, R. (2000). Assimilating new technologies: The role of
organizational culture. Information Systems Management., 17(3), 36-42.
Hofstede, G. (1980). Culture's consequences: International differences in work-
related values. Beverly Hills, CA: Sage.
Hollingshead, A. B. (1998). Communication, learning, and retrieval in transactive
memory systems. Journal of Experimental Social Psychology, 34, 423-442.
Hutchins, H. (1996). Cognition in the wild. Cambridge, MA: MIT Press.
Inkpen, A., & Crossan, M. M. (1995). Believing is seeing: Joint ventures and
organizational learning. Journal of Management Studies, 32, 595-618.
235

Inkpen, A., & Dinur, A. (1998). Knowledge management process and international
joint ventures. Organization Science, 9(4), 454-468.
Inkpen, A., & Li, K. Q. (1999). Joint venture formation: Planning and knowledge-
gathering for success. Organizational Dynamics, 27(4), 33-47.
Iverson, J. O., & McPhee, R. D. (2002). Knowledge management in community of
practice: Being true to the communicative character of knowledge.
Management Communication Quarterly, 16(2), 259-266.
Jaffe, A. (1989). The real effects of academic research. American Economic
Review, 79(5), 957-970.
Jaffe, A., & Trajtenberg, M. (1999). International knowledge flows: evidence from
patent citations. Economics of Innovation and New Technology, 8(1/2), 105-
136.
Jaffe, A., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of
knowledge spillovers by patent citations. Quarterly Journal of Economics,
577-598.
Jarillo, J. C. (1988). On strategic networks. Strategic Management Journal, 9, 31-
41.
Jensen, R., & Thursby, M. C. (1998). Proofs and prototypes for sale: The tale of
university licensing. NBER Working Paper 6698.
Juanillo, N. K. (2001). The risks and benefits of agricultural biotechnology. The
American Behavioral Scientist, 44(8), 1246-1266.
Kale, R., Singh, H., & Perlmutter, H. (2000). Learning and protection of
proprietary assets in strategic alliances: Building relational capital. Strategic
Management Journal, 21(Special issue), 2137-2237.
236

Kalling, T., & Styhre, A. (2003). Knowledge sharing in organizations.
Copenhagen, Denmark: Copenhagen Business School Press.
Kanter, R. M. (1989). When giants learn to dance. New York: Simon and Schuster.
Kaplan, S. N., & Milton, B. A. (1994). Appointment of outsiders to Japanese
boards: Determinants and implications for managers. Journal of Financial
Economics, 36, 766-794.
Kenny, M. (1986). Biotechnology: The university-industrial complex. New Haven:
Yale University Press.
Kim, T. Y., Oh, H., & Swaminathan, A. (2006). Framing interorganizational
network change: A network inertia perspective. Academy of Management
Review, 31(3), 704-720.
Knapp, E., & Yu, D. (1999). Understanding organizational culture: How culture
helps or hinders the flow of knowledge. Knowledge Management Review, 7,
16-21.
Kogut, B. (1988). Joint ventures: Theoretical and empirical perspectives. Strategic
Management Journal, 9(310-332).
Kogut, B. (1989). The stability of joint ventures: Reciprocity and competitive
rivalry. The Journal of Industrial Economics, 38, 183-198.
Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities,
and the replication of technology. Organization Science, 3(3), 383-397.
Koh, J., & Venkatraman, N. (1991). Joint venture formations and stock market
reactions: Assessment in the information technology sector. Academy of
Management Journal, 34, 869-892.
237

Kostova, T., & Roth, K. (2002). Adoption of an organizational practice by
subsidiaries of multinational corporations: Institutional and relational
effects. Academy of Management Journal, 45(1), 215-233.
Koza, M. P., & Lewin, A. Y. (1998). The co-evolution of strategic alliances.
Organization Science, 9, 255-264.
Krackhardt, D. (1988). Predicting with networks: Nonparamentric multiple
regression analysis of dyadic data. Social Networks, 10, 359-381.
Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives,
enduring questions. Annual Review of Psychology, 50, 569-598.
Larson, A. (1991). Partner networks: Leveraging external ties to improve
entrepreneurial performance. Journal of Business Venturing, 6, 173-188.
Laumann, E. O., Galaskiewicz, J., & Marsden, P. V. (1978). Community structure
as interorganizational linkages. Annual Review of Sociology, 4, 455-484.
Lazega, E., & Pattison, P. (1999). Social capital, multiplex generalized exchanges
and cooperation in organizations: A case study. Social Networks, 21, 67-90.
Leonard, D., & Sensiper, S. (1998). The role of tacit knowledge in group
innovation. California Management Review, 40(3), 112-132.
Lerner, J., & Merges, R., P. (1998). The control of technology alliances: An
empirical analysis of the biotechnology industry. The Journal of Industrial
Economics, 46(2), 125-156.
Liebeskind, j. P., Oliver, A. L., Zucker, L. G., & Brewer, M. B. (1996). Social
network, learning, and flexibility: Sourcing scientific knowledge in new
biotechnology firms. Organization Science, 7(4), 428-443.
238

Liebeskind, j. P., Oliver, A. L., Zucker, L. G., & Brewer, M. B. (1998). From
handshake to contract: Trust, intellectual property, and the social structure
of academic research. In C. Lane & R. Bachmann (Eds.), Trust within and
between organizations (pp. 118-145). New York: Oxford University Press.
Lim, K. (2000). The many faces of absorptive capacity: Spillovers of copper
interconnect technology for semiconductor chips. (MIMEO, MIT).
Liyanage, S., & Barnard, R. (2003). Valuing of firm's prior knowledge: A measure
of knowledge distance. Knowledge and Process Management, 10(2), 85-98.
Lomi, A. (1997). Markets with hierarchies and the network structure of
organizational communities. Journal of Management and Governance, 1,
49-66.
Lopez, S. P., Peon, J. M. M., & Ordas, C. J. V. (2004). Managing knowledge: The
link between culture and organizational learning. Journal of Knowledge
Management, 8(6), 93-104.
Lorrain, F., & White, H. C. (1971). The structural equivalence of individuals in
social networks. Journal of Mathematical Sociology, 1, 49-80.
Lowe, R. (2002). Invention, innovation, and entrepreneurship: The
commercialization of university research by inventor-founded firms.
Unpublished Phd dissertation, University of California, Berkeley, Berkeley,
CA.
Macaulay, S. (1963). Non-contractual relations in business. American Sociological
Review, 28, 55-70.
Macneil, I. R. (1978). Contracts: Adjustments of long-term economic relations
under classical, neoclassical, and relational contract law. Northwestern
University Law Review, 72, 854-906.
March, J. G. (1991). Exploration and exploitation in organizational learning.
Organization Science, 2(71-87).
239

March, J. G., & Simon, H. A. (1958). Organizations: Wiley.
Mariti, P., & Smiley, R. H. (1983). Co-operative agreements and the organization
of industry. Journal of Industrial Economics, 31, 437-451.
Markovsky, B., Willer, D., & Patton, T. (1988). Power relations in exchange
networks. American Sociological Review, 53, 220-236.
Marrett, C. B. (1971). On the specification of interorganizational dimensions.
Sociology and Social Research, 56(1), 83-99.
Martin, J. (1982). Stories and scripts in organizational settings. In A. H. Hastorf &
A. M. Isen (Eds.), Cognitive social psychology (pp. 255-305). New York:
Elsevier.
Martin, J. (1992). Cultures in organizations: Three perspectives. New York:
Oxford University Press.
Martin, J. (2002). Organizational culture: Mapping the terrain. Thousand Oaks,
CA: Sage.
Marwell, G., Oliver, P. E., & Prahl, R. (1988). Social networks and collective
action: A theory of the collective mass. III. American Journal of Sociology,
94, 502-534.
Matusik, S. F., & Hill, C. W. L. (1998). The utilization of contingent work,
knowledge creation, and competitive advantage. Academy of Management
Review, 23(4), 680-697.
Mayer, R., Davis, J., & Schoorman, F. (1995). An integrative model of
organizational trust. Academy of Management Review, 20, 709-734.
McLure, M., & Faraj, S. (2000). It is what one does: Why people participate and
help others in electronic communities of practice. The Journal of Strategic
Information Systems, 9(2-3), 55-173.
240

Merton, R. (1973). The sociology of Science. Chicago, IL: Chicago University
Press.
Meyer, J. W., & Rowan, B. (1977). Institutional organizations: Formal structure as
myth and ceremony. American Journal of Sociology, 83(2), 340-363.
Meyer, J. W., & Scott, W. R. (1983). Organizational environments: Ritual and
rationality. Beverly Hills, CA: Sage.
Meyerson, D. E. (1991). "Normal" ambiguity? A glimpse of an occupational
culture. In P. J. Frost, L. F. Moore, M. R. Louis, G. C. Lundberg & J.
Martin (Eds.), Reframing Organizational Culture. (pp. 131-144). Thousand
Oaks: CA: Sage.
Mizruchi, M. S. (1982). The American corporate network: 1904-1974. Beverly
Hills, CA: Sage.
Mizruchi, M. S. (1996). What do interlocks do? An analysis, critique, and
assessment of research on interlocking directorates. Annual Review of
Sociology, 22, 271-298.
Mizruchi, M. S., & Yoo, M. (2002). Interorganizational power and dependence. In
J. A. C. Baum (Ed.), The Blackwell Companion to Organizations (pp. 599-
620). Oxford: Blackwell.
Monge, P., & Contractor, N. (2003). Theories of communication networks. New
York, NY: Oxford University Press.
Monge, P., Fulk, J., Kalman, M. E., Flanagin, A. J., Parnassa, C., & Rumsey, S.
(1998). Production of collective action in alliance-based interorganizational
communication and information systems. Organization Science, 9(3), 411-
433.
241

Mowery, D., Nelson, R. R., Sampat, B., & Ziedonis, A. A. (1999). The effects of
the Bayh-Dole Act on U.S. University research and technology transfer: An
analysis of data from Columbia University, the University of California,
and Stanford University. Research Policy.
Murray, F. (2002). Innovation as the co-evolution of scientific and technological
networks: Exploring tissue engineering. Research Policy, 31, 1389-1403.
National Research Council. (1994). Information technology in the service society.
National Science Board. (2004). Science and engineering indicators 2004.
Arlington, VA: National Science Foundation.
National Science Foundation. (2003). Federal obligations to universities and
colleges for total research, by detailed field of science & engineering and
selected agency:. Retrieved March 25th, 2006, from
http://www.nsf.gov/statistics/nsf03325/tables/hist62.xls
Nelson, R. R. (2001). Observations on the post-Bayh-Dole rise of patenting at
American universities. Journal of Technology Transfer, 26, 13-19.
Newman, M. E. J. (2001). The structure of scientific collaboration networks.
Proceedings of the National Academy of Science of the United States of
America, 98(2), 404-409.
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation.
Organization Science, 5(1), 14-37.
Nonaka, I., & Konno, N. (1998). The concept of "Ba": Building a foundation for
knowledge creation. California Management Review, 40(3), 40-54.
Nonaka, I., & Takeuchi, H. (1995). The knowledge-creating company. New York:
Oxford University Press.
242

Nonaka, I., Toyama, R., & Byosiere, P. (2001). A theory of organizational
knowledge creation: Understanding the dynamic process of creating
knowledge. In M. Dierkes, A. B. Anal, J. Child & I. Nonaka (Eds.),
Handbook of organizational learning and knowledge (pp. 491-517).
Oxford, UK: Oxford University Press.
O'Dell, C., & Grayson, C. J. (1998). If we only knew what we know: Identification
and transfer of internal best practices. California Management Review,
40(3), 154-174.
Ohmae, K. (1987). The triad world view. Journal of Business Strategies, 7(4), 8-19.
Ohmae, K. (1989). The global logic of strategic alliances. Harvard Business
Review, Marsh/April, 143-154.
Oliver, C. (1990). Determinants of interorganizational relationships: Integration
and future directions. Academy of Management Review, 15(2), 241-265.
Owen-Smith, J., & Powell, W. W. (2001a). To patent or not: Faculty decisions and
institutional success at technology transfer. Journal of Technology Transfer,
26(1/2), 99-114.
Owen-Smith, J., & Powell, W. W. (2001b). To patent or not: Faculty decisions and
institutional success at technology transfer. Journal of Technology Transfer,
26(1/2), 99-114.
Owen-Smith, J., & Powell, W. W. (2004). Knowledge networks as channels and
conduits: The effects of spillovers in the Boston Biotechnology
Community. Organization Science.
Oxley, J. E. (2004). Learning versus protection in inter-firm alliances: A false
dichotomy. In V. Mahnke & T. Pedersen (Eds.), Knowledge flows,
governance and multinational enterprise (pp. 108-147). New York, NY:
Palgrave Macmillan.
243

Perakyla, A. (1997). Reliability and validity in research based on tapes and
transcripts. In D. Silverman (Ed.), Qualitative research: Theory, method
and practice (pp. 201-220). Thousand Oaks, CA: Sage.
Pfeffer, J., & Salancik, C. (1978). The external control of organizations. New
York: Harper and Row.
Phene, A., & Tallman, S. (2002). Knowledge flows and geography in
biotechnology. International Journal of Medical Marketing, 2(3), 241-254.
Podolny, J. M. (1994). Market uncertainty and the social character of economic
exchange. Administrative Science Quarterly, 39, 458-483.
Podolny, J. M., Hannen, M., & Stuart, T. E. (1996). Networks, knowledge and
niches: Competition in the world-wide semiconductor industry (1984-
1991). American Journal of Sociology, 102(3), 659-689.
Polanyi, M. (1958). Personal knowledge: Towards a post-critical philosophy.
London: Routledge & Kegan Paul.
Powell, W. W. (1990). Neither market nor hierarchy: Network forms of
organization. Research in Organizational Behavior, 12, 295-336.
Powell, W. W. (1998). Learning from collaboration: Knowledge and networks in
biotechnology and pharmaceutical industries. California Management
Review, 40(3), 228-240.
Powell, W. W., & Brantley, P. (1992). Competitive cooperation in biotechnology:
Learning through networks. In N. Nohira & R. G. Eccles (Eds.), Networks
in organizations. Boston, MA: Harvard Business Review Press.
Powell, W. W., Koput, K. W., & Smith-Doerr, L. (1996). Interorganizational
collaboration and the locus of innovation: Networks of learning in
biotechnology. Administrative Science Quarterly, 41, 116-145.
244

Powell, W. W., White, D. R., Koput, K. W., & Owen-Smith, J. (2005). Network
dynamics and field evolution: The growth of interorganizational
collaboration in life sciences. American Journal of Sociology, 110(4), 1132-
1205.
Quere, M. (2003). Knowledge dynamics: Biotechnology's incursion into the
pharmaceutical industry. Industry and Innovation, 10(3), 255-273.
Quere, M., & Saviotti, P. P. (2002, June 6-8, 2002). Knowledge dynamics and the
organization of the life science industries. Paper presented at the DRUID
Summer Conference on Industrial Dynamics of the New and Old Economy-
Who is embracing whom?, Copenhagen/Elsinore.
Quinn, J. B., Anderson, P., & Finkelstein, S. (1996). Managing professional
intellect: Making the most of the best. Harvard Business Review, March-
April, 71-82.
Rand. (2003). Technology transfer of federally funded R&D. Santa Monica, CA:
Rand.
Ratliff, A. (2003). Biotechnology and pharmaceutical R&D and licensing trends:
You pay your money and takes your chances. Journal of Commercial
Biotechnology, 10(1), 54-59.
Ring, P. S., & Van De Ven, A. H. (1994). Developmental processes of cooperative
interorganizational relationships. Academy of Management Review, 19(1),
90-118.
Roberts, E. (1991). Entrepreneurs in high technology. New York, NY: Oxford
University Press.
Robins, G., & Pattison, P. (2006). Multiple networks in organisations. Retrieved
2006, March 10, from
http://www.sna.unimelb.edu.au/publications/publications.html
245

Robins, G., Pattison, P., & Woolcock, J. (2005). Small and other worlds: Global
network structure from local processes. American Journal of Sociology,
110(4), 894-936.
Rockwood, A. (1983). The impact of joint ventures on the market for OCS oil and
gas lease. Journal of Industrial Economics, 31, 452-468.
Rogers, E. (1983). Diffusion of innovations. New York: Free Press.
Rothaermel, F. T., & Deeds, D. L. (2004). Exploration and exploitation alliances in
biotechnology: A system of new product development. Strategic
Management Journal, 25, 201-221.
Rumelt, R. P. (1974). Strategy, structure and economic performance. Boston:
Division of research, Graduate School of Business Administration, Harvard
University.
Ruppel, C. P., & Harrington, S. J. (2001). Sharing knowledge through intranets: A
study of organizational culture and intranet implementation. IEEE
Transactions on Professional Communication, 44(1), 37-52.
Ryle, G. (1949). The concept of mind. Chicago: University of Chicago Press.
Santoro, M. D., & Gopalakrishnan, S. (2001). Relationship dynamics between
university research centers and industrial firms: Their impact on technology
transfer activities. Journal of Technology Transfer, 26, 163-171.
Saxenian, A. (1994). Regional advantage: Culture and competition in Silicon
Valley and Route 128. Cambridge, MA: Harvard University Press.
Schacter, B. (1999). Issues and dilemmas of biotechnology. Westport, Connecticut:
Greenwood Press.
246

Schein, E. (1991). What is culture? In P. J. Frost, L. F. Moore, M. R. Louis, G. C.
Lundberg & J. Martin (Eds.), Reframing Organizational Culture. (pp. 243-
253). Thousand Oaks: CA: Sage.
Schein, E. (2000). Don't make culture another item on the KM checklist.
Knowledge Management Review, 3(4), 8-9.
Schermerhorn, J. R. (1975). Determinants of interorganizational cooperation.
Academy of Management Journal, 18(4), 846-856.
Schumpeter, J. A. (1947). Capitalism, socialism, and Democracy. New York:
Harper & Row.
Scott, W. R. (1992). Organizations: Rational, natural and open systems (3rd ed.).
Englewood Cliffs, NJ: Prentice Hall.
Segil, L. (1998). Strategic alliances for the 21st century. Strategy and Leadership,
September/October, 12-16.
Senge, P. (1990). The fifth discipline: The age and practice of the learning
organization. London, UK: Century Business.
Shahi, G. (2004). Biobusiness in Asia. Singapore: Pearson Prentice Hall.
Shan, W., Walker, B., & Kogut, B. (1994). Interfirm cooperation and startup
innovation in the biotechnology industry. Strategic Management, 15, 387-
394.
Shane, S. (2004). Academic entrepreneurship: University spinoffs and wealth
creation. Cheltenham, IK: Edward Elgar.
Shaprio, C., & Willing, R. D. (1990). On the antitrust treatment of production joint
ventures. Journal of Economic Perspective, 4(3), 113-130.
247

Siegel, D. S., Thursby, J. G., Thursby, M. C., & Ziedonis, A. A. (2001a).
Organizational issues in university-industry technology transfer: An
overview of the symposium issue. Journal of Technology Transfer., 26, 5-
11.
Siegel, D. S., Thursby, J. G., Thursby, M. C., & Ziedonis, A. A. (2001b).
Organizational issues in university-industry technology transfer: An
overview of the symposium issue. Journal of Technology Transfer, 26, 5-
11.
Smircich, L. (1983). Concepts of culture and organizational analysis.
Administrative Science Quarterly, 28, 339-358.
Smith-Doerr, L., & Powell, W. W. (2005). Networks and economic life. In N. J.
Smelser & R. Swedberg (Eds.), The handbook of economic sociology. (pp.
379-402): Princeton University Press.
Sobocinski, P. (1999). Creating hi-tech business growth in Wisconsin: University
of Wisconsin-Madison technology transfer and entrepreneurship. Madison,
WI: University of Wisconsin System Board of Regents.
Srivastava, S., & Barrett, F. J. (1988). The transforming nature of metaphors in
group development: A study in group theory. Human Relations, 41, 31-64.
Stenmark, D. (2000-2001). Leveraging tacit organizational knowledge. Journal of
Management Information Systems, 17(3), 9-24.
Stinchcombe, A. L. (1990). Weak structural data (Review of Mizruchi and
Schwartz). Contemporary Sociology, 19, 380-382.
Stuart, T. E. (2000). Interorganizational alliance and the performance of firms: A
study of growth and innovation rates in a high-technology industry.
Strategic Management Journal, 21, 791-811.
Szulanski, G. (1996). Exploring internal stickiness: Impediments to the transfer of
best practice within the firm. Strategic Management Journal, 17, 27-43.
248

Szulanski, G. (2003). Sticky knowledge: Barriers to knowing in the firm. Thousand
Oaks, CA: Sage.
Szulanski, G., Cappetta, R., & Jensen, R. J. (2004). When and how trustworthiness
matters: Knowledge transfer and the moderating effect of causal ambiguity.
Organization Science, 15(5), 600-613.
Teece, D. J. (2000). Managing intellectual capital: Organizational, strategic and
policy dimensions. Oxford: Oxford University Press.
Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill.
Thursby, J. G., Jensen, A. R., & Thursby, M. C. (2001). Objectives, characteristics
and outcomes of university licensing: A survey of major U.S. universities.
Journal of Technology Transfer, 26, 59-72.
Thursby, J. G., & Thursby, M. C. (2000). Who is selling the ivory tower? Sources
of growth in university licensing. NBER Working Paper 7718.
Thursby, J. G., & Thursby, M. C. (2001). Industrial perspectives on licensing
university technologies: Sources and problems. Journal of Association of
University Technology Managers, 13.
Tolbert, P. S., & Zucker, L. G. (1983). Institutional sources of change in the formal
structure of organizations: The diffusion of civil service reform, 1880-1935.
Administrative Science Quarterly, 28(22-39).
Tsai, W., & Ghoshal, S. (1998). Social capital and value creation: The role of
intrafirm networks. Academy of Management Journal, 41(4), 464-476.
Tyler, T. R., & Degoey, P. (1996). Trust in organizational authorities: The
influence of motive attributions in willingness to accept decisions. In R. M.
Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory
and research. Thousand Oaks, CA: Sage.
249

Uzzi, B. (1996). The sources and consequences of embeddedness for the economic
performance of organizations. American Sociological Review, 61, 674-698.
Uzzi, B. (1997). Social structure and competition in interfirm networks: The
paradox of embeddedness. Administrative Science Quarterly, 42, 35-67.
Van Maanen, J. (1991). The smile factory: Work at Disneyland. In P. J. Frost, L. F.
Moore, M. R. Louis, G. C. Lundberg & J. Martin (Eds.), Reframing
Organizational Culture. (pp. 58-76). Thousand Oaks: CA: Sage.
Von Hippel, E. (1988). The sources of innovation. New York: Oxford University
Press.
Von Krogh, G., Ichijo, K., & Nonaka, I. (2000). Enabling knowledge creation:
How to unlock the mystery of tacit knowledge and release the power of
innovation. New York: Oxford University Press.
Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and
applications. Cambridge: Cambridge University Press.
Webb, G. (1996). Trust and crises. In R. M. Kramer & T. R. Tyler (Eds.), Trust in
organizations: Frontiers of theory and research. Thousand Oaks, CA: Sage.
Weick, K. E. (1979). The social psychology of organizing. Reading, MA: Addison-
Wesley.
Weiss, L. (1999). Collection and connection: The anatomy of knowledge sharing in
professional service firms. Organization Development Journal, 17(4), 61-
77.
Wenger, E. (1998). Communities of practice: Learning, meaning and identity. New
York: Cambridge University Press.
Wenger, E., McDermott, R., & Snyder, W. M. (2002). Cultivating community of
practice. Boston, MA: Harvard Business School Press.
250

Wenger, E., & Snyder, W. M. (2000). Community of practice: The organizational
frontier. In Harvard business review on organizational learning (pp. 1-20).
Cambridge, MA: Harvard Business School Press.
Wengraf, T. (2001). Qualitative research interviewing. London: Thousand Oaks.
Wiewel, W., & Hunter, A. (1985). The interorganizational network as a resource: A
comparative case study on organizational genesis. Administrative Science
Quarterly, 30, 482-496.
Wilber, K. (2000). Integral psychology. Boston: Shambhala.
Willer, D., & Skvoretz, J. (1997). Network connection and exchange ratios:
Theory, predictions, and experimental tests. Advances in group process, 14,
199-234.
Williamson, O. E. (1975). Markets and hierarchies: Analysis and antitrust
implications. New York, NY: The Free Press.
Williamson, O. E. (1979). Transaction-cost economics: The governance of
contractual relations. Journal of Law and Economics, 22, 233-261.
Williamson, O. E. (1985). The economic institutions of capitalism. New York, NY:
The Free Press.
Williamson, O. E. (1991). Comparative economic organization: The analysis of
discrete structural alternatives. Administrative Science Quarterly, 36, 269-
296.
Williamson, O. E. (1993). Opportunism and its critics. Managerial and Decision
Economics, 14, 97-107.
Winter, S. G. (1987). Knowledge and competence as strategic assets. In D. J. Teece
(Ed.), The competitive challenge: Strategies for industrial innovation and
renewal. (pp. 159-184). Cambridge, MA: Ballinger.
251

Yamagishi, T., Gillmore, M. R., & Cook, K. S. (1988). Network connections and
the distribution of power in exchange networks. American Journal of
Sociology, 93, 833-851.
Zander, U., & Kogut, B. (1995). Knowledge and the speed of the transfer and
imitation of organizational capabilities: An empirical test. Organization
Science, 6, 76-92.
Zucker, L. G. (1987). Institutional theories of organization. Annual Review of
Sociology, 13, 443-464.
Zucker, L. G., Darby, M. R., & Armstrong, J. S. (1998). Geographically localized
knowledge: Spillovers or markets? Economic Inquiry, 36(1), 65-86.
Zucker, L. G., Darby, M. R., & Armstrong, J. S. (2002). Commercializing
knowledge: University science, knowledge capture, and firm performance
in biotechnology. Management Science, 48(1), 138-153.
Zucker, L. G., Darby, M. R., & Brewer, M. B. (1998). Intellectual human capital
and the birth of U.S biotechnology enterprises. The American Economic
Review, 88(1), 290-306.

252

Appendices
Appendix A: Interview protocol-industry
Section 1: Goals/purpose of academic-industry knowledge sharing
1. Has your company been engaged in knowledge sharing with academic institutes
such as universities?

2. Why does your company decide to engage in academic-industry knowledge
sharing? What are your goals?

3. In the case of your company, how well have these goals been achieved through
academic-industry knowledge sharing?

4. In your opinion, why do academic institutes such as universities want to share
knowledge with commercial companies?

Section 2: Channels of academic-industry knowledge sharing
Some channels of university-industry knowledge sharing include patent licensing,
strategic alliances, spin-offs, and personal/informal communication.

5. How many of these channels of academic-industry knowledge sharing are
adopted in your company?

6. When your company is partnering with academic institutes or researchers how
do you determine which channel to use, in different circumstance/contexts? Give an
example.

7. Do you identify any other channels of knowledge sharing that we have not
discussed so far?

Section 3: IP issues
8. Biotechnology industry is an industry where intellectual property is of
paramount importance and is strictly guarded. How do you delineate between
knowledge that can be shared and knowledge that cannot be shared?

9. In your own engagement in academic-industry knowledge sharing in
biotechnology industry, how concerned are you about the possibility that your
253

collaborators will take advantage of your IP? How much do you trust them? Or is it
a problem at all?

Section 4: Effectiveness of academic-industry knowledge sharing
10. From your own experience, how would you rate the effectiveness of academic-
industry knowledge sharing of your company on a scale of 1 to 5 with 1 meaning
extremely ineffective and 5 meaning extremely effective?

11. In your opinion, what are the major barriers to university-industry knowledge
sharing in biotech sector? (Ask respondents to discuss each of them briefly)

12. Based on your own experience, could you please give me one example of most
successful incident of academic-industry knowledge sharing and one example of
least successful incident of academic-industry knowledge sharing? Why do you
think it is a success/failure?

13. In your opinion, what could biotech companies do to facilitate effective
knowledge sharing?

14. In your opinion, what could academic institutes/researchers/scientists do to
facilitate effective knowledge sharing?

Section 5: Culture
People often talk about the culture of academia and industry. By culture, we usually
mean the shared value and ways of conduct by a group of people. For instance,
people frequently talk about the culture of a country, the culture of a profession, or
the culture of an organization.

15. In your opinion, what are the similarities and differences between the culture of
academia and the culture of industry?

16. Do you see cultural difference you mentioned above affecting the process and
outcome of academic-industry knowledge sharing? If so, can you tell me an
example where cultural difference impacted the outcome of university-industry
knowledge sharing?

Section 6: Informal knowledge sharing
Researchers have noticed that people also share personal/individualized knowledge,
such as technical know-how, through informal knowledge sharing channels.
254

17. What is the role of informal/personal communication in academic-industry
knowledge sharing? How would you rate its importance on a scale from 1 to 5 with
1 meaning not significant at all and 5 meaning extremely significant?

18. Based on your own experience in academic-industry knowledge sharing, what
kind of knowledge do you share through informal channels. (I mean other than the
sharing of knowledge through formal channels such as patent licensing)

19. Could you give me an example of informal knowledge sharing?

Section 7: The use of communication technology
20. Have you participate in academic-industry knowledge sharing across long
distance, such as across national borders? If so, in your opinion, how is it similar to
or different from knowledge sharing within a local context? Is long distance a
barrier to successful knowledge sharing? What kinds of communication
technologies do you use in this process? In your opinion, how well do these
communication technologies replace the need for face-to-face communication?

255

Appendix B: Interview protocol-academia
Section 1: Goals/purpose of academic-industry knowledge sharing
1. Could you tell me something about your experience with regard to
industry/academia?

2. In your opinion, why do biotech companies decide to engage in academic-
industry knowledge sharing? What are their goals?

3. How well have these goals been achieved through academic-industry knowledge
sharing?

4. In your opinion, why do academic institutes such as universities want to share
knowledge with commercial companies?

5. Based on your experience, how well have these goals been achieved through
university-industry knowledge sharing?

Section2: Channels of academic-industry knowledge sharing
Some channels of university-industry knowledge sharing include patent licensing,
strategic alliances, spin-offs, and personal/informal communication.

6. How many of the channels of academic-industry knowledge sharing are adopted
by your university/institute that you are aware of?

7. Do you identify any other channels of knowledge sharing that we have not
discussed so far?

Section 3: IP issues
8. Biotechnology industry is an industry where intellectual property is of
paramount importance and is strictly guarded. How do you delineate between
knowledge that can be shared and knowledge that cannot be shared?

9. In your own engagement in academic-industry knowledge sharing in
biotechnology industry, how concerned are you about the possibility that your
collaborators will take advantage of your IP? How much do you trust them? Or is it
a problem at all?

Section 4: Effectiveness of academic-industry knowledge sharing
256

10. From your own experience, how would you rate the effectiveness of academic-
industry knowledge sharing of your university/your research group/yourself on a
scale of 1 to 5 with 1 meaning extremely ineffective and 5 meaning extremely
effective?

11. In your opinion, what are the major barriers to university-industry knowledge
sharing in biotech sector? (Ask respondents to discuss each of them briefly)

12. Based on your own experience, could you please give me one example of most
successful incident of academic-industry knowledge sharing and one example of
least successful incident of academic-industry knowledge sharing? Why do you
think it is a success/failure?

13. In your opinion, what could biotech companies do to facilitate effective
knowledge sharing?

14. In your opinion, what could academic institutes/researchers/scientists do to
facilitate effective knowledge sharing?

Section 5: Culture
People often talk about the culture of academia and industry. By culture, we usually
mean the shared value and ways of conduct by a group of people. For instance,
people frequently talk about the culture of a country, the culture of a profession, or
the culture of an organization.

15. In your opinion, what are the similarities and differences between the culture of
academia and the culture of industry?

16. Do you see cultural difference you mentioned above affecting the process and
outcome of academic-industry knowledge sharing? If so, can you tell me an
example where cultural difference impacted the outcome of university-industry
knowledge sharing?

Section 6: Informal knowledge sharing
10. Researchers have noticed that people also share personal/individualized
knowledge, such as technical know-how, through informal knowledge sharing
channels.

257

17. What is the role of informal/personal communication in academic-industry
knowledge sharing? How would you rate its importance on a scale from 1 to 5 with
1 meaning not significant at all and 5 meaning extremely significant?

18. Based on your own experience in academic-industry knowledge sharing, what
kind of knowledge do you share through informal channels. (I mean other than the
sharing of knowledge through formal channels such as patent licensing)

Section 7: The use of communication technology
19. Have you participate in academic-industry knowledge sharing across long
distance, such as across national borders? If so, in your opinion, how is it similar to
or different from knowledge sharing within a local context? Is long distance a
barrier to successful knowledge sharing? What kinds of communication
technologies do you use in this process? In your opinion, how well do these
communication technologies replace the need for face-to-face communication?

258

Appendix C: List of Codes with Frequencies
Code Frequencies
Number of codes
related
academic-academic KS 8 11
academic-industry knowledge sharing 133 22
academic-industry partnership 5 1
academic freedom 19 5
barrier-institutional change 7 1
barrier-lack of trust 8 2
barrier-mission 13 2
barrier-misunderstanding of expectation 16 1
barrier-noncompete contract 2 1
barrier-organizational culture 9 1
barrier-overprice IP 24 2
barrier-perceived difficulty in dealing with TTO 4 0
barrier-transaction cost 11 1
barrier-unclear IP 7 2
barrier-university's constraints 19 2
battier-TTO-Incompetence 20 4
capital 19 1
Channel-conference and seminars 26 2
Channel-consulting 23 2
Channel-education and training program 7 1
Channel-invention assignment 3 0
Channel-invited visit or talk 10 1
Channel-knowledge brokers 4 0
Channel-KS by hiring 3 1
Channel-material and data transfer 11 0
Channel-patent licensing 61 0
Channel-publication 27 3
Channel-research collaboration 45 4
Channel-research contract 13 1
Channel-research grants 12 2
Channel-special research centers 7 0
Channel-spin-off 20 0
Channel-Strategic alliance 3 0
characteristics of patent 11 5
commitment 1 2
communication before KS 15 0
259

company's greediness 9 0
confidential information 6 1
continuity in research 7 2
cultural convergence 9 0
cultural difference 46 1
culture-characteristics of research 18 5
culture-different timeline 15 1
culture-different values 8 0
culture-discipline 18 2
culture-dress code 1 0
culture-emphasis on IP 11 0
culture-emphasis on teamwork or independence 6 1
culture-honest 2 2
culture-institutional incentive 26 1
culture-personal reward system 15 3
culture-strive for scientific excellence 7 0
dual-diligence 8 3
ease of collaboration (unrelated to result) 3 3
entrepreneurial culture 7 1
entrepreneurially minded faculty 1 0
ethics 7 1
expectation 9 0
face-to-face communication 17 6
faculty's involvement 10 7
failed deal 25 12
failure in delivery research product 1 0
Failure to cash out 13 4
flexibility 7 3
formal interaction 5 1
geographic distance 15 4
goal-access to future employees 4 4
goal-access to knowledge and technology 28 3
goal-access to proprietary data and materials 10 1
goal-comply to grant policy 2 3
goal-involved in cutting edge research 7 1
goal-lower risk 1 0
goal-monetary reward 32 3
goal-more research funding 35 4
goal-product 25 7
goal-profit 8 8
260

goal-promote the development of science 11 6
goal-public good 22 7
goal-save money 6 2
goal-save time 2 2
goal-see science leads to real products 13 3
goal-social capital 12 6
goal-student placement 3 2
goal-technical problem solving 12 5
grant system 8 2
Industry's dependence on academia 3 0
industry cluster 7 3
informal knowledge sharing 39 24
information leaking 23 6
intellectual property 71 1
Inter-firm competition 11 4
IP-copy right 1 0
IP-formal contract 22 3
IP-institutional history 1 2
IP-patent 19 1
IP-published or unpublished work 5 1
IP-reputation 2 2
IP-trade secret 6 0
IP-training 2 0
IP-University ownership 6 1
IP infringement 17 6
knowledge-industry trend 2 0
Knowledge-interpretation of result 1 1
knowledge-methodology 3 1
Knowledge-research tools 3 0
Knowledge-technical know-how 9 1
Knowledge-way of working 2 1
knowledge hoarding 5 6
knowledge of industry (or lack of it) 9 0
leadership 6 1
legal action 5 3
long-term relationship 9 5
management support 3 0
market environment 1 0
market value of technology 13 0
national culture 1 1
261

negative effect of KS 2 1
New tradition 5 0
Non-disclosure agreement 12 0
old tradition 7 0
one-way KS 4 0
openness 35 10
past experience 3 1
Patent 5 1
patent system 2 2
personal communication 21 3
personal relationship 2 4
personality 5 0
public knowledge 1 0
public vs. private university 10 2
relationship 42 5
research capacity 3 0
research quality 2 0
SBIR 6 0
Secrecy 30 9
sharing of insider information 1 2
success story 25 11
successful commercialization 7 4
tacit knowledge 1 0
Technology-Conference call 4 0
Technology-email 14 0
Technology-Internet 7 0
Technology-telephone 10 0
Technology-transportation 2 0
Technology-video conference 7 0
termination of collaboration 1 1
time difference 3 1
training on KS 3 0
transaction-based relationship 12 1
trust 18 10
TTO 25 1
two-way KS 11 1
uncertainty in KS 1 1
understanding of market 6 1
University reach out 7 0
within-company KS 2 1
262

Appendix D: Sample Codes and Quotations

Codes Sample quotations
Academic-industry KS
Goal-Access to knowledge
and technology
Goal-Product
Goal-profit
Well, I think that like all the industrial
companies, companies are for profit, right? They
make a product. So if you are a science-based
company, basically you want to see what
advances in science can lead to new products. So
that is basically the goal for not only the biotech,
but also pharmaceutical and high tech
companies. So essentially the majority of the
breakthrough science is done in academia, so
basically you want to collaborate with academics
so that the breakthrough science hopefully can
help you generate more products in the future. So
that's the goal, I think, for majority of the
collaborations.
Academic-industry KS
Goal-Technical problem
solving
Goal-Product
Transaction-based
relationship
Long term relationship
So there may be two tier of collaborations, some
with very short-term timeline in mind and to
solve certain technical challenge, and some with
very long-term prospect kind of mindset that if
you keep working on this, and keep
collaborating, eventually you will have a
product.
Academic-industry KS
Academic-academic KS
Openness
Culture-Personal reward
system
Culture-Emphasis on IP
Culture-Teamwork or
independence
Secrecy
Within-company KS

Well, interesting enough, academic culture
ideally is supposed to be a lot of sharing of ideas
and approaches. For whatever reason, I think
there are a lot of reasons for this, the reality is
that there is less information sharing culturally in
academic organizations than there needs to be to
move science along. And I've talked to people
both in industry and academics about this. In
industry internally within a particular company,
there is much more information sharing than
there is back at the academic organization
because everyone has a common interest in
seeing a good product developed. There isn't that
common interest in academics. It's much more
fragmented. The self- interest is much more
fragmented academics than it is in a company.
So the information sharing, and it's just the
263

reverse of what you might think it would be in
some sense. Now clearly companies are
extremely careful about sharing their own
proprietary information with other companies,
but within the company itself, there is more of an
openness.
264

Appendix E: List of Conceptual Networks Created
Motivations to academic-industry knowledge sharing
Informal knowledge sharing
Academic-academic industry knowledge sharing
Academic-industry knowledge sharing
Barriers to academic industry knowledge sharing
265

Appendix F: Visualization of interactive and transactional knowledge network
1995, 2000, and 2006
1995 License network

266

1995 Research network

267

2000 License network

268

2000 Research network

269

2006 License network

270

2006 Research Network

271

Abstract (if available)

Abstract We live in a knowledge society where intellectual capital is the main driving force of social development and wealth creation. The competitiveness of today's organization lies in its ability to create, transfer, assemble, integrate, protect, and exploit knowledge. This dissertation studies the interorganizational knowledge sharing in the biotech industry.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Knowledge sharing in Chinese surgical teams

PDF

The role of capabilities in new alliance creation and performance: a study of the biotechnology industry

PDF

The evolution of knowledge creation online: what is driving the dynamics of Wikipedia networks

PDF

The geopolitics of the news: the case of the Al Jazeera network

PDF

The patterns, effects and evolution of player social networks in online gaming communities

PDF

An organizational culture approach to the study of individual power and its use

PDF

A system of participation and allocation preferences in organizations

PDF

Information sharing in work groups: A transactive memory approach

PDF

Effects of economic incentives on creative project-based networks: communication, collaboration and change in the American film industry, 1998-2010

PDF

The coevolution of multimodal, multiplex, and multilevel organizational networks in development communities

PDF

Organizational mimicry in American social movement communities: an analysis of form communication effects on the evolution of crisis pregnancy centers, 1989-2009

PDF

Collaborative care capacity: developing culture, power relationships and leadership support for team care in a hospital

PDF

Hybrid communication: a structurational analysis of computer gaming teams

PDF

Hurricane Katrina: framing leadership through communication

PDF

The gene is out of the bottle: the communication of genetic complexity in direct-to-consumer genetics

PDF

The globalization of cultural industries: Nollywood – the view from the South

PDF

Staying ahead of the digital tsunami: strategy, innovation and change in public media organizations

PDF

How social and human capital create financial capital in crowdfunding projects

PDF

The influence of social networking sites on high school students' social and academic development

PDF

Advocacy of public interest content in corporate entertainment media

Asset Metadata

Creator Tang, Lu (author)

Core Title Interorganizational knowledge networks: the case of the biotechnology industry

School Annenberg School for Communication

Degree Doctor of Philosophy

Degree Program Communication

Publication Date 06/20/2007

Defense Date 05/23/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag interorganizational knowledge sharing,knowledge networks,knowledge sharing,OAI-PMH Harvest,social network analysis

Language English

Advisor Riley, Patricia (committee chair), Fulk, Janet (committee member), Mitroff, Ian I. (committee member)

Creator Email ltang@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m541

Unique identifier UC1488335

Identifier etd-Tang-20070620 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-512067 (legacy record id),usctheses-m541 (legacy record id)

Legacy Identifier etd-Tang-20070620.pdf

Dmrecord 512067

Document Type Dissertation

Rights Tang, Lu

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu