social capital, networks, and knowledge · pdf filesocial capital, networks, and knowledge...
TRANSCRIPT
Social Capital, Networks, and Knowledge TransferAuthor(s): Andrew C. Inkpen and Eric W. K. TsangReviewed work(s):Source: The Academy of Management Review, Vol. 30, No. 1 (Jan., 2005), pp. 146-165Published by: Academy of ManagementStable URL: http://www.jstor.org/stable/20159100 .Accessed: 03/02/2012 03:34
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].
Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academyof Management Review.
http://www.jstor.org
? Academy o? Management Review 2005, Vol. 30, No. 1, 146-165.
SOCIAL CAPITAL NETWORKS. AND KNOWLEDGE TRANSFER
ANDREW C. INKPEN Thunderbird and Nanyang Business School
ERIC W. K. TSANG Wayne State University
We examine how social capital dimensions of networks affect the transfer of knowl
edge between network members. We distinguish among three common network types:
intracorporate networks, strategic alliances, and industrial districts. Using a social
capital framework, we identify structural, cognitive, and relational dimensions for the
three network types. We then link these social capital dimensions to the conditions
that facilitate knowledge transfer. In doing so, we propose a set of conditions that
promote knowledge transfer for the different network types.
Networks provide firms with access to knowl
edge, resources, markets, or technologies. In this
article we focus on networks and how firms ac
quire knowledge through their positions within networks. Various scholars interested in net
work relationships have recognized the knowl
edge dimension of networks and its link with
competitive success (e.g., Baum, Calabrese, &
Silverman, 2000; Dyer & Nobeoka, 2000; Gupta &
Govindarajan, 2000; Nishiguchi, 1994). A key ar
gument is that, through membership in a net
work and the resulting repeated and enduring
exchange relationships, the potential for knowl
edge acquisition by the network members is cre
ated. Our interest is in knowledge acquisition, how knowledge transfer between network mem
bers occurs, and what role social capital plays in the transfer.
The primary motivator for us was a theoretical
gap in the research where the key concepts of
networks, social capital, and organizational knowledge transfer intersect. This gap is the
result of four interconnected theoretical re
search threads operating at an organizational level. First, there is a well-established body of
literature underscoring important relationships between knowledge and networks. Second, in
the network area there is increasing interest in
understanding how the social context in which
firms are embedded influences their behavior
and performance (Gulati, Nohria, & Zaheer, 2000; Uzzi & Gillespie, 2002). Third, social capital has
been identified as a concept that can add value to the study of network social processes (Lee, Lee, & Pennings, 2001; Leenders & Gabbay, 1999). Fourth, in various academic (e.g., Adler & Kwon,
2002; Gargiulo & Benassi, 2000; Nahapiet & Ghoshal, 1998) and practitioner-oriented publi cations (e.g., Anand, Glick, & Manz, 2002; Baker,
2000), researchers recently have argued that ac
cess to new sources of knowledge is one of the
most important direct benefits of social capital. Moreover, there is evidence suggesting that
knowledge transfer is facilitated by intensive
social interactions of organizational actors
(Lane & Lubatkin, 1998; Yli-Renko, Autio, & Sa
pienza, 2000; Zahra, Ireland, & Hitt, 2000). In examining these four conceptual threads, it
became apparent to us that a systematic theo
retical analysis of social capital and the transfer
of knowledge between network members did not
exist. Although various variables that affect net
work knowledge exchange and transfer have
been posited (such as firm intent, absorptive
capacity, and control systems), there are few
studies that examine how the social capital di
mensions of networks affect an organization's
ability to acquire new knowledge from the net
work and facilitate the transfer of knowledge among network members. The literature on so
cial capital identifies knowledge access as a
key benefit but does not address the manageri
ally relevant question of how social capital ac
We thank Richard Osborn, Arvind Parkhe, Henry Yeung, and the anonymous AMR reviewers for their valuable com
ments. Both authors contributed equally to the article.
146
2005 Inkpen and Tsang 147
tually affects knowledge transfer between net
work actors.
By explicitly linking social capital, networks,
and knowledge transfer, we strive to achieve
several objectives. The primary objective is to
examine how the social capital dimensions of
networks affect an organization's ability to ac
quire new knowledge from the network and fa
cilitate the transfer of knowledge among net
work members. A key premise is that networks
create access to knowledge for the network ac
tors. Since access is a necessary but not suffi
cient condition leading to transfer,1 we are in
terested in conditions that facilitate knowledge transfer.
A second objective is to integrate the diverse
literature on networks and knowledge transfer.
Because the network concept is broad and mul
tidimensional, we have chosen to distinguish among three common network types: intracorpo rate networks, strategic alliances, and indus
trial districts. Using a social capital framework
derived from Nahapiet and Ghoshal (1998), we
identify structural, cognitive, and relational di
mensions for the three network types. We then
link these social capital dimensions to the con
ditions that facilitate knowledge transfer. In do
ing so, we propose a set of conditions related to
knowledge transfer for the different network
types. Finally, a third objective is to help advance
the study of social capital beyond that of an
umbrella concept (Adler & Kwon, 2002) to a use
ful and valid concept with the potential for un
derstanding network processes. We organize the paper as follows. In the first
three sections we discuss the three core con
cepts?namely, network types, knowledge transfer, and social capital. We then identify how the social capital dimensions are embed ded in each network type. This is followed by a
section in which we examine how the social
capital dimensions influence knowledge trans
fer for the network types. Finally, we discuss
implications and conclusions.
NETWORK TYPES
In this article we focus on strategic networks, which are composed of interorganizational ties
that are enduring and of strategic significance for the firms entering them (Gulati et al., 2000). A
key characteristic of networks is repeated and
enduring exchange relationships between the
actors in the network (Podolny & Page, 1998). This definition of networks includes a wide
range of forms, including intracorporate busi ness units, strategic alliances, franchises, R&D
consortia, buyer-supplier relationships, busi ness groups, trade associations, government
sponsored technology programs, and so on.
Figure 1 shows a typology of some common
network types along two dimensions.2 The ver
tical-horizontal dimension represents the extent
to which network members occupy different po sitions along the network's value chain. The
structured-unstructured dimension represents the extent to which network governance is struc
tured. In a structured network, members' roles
and relationships are clearly defined, and mem
bers are well organized to achieve certain goals. The reverse is true for an unstructured network.
A challenge in studying networks is ade
quately specifying the boundaries of the net
works (Gulati, 1995). The three network types defined below are not intended to be exhaustive
in coverage. Doing so is beyond the scope of this
paper. Our intent is to cover a spectrum of hor izontal and vertical relationships that go from the single-node divisionalized firm (the intracor
porate network) to interfirm relationships (the alliance) to an unstructured collection of firms
(industrial district). As indicated by Figure 1, the
three network types cover both ends of each of
the two dimensions. Moreover, they are among the most researched and discussed network
types. Although it is not feasible to examine all
organizational network types, our discussion of
multiple types raises key issues to help under
stand the relationships between knowledge and
network types not specifically discussed.
1 Inkpen and Beamish (1997) examine the issue of knowl
edge access versus knowledge transfer in detail. With a
focus on alliances, the authors examine the conditions under
which knowledge is transferred from one partner to another
and the resulting impact on alliance stability. The authors
argue that the formation of the alliance creates knowledge access for the partners, but transfer only occurs under cer
tain conditions.
2 We thank an anonymous reviewer for suggesting the
typology of networks. Note that the location and shape of
each network type in the figure are approximations only.
148 Academy of Management Review January
FIGURE 1 A Typology of Network Types
Vertical
Horizontal
Structured Unstructured
Intracorporate Network
An intracorporate network consists of a group of organizations operating under a unified cor
porate identity, with the headquarters of the net work having controlling ownership interest in its subsidiaries. Following Ghoshal and Bartlett
(1990), we conceptualize an intracorporate net work as an interorganizational grouping, rather than a unitary organization, because valuable
insights on the internal structures and opera tions of such an entity can be gained from net
work-related concepts used for investigating in
terorganizational phenomena. There is a clear linkage between ownership
and hierarchical power in an intracorporate net
work. Nevertheless, the strength of the link var
ies greatly along several dimensions, such as
the extent of decentralizing decision-making au
thorities to subsidiaries, the nature of the indus
try concerned, and the physical and cultural dis tances between headquarters and subsidiaries.
Strategie Alliance
A strategic alliance is a group of firms enter
ing into voluntary arrangements that involve
exchange, sharing, or codevelopment of prod ucts, technologies, or services (Gulati, 1998). The last two decades have witnessed a proliferation of strategic alliances among firms as a result of
technological development and globalization. An alliance can be formed by firms located in different positions or in the same position of the value chain. In the latter case, the firms con
cerned may produce similar products and com
pete in similar geographical markets (see Hamel, Doz, & Prahalad, 1989).
It is common for firms to enter into multiple alliances with a number of partners?a phenom enon that has been called an "alliance network"
(Koka & Prescott, 2002). For the sake of clear and concise exposition, we base our discussion on
the context of a strategic alliance rather than a
collection of alliances. That said, the issue of
2005 Inkpen and Tsang 149
knowledge transfer is conceptually the same, whether there is one alliance or multiple alli
ances, because it still involves knowledge mov
ing between alliance partners. Also for clarity, we note that this context includes the complex form of alliance known as an alliance constella
tion, which is an alliance involving multiple firms, such as the code-sharing alliances among airlines. Although these alliances often involve
quite complex design and governance, they have the same value creation logic as bilateral alliances (Das & Teng, 2002).
Industrial District
An industrial district is "a network comprising independent firms operating in the same or re
lated market segment and a shared geographic locality, benefiting from external economies of scale and scope from agglomeration" (Brown &
Hendry, 1998: 133). Some famous examples in clude Silicon Valley, Route 128, the Third Italy, and the City of London. An industrial district consists of a network of producers, supporting organizations, and a local labor market (Scott, 1992). There may or may not be a vertical divi sion of labor among the producers. For instance,
Storper (1993) found that clusters of firms in northeast central Italy exhibited a marked divi sion of labor, whereas other Italian clusters
comprised groups of firms doing more or less the same thing. There are usually major universities located inside or close to industrial districts. The universities train skilled personnel and provide technical and research support to firms in the districts.
KNOWLEDGE TRANSFER
We are interested in conditions that facilitate
knowledge transfer in networks. Knowledge transfer is the process through which one net
work member is affected by the experience of another (Argote & Ingram, 2000). Knowledge transfer manifests itself through changes in
knowledge or performance of the recipient unit. In a growing body of research, scholars argue
that organizations able to transfer knowledge effectively from one organizational unit to an
other are more productive than organizations that are less capable of knowledge transfer (e.g.,
Almeida & Kogut, 1999; Argote, Beckman, &
Epple, 1990; Baum & Ingram, 1998; Hansen, 2002;
Kostova, 1999). New knowledge, especially
knowledge from outside the firm, can be an im
portant stimulus for change and organizational
improvement. Related to the network context more specifically, Kotabe, Martin, and Domoto
(2003) found that organizational benefits can
arise from knowledge transfer between network
firms.
Gupta and Govindarajan (1991) argue that the
MNC can be regarded as a network of capital,
product, and knowledge transactions among units operating in different countries and that "the primary reason why MNCs exist is because
of their ability to transfer and exploit knowledge more effectively and efficiently in the intracor
porate context than through external market mechanisms" (Gupta & Govindarajan, 2000: 473). For each type of transaction, subsidiaries can
differ in the extent to which they engage in
intracorporate transactions and whether they are receivers or providers of what is being trans
acted. Gupta and Govindarajan (2000) further maintain that MNCs exist primarily because of their superior ability to transfer knowledge in
ternally, relative to the ability of markets. In this article we focus on conditions that facilitate in
tracorporate knowledge flows among subsidiar ies.
In a strategic alliance, knowledge transfer can be viewed from several perspectives. First, firms may acquire knowledge useful in the de
sign and management of other alliances (Lyles, 1988). This collaborative know-how may be ap
plied to the management of future alliances.
Second, firms may acquire knowledge about an
alliance partner that supports the firm's ability to manage the collaborative task. The knowl
edge obtained can be central to the evolution of the alliance (Ari?o & de la Torre, 1998; Doz, 1996). Third, firms learn with an alliance partner when the partners jointly enter a new business area
and develop new capabilities. Last, firms ac
quire knowledge from an alliance partner by gaining access to the skills and competencies the partner brings to the alliance (Baum et al., 2000; Kogut, 1988). In this context, alliances pro vide opportunities to create redeployable knowledge (or private benefits), such as techni cal knowledge or market knowledge. For the
purpose of our discussion, we focus on the last
150 Academy of Management Review January
two perspectives, which concern knowledge flows between alliance partners.3
A large number of recent studies of industrial
districts have emphasized the capacity of dis tricts to support processes of knowledge acqui sition and innovation as the basis for creating competitive advantage (for a review, see Mac
Kinnon, Cumbers, & Chapman, 2002). Firms in an industrial district have various opportunities to tap into a larger knowledge resource base. The knowledge of primary interest is usually highly tacit, difficult to replicate, and not easily purchased. Helmsing, citing Lawson (1999), de scribes industrial district learning as the "emer
gence of basic shared knowledge and proce dures amongst a group of (geographically close) firms which facilitates co-operation and prob lem-solving" (2001: 289). Geographic proximity facilitates knowledge flows and technical ex
changes among firms (Marshall, 1920). Keeble and Wilkinson (1999) identify three ma
jor mechanisms for the spatial transfer of knowl
edge within the boundaries of an industrial dis trict: (1) interfirm mobility of the labor force
within the district; (2) interactions between sup
pliers and customers and the makers and users
of capital equipment; and (3) spin-off of new
firms from existing firms, universities, and pub lic sector research laboratories. The knowledge transfer processes taking place in an industrial
district give rise to cumulative local know-how
that goes beyond the boundaries of the firm but that remains within the spatial boundaries of the district (Capello, 1999).
SOCIAL CAPITAL
Those studying interfirm relationships in
creasingly focus on how firms are socially em
bedded in networks of relationships that incor
porate a diverse set of organizational actors.
Social capital is gaining prominence as a con
cept that provides a foundation for describing
and characterizing a firm's set of relationships. However, although the concept of social capital has found widespread acceptance, there re
mains widespread uncertainty about its mean
ing and effects (Koka & Prescott, 2002). In his review, Portes (1998) identifies Pierre
Bourdieu's (1986) analysis as the first systematic analysis of social capital. Bourdieu defined the
concept as "the aggregate of the actual or po tential resources which are linked to possession of a durable network of more or less institution alized relationships of mutual acquaintance or
recognition" (1986: 248). As the concept evolved,
through work by Coleman (1988), Burt (1992), and others, a consensus emerged that social capital represents the ability of actors to secure benefits
by virtue of membership in social networks or
other social structures (Portes, 1998). At an orga nizational level, benefits include privileged ac cess to knowledge and information, preferential opportunities for new business, reputation, in
fluence, and enhanced understanding of net
work norms.
Although Adler and Kwon's (2002) comprehen sive review identifies many different ap
proaches used in studying social capital, two
patterns emerge from the various definitions
(Leana & Van Buren, 1999). The first is derived from social network theorists (e.g., Belliveau,
O'Reilly, & Wade, 1996; Burt, 1997; Useem &
Karabel, 1986), who emphasize personal bene
fits, such as career advancement, that actors
gain directly from their social capital. Propo nents of this perspective consider social capital a private good possessed by individuals. Other scholars conceptualize social capital as a public good (e.g., Bourdieu, 1986; Coleman, 1988; Put
nam, 1993). They regard social capital as an
attribute of a social unit, rather than an individ ual. As a public good, social capital is available to and benefits not only those who create it but
also group members at large (Kostova & Roth,
2003). For this paper we adopt a definition of social
capital similar to that offered by Nahapiet and
Ghoshal (1998) and used by Bolino, Turnley, and
Bloodgood (2002).4 We define social capital as
3 We do not assume the existence of a dominant partner in
the strategic alliance. Power dynamics in a strategic alli
ance are complicated and beyond the scope of this paper. For instance, a joint venture partner with minority equity can
use various ways to control the venture (Hamel, 1991;
Schaan, 1988). Moreover, bargaining power may shift from
one partner to another, depending on their relative learning
speeds (Inkpen & Beamish, 1997). As can be seen from our
discussions, virtually all of our arguments apply to alliances
with or without dominant partners.
4 In their footnote 1 Bolino et al. (2002) acknowledge that
alternative social capital frameworks exist. Their rationale
for using Nahapiet and Ghoshal's (1998) framework is four
fold: (1) it integrates many of the social capital facets dis
cussed in previous work, (2) it is useful for examining social
2005 Inkpen and Tsang 151
the aggregate of resources embedded within, available through, and derived from the network of relationships possessed by an individual or
organization?a definition that accommodates both the private and public good perspectives of social capital. The central proposition in this view of social capital is that networks of rela
tionships are a valuable resource (i.e., capital) for the individual or organization. The logic of this view can be seen in the example of a firm that establishes a network tie with another firm, such as a supply contract. This network tie be comes a social capital resource of the two firms. As time passes, trust between the firms may
develop, and such trust, in addition to the formal tie between the firms, will also constitute a so
cial capital resource. The social capital of the firms is thus enhanced. From the social capital, various benefits, such as preferential knowl
edge access, may flow to the firms. Individual social capital originating from an
individual's network of relationships can be dis
tinguished from organizational social capital derived from an organization's network of rela
tionships.5 The former has the property of a pri vate good, whereas the latter takes on the na
ture of a public good. With social capital as a
public good, members of an organization can
tap into the resources derived from the organi zation's network of relationships without neces
sarily having participated in the development of those relationships (Kostova & Roth, 2003). These two levels of social capital are often interre lated. For example, a manager, through his or
her own social relationships and personal con
nections, can help his or her company set up a
joint venture with another company. In this case,
organizational social capital is created on the basis of individual social capital.
For a systematic analysis of organizational social capital across multiple network types, we need to distinguish among (1) the possessors of social capital, (2) the dimensions of social capi tal, (3) the benefits of social capital, and (4) the
factors that operate as determinants of social
capital benefits. In this article the possessors of
organizational social capital are the members of the three network types identified above. The social capital benefit examined is the opportu nity to acquire knowledge from other network members. As discussed in the following section, the dimensions of social capital refer to the clus ters of the main facets of social capital (Na
hapiet & Ghoshal, 1998). After the discussion of these dimensions, our attention shifts to the de terminants of knowledge transfer, which is the social capital benefit examined in this paper.
We propose conditions that facilitate knowledge transfer. Our fundamental argument is that, de
pending on the network type, different condi tions will affect how the social capital dimen sions influence knowledge transfer. This section links the social capital dimensions with network
types and provides a set of theoretical relation
ships. Although we focus on analyzing organi zational social capital across network types, we
also incorporate individual social capital in our
discussion, because the interplay between the two affects knowledge transfer between net
work members.
DIMENSIONS OF SOCIAL CAPITAL AND NETWORK TYPES
Knowledge acquisition has been identified as a direct benefit of social capital (Adler & Kwon, 2002; Nahapiet & Ghoshal, 1998). In this paper we seek to understand how knowledge moves within networks and how social capital affects the knowledge movement. To achieve this objec tive, we adopt Nahapiet and Ghoshal's (1998) three dimensions of social capital: structural,
cognitive, and relational. Table 1 shows the three network types and the
three social capital dimensions. We draw from the literature to illustrate how these social cap ital dimensions are embedded in each network
type. Depending on the network type, the nature of social capital varies. We note that within each network type there is substantial variance, in that there are different forms of intracorporate networks, strategic alliances, and so on. The characteristics of social capital in Table 1 are associated with the more typical forms of each network type.
capital at the organizational level, (3) it incorporates a cog nitive dimension, and (4) it establishes a relationship be
tween social capital and intellectual capital (i.e., organiza tional knowledge).
5 Our conceptualization of organizational social capital is
different from that of Leana and Van Buren, who define it as
"a resource reflecting the character of social relations within
the firm" (1999: 538).
152 Academy of Management Review January
TABLE 1 Social Capital Dimensions Across Network Types
Social Capital Dimensions Intracorporate Network Strategic Alliance Industrial District
Structural
Network ties
Network
configuration
Network stability
Cognitive Shared goals
Shared culture
Relational: Trust
Fuzzy distinction between
intramember and
intermember ties
Hierarchical, easy to
establish connectivity between network
members
Stable membership
Members working toward
a common goal set by
headquarters
Overarching corporate culture
Little risk of opportunism, institutional-based trust
Intermember ties
determining social ties
within an alliance
Nonhierarchical, possibility of exploiting structural
hole positions
High rate of instability
Compatible goals but rarely common goals
Cultural
compromise/conflict
among members
Significant risk of
opportunism, behavioral
based trust
Social ties as a foundation
for intermember ties
Nonhierarchical and dense
networks in a
geographical region
Dynamic, with members
joining and leaving the
district
Neither shared nor
compatible goals
Industry recipe
Process-based personal trust
Structural Dimension
The structural dimension of social capital in
volves the pattern of relationships between the network actors and can be analyzed from the
perspective of network ties, network configura tion, and network stability.6 Network ties deal
with the specific ways the actors are related. Ties are a fundamental aspect of social capital, because an actor's network of social ties creates
opportunities for social capital transactions
(Adler & Kwon, 2002). A key feature of intracor
porate networks is that members of a network
belong to the same corporation. As such, ties within a member, such as interdepartmental and interpersonal relationships, may not be
very different in nature from those between members. In other words, boundaries of network
members are more porous than those of other network types. The nature of ties between alli ance partners will impact the social ties be tween managers who are assigned to the alli ance by the partners. For instance, in an
alliance formed between keen competitors, such social ties will likely be cautious and tense, because each partner is wary of divulging valu able knowledge to other partners. That is, the nature of organizational social capital sets the tone for individual social capital. An important characteristic of network ties between members in an industrial district is that, more often than
not, they are established as a result of interper sonal relationships developed from informal so
cial gatherings and meetings (Brown & Hendry, 1998; Paniccia, 1998). As such, individual social
capital forms the basis of organizational social
capital. The configuration of a network structure de
termines the pattern of linkages among network
members. Such elements of configuration as hi
erarchy, density, and connectivity affect the
flexibility and ease of knowledge exchange
through their impact on the extent of contact and
accessibility among network members (Krack hardt, 1992). Intracorporate networks are often
6 The facets of each social capital dimension discussed in
this paper are by no means exhaustive. Owing to space
limitations, we focus on facets that are most related to
knowledge transfer between network members. For in
stance, we replace the facet "appropriable organization," which Nahapiet and Ghoshal (1998) include in their struc
tural dimension, with "network stability." Our rationale is
that stability varies greatly across network types and has
serious implications for knowledge transfer. These points are explained in detail in the following discussion.
2005 Inkpen and Tsang 153
arranged in a hierarchical way, with headquar ters being at the top of the hierarchy. Depending on the overall corporate structure, some mem
bers of the network may not be connected to
other members. Connectivity can be easily es
tablished either through the headquarters or on
the members' own initiatives. Although alli ances such as equity joint ventures are broader and deeper in interactions than alliances such as technology licenses, the structure of a strate
gic alliance is nonhierarchical. Connectivity is
less straightforward than the case of intracorpo rate networks, since it needs to be established across firm boundaries. Thus, some members
may span the structural holes of the network and enjoy the associated informational advan
tages (Burt, 1992). Connectivity between network members in an industrial district is usually es
tablished through informal interpersonal rela tions. While a general pattern, in terms of net
work density, for intracorporate networks and alliances would be unusual, a characteristic of an industrial district is dense, nonhierarchical networks of firms located within the district, with some of them forming cliques.
Network stability is defined as change of
membership in a network. A highly unstable network may limit opportunities for the creation of social capital, because when an actor leaves the network, ties disappear. While stability is not a major issue in intracorporate networks un
less there are frequent corporate restructuring activities, it is a much studied concept in the alliance area, perhaps because of the high in
stability rate usually attributed to this particu lar network form (Inkpen & Beamish, 1997; Yan &
Zeng, 1999). Last, membership in an industrial district is often unstable, with firms joining and
leaving the district continuously.
Cognitive Dimension
The cognitive dimension represents the re sources providing shared meaning and under
standing between the network members (Na
hapiet & Ghoshal, 1998). The two facets of the dimension we address are shared goals and shared culture among network members. Shared goals represent the degree to which net work members share a common understanding and approach to the achievement of network tasks and outcomes. Depending on the network
type, the tasks and outcomes may vary in clarity
and definition. Members of an intracorporate network usually work toward a common goal set
by headquarters, although they may have to fulfill certain secondary goals related to their own products and markets. Partner firms often have different goals in mind when they enter a
strategic alliance. Negotiation helps partners arrive at goals that are acceptable to most, if not
all, of them. In an industrial district there are
likely to be few shared or even compatible goals, owing to the complexity of the network ties.
Shared culture refers to the degree to which norms of behavior govern relationships. This facet is similar to tie modality, which is "the set of institutionalized rules and norms that govern
appropriate behavior in the network. While these are sometimes spelled out in formal con
tracts, most often they are simply understand
ings that evolve within the dyad and the net work" (Gulati et al., 2000: 205). In some cases
shared norms may create excessive expecta
tions of obligatory behavior and may possibly result in problems of free riding and unwilling ness to experiment beyond the network. Mem bers of an intracorporate network work under an
overarching corporate culture. For instance, all the operations of Johnson & Johnson worldwide subscribe to the strong ethical culture set by the
headquarters. Since partner firms usually have distinct cultures, strategic alliances are often formed on the basis of cultural compromise among the partners concerned. Cultural conflict will arise if certain partners rigidly push for ward their own ways of doing things. Although firms in an industrial district may have various distinct cultures, they tend to share an industry recipe. As organizational ecologists (e.g., Han nan & Freeman, 1977) argue, firms in the same
line of business experience substantial pressure to adopt similar policies. Similarly, Spender's (1989) study indicates that, in the face of uncer
tainty, managerial recipes specific to task envi ronments gradually evolve and are adopted by firms operating in the industries concerned.
Relational Dimension
The relational dimension focuses on the role of direct ties between actors and the relational, as opposed to structural, outcomes of interac tions. Among the facets of this dimension, such as trust, norms, and identification, we focus on
154 Academy of Management Review January
trust, both because of space limitations and be cause trust is a critical factor affecting interfirm
knowledge transfer and creation (Dodgson, 1993; Doz, 1996). Trust is based on social judgments (e.g., assessment of the other party's benevo
lence, competence, etc.), together with assess ment of the costs (i.e., risk) if the other party turns out to be untrustworthy (Rousseau, Sitkin, Burt, & Camerer, 1998). Under a risky condition, a
party's trust is signified by a decision to take action that puts its fate in the hands of the other
party. Trust plays a key role in the willingness of
network actors to share knowledge. A lack of trust may lead to competitive confusion about
whether or not a network firm is an ally (Powell,
Koput, & Smith-Doerr, 1996). Conversely, an at
mosphere of trust should contribute to the free
exchange of knowledge between committed ex
change partners, because decision makers should not feel that they have to protect them selves from others' opportunistic behavior (Blau, 1964; Jarillo, 1988). As trust develops over time,
opportunities for knowledge transfer between network members should increase. With the de
velopment of a pattern of interactions, organiza tions will decrease their efforts to protect their
knowledge and skills. Trust in an intracorporate network is institu
tional based: the fact that an organization is a
member of the network signifies to other mem
bers that the former should be trustworthy. While risk of opportunism is normally not a con cern for intracorporate networks, it is a serious concern for strategic alliances. Unlike intracor
porate networks, trust in strategic alliances is
behaviorial based. A partner firm needs to sig nify its trustworthiness through the way it be haves in the alliance. For industrial districts,
interpersonal trust plays a critical role, since, as
mentioned earlier, individual social capital drives the development of organizational social
capital. Moreover, trust is process based, in the sense that firms regularly test each other's in
tegrity, moving from small, discrete exchanges of limited risk to more open-ended deals that
subject the parties to substantial risk (Lazerson & Lorenzoni, 1999).
NETWORKS AND KNOWLEDGE TRANSFER
In this section we examine social capital de terminants of network knowledge transfer.
Through the various ties that firms have with other firms, network members are exposed to various types of knowledge that are potentially valuable. As Powell states, "The most useful information is rarely that which flows down the formal chain of command in an organization, or
that which can be inferred from price signals. Rather, it is that which is obtained from some one you have dealt with in the past and found to be reliable" (1990: 304).
From a network perspective, Podolny and
Page (1998) identify two types of learning.7 First, networks can facilitate learning via the transfer of knowledge from one firm to another. In other
words, the network acts as a conduit for process
ing and moving knowledge; learning from an
alliance partner is this type of learning. Second, networks may become the locus of novel knowl
edge creation at the network level, rather than within the nodes?firms?of the network. Al
though our focus is primarily on the network as a conduit for knowledge transfer, the two forms of network learning are not always easy to dis
tinguish in practice. The dependent variable in this discussion is
knowledge transfer between network members. Based on the key argument that social capital plays a critical role in the transfer and exchange of network knowledge, we propose a set of conditions that facilitate knowledge transfer in networks. These facilitating conditions, as sum marized in Table 2, are factors specifically as
sociated with the respective facets of the three social capital dimensions. Our objective in this section is to identify specific relationships be tween social capital and knowledge transfer. Because the facilitating conditions that influ ence knowledge transfer differ across network
types, we show that developing an understand
ing of social capital and networks requires an
analysis of the specific features of the different network types.
7 In integrating various bodies of literature, we were con
fronted with differences in how such terms as knowledge transfer and learning have been used. The social capital literature usually discusses knowledge transfer (and access) rather than learning. The network literature, however, as the
Podolny and Page (1998) reference indicates, uses the term
learning in reference to the knowledge acquisition process. We have tried to be as consistent as possible in the use of
terminology.
2005 Inkpen and Tsang 155
TABLE 2 Conditions Facilitating Knowledge Transfer
Social Capital Dimensions Intracorporate Network Strategic Alliance Industrial District
Structural
Network ties
Network
configuration
Network stability
Cognitive Shared goals
Shared culture
Relational: Trust
Personnel transfer
between network
members
Decentralization of
authority by headquarters
Low personnel turnover
organization wide
Shared vision and
collective goals Accommodation for local
or national cultures
Clear and transparent reward criteria to
reduce mistrust among network members
Strong ties through repeated
exchanges
Multiple knowledge connections
between partners
Noncompetitive approach to
knowledge transfer
Goal clarity
Cultural diversity
Shadow of the future
Proximity to other members
Weak ties and boundary
spanners to maintain
relationships with
various cliques Stable personal
relationships
Interaction logic derived
from cooperation Norms and rules to govern
informal knowledge
trading Commercial transactions
embedded in social ties
The focus in Table 2 is on organizations within the network, rather than the network itself. We discuss knowledge acquired by network mem
bers through participation in the network's
knowledge-sharing activities. The hypothesized conditions, elaborated in the following section, can be readily converted into testable proposi tions and provide suggested directions for fu ture research.
Before getting to the detailed discussion, we
illustrate the logic of the table. One of the struc tural facets is network configuration. At a broad level, network configuration affects the flexibil
ity and ease of knowledge exchange between network members. For example, for intracorpo rate networks, a facilitating condition (for net
work configuration) is headquarter's decentrali zation of authority to network members such that the development of lateral network ties and
knowledge transfer is enhanced. Expressed as a
proposition, the greater headquarter's decen tralization of authority to intracorporate network
members, the more likely ties between the mem
bers will develop that lead to knowledge trans fer. As another example, having boundary span ners maintain weak ties with various cliques for
exploration purposes is an important facilitat
ing condition for firms operating in an industrial district. Expressed as a proposition, the greater
the presence of boundary spanners with weak ties to various cliques, the more likely a pattern of linkages among network members will de
velop that lead to knowledge transfer.
Structural Dimension
Network ties. Since the boundaries between
intracorporate network members are more po rous than those between members of other net
work types, personnel transfer between mem bers should take place more readily. Such transfers establish social network ties on top of the more formal intermember ties; the latter, in turn, are strengthened by the existence of the former. The social network ties facilitate inter
member social interactions and provide chan nels for knowledge exchange. Ghoshal, Korine, and Szulanski's (1994) research on MNCs docu ments the importance of such social interactions for diffusing new ideas within the corporations. In an in-depth case study of a Dutch multina tional software company, Orlikowski (2002) found that expatriates brought with them skills and techniques to share with local engineers.
For knowledge transfer to occur in alliances,
strong ties between the partners are necessary (Inkpen & Dinur, 1998). Factors supporting the
development of strong ties include prior partner
156 Academy of Management Review January
relationships and repeated transactions (Gulati, 1995). In the absence of strong ties, especially in
alliances between competitors, partners may not develop the necessary relationships that al
low managers to share knowledge willingly. Larson (1992) has shown that strong ties promote and enhance trust, reciprocity, and long-term perspectives. Kale, Singh, and Perlmutter (2000) found a positive relationship between the
strength of ties and the degree of learning in
alliances.
For industrial districts, Camagni (1995) identi fies spatial proximity as a key characteristic of
what he calls "a local network." From the per
spective of an individual firm, it is beneficial to be located physically close to other firms in the district. Proximity helps the formation of net
work ties and facilitates interfirm and espe
cially interpersonal interactions through which
knowledge is exchanged. The more tacit the
knowledge involved, the more important spatial proximity is between the parties taking part in
the exchange (Maskell & Malmberg, 1999). This
implies that firms occupying more central loca tions of the district have the edge over those located at the periphery.
Network configuration. As argued by Grant, "Once firms are viewed as institutions for inte
grating knowledge, a major part of which is tacit
and can be exercised only by those who possess it, then hierarchical coordination fails" (1996: 118). Thus, the headquarters of an intracorporate
network must decentralize authority to members of the network so that they can determine how to make the best use of the knowledge they pos sess. Moreover, decentralization enables mem
bers to establish lateral ties on their own initia
tive, without first seeking approval from
headquarters. Decentralization can facilitate
timely knowledge sharing among the members.
Tsai's (2002) study of a large, multiunit company confirms that centralization is negatively asso
ciated with intracorporate knowledge sharing. Based on the notion of connections through
which alliance managers can share their obser vations and experiences (Von Krogh, Roos, &
Slocum, 1994), Inkpen and Dinur (1998) identify four types of alliance structural ties that can
lead to knowledge sharing: technology link
ages, alliance-parent interaction, personnel transfers, and strategic integration. The four
processes share a conceptual underpinning in
that each creates the potential for individuals to
share their observations and experiences. Each of the four processes provides an avenue for
managers to gain exposure to knowledge and ideas outside their traditional organizational boundaries, and each creates a connection for individual managers to communicate their alli ance experiences to others.
In industrial districts, cliques of firms with
strong ties may be formed. For instance, in the
City of London, Japanese banks may form one
clique and American banks another. While there are intense knowledge exchanges within a
clique, there may be little between cliques. From the perspective of an individual firm oper ating in the district, it is crucial to have bound
ary spanners who maintain weak network ties with various cliques for exploration purposes
(Rowley, Behrens, & Krackhardt, 2000). One sim
ple way to achieve this is through participating in the activities of professional associations. In the case of the City of London, one such associ ation is the Chartered Institute of Bankers, which is a British association of banking profes sionals.
Network stability. Although the membership of an intracorporate network is usually more
stable than that of other network types, this sta
bility may not help knowledge transfer if there is a high personnel turnover rate. Organiza tional learning depends, at least partially, on
memories of individuals and their learning abil ities (Carley, 1992). Individuals leaving a net
work take with them knowledge that may be crucial for organizational success. In addition,
personnel turnover affects intracorporate knowl
edge sharing, which often takes place through formal or informal exchanges on an individual basis. Such exchanges are facilitated by estab lished rapport and friendship. Maintaining a
stable pool of personnel within a network can
help individuals develop long-lasting interper sonal relationships.
Learning from an alliance partner is a key determinant of bargaining power and alliance
stability (Hamel, 1991; Inkpen & Beamish, 1997; Yan, 1998). Hamel (1991) proposes that the most
important determinant of partner bargaining power in alliances is the ability to learn. Firms that can learn quickly are able to acquire part ner skills, reducing their dependence and in
creasing their bargaining power. Inkpen and Beamish (1997) used these arguments to develop a framework of instability and international
2005 Inkpen and Tsang 157
joint ventures. In an alliance, dependence can
be a source of power for the firm controlling the
key resources. When knowledge acquisition shifts the dependency relationship, the cooper ative basis for the alliance may erode, leading to instability. If one partner acquires knowledge faster than the other, the faster-learning partner no longer has the same need, which can lead to a situation of partner asymmetry (Makhija &
Ganesh, 1997). Thus, if partner firms regard their alliance as a learning race field, the partner learning faster will be motivated to leave the alliance. If the attitude toward learning is non
competitive, however, destabilizing forces will be less likely and greater symmetric learning may occur.
Industrial districts are characterized by the constant entry and exit of firms. From the per
spective of firms remaining in the district, out
going firms take with them not only knowledge but also important personal contacts (unless the
employees concerned choose to stay behind).
Boundary spanners of the firms that remain may try to establish more intimate and stable per sonal relationships with their counterparts in the exiting firms. In this way, personal contacts can be maintained and may continue to serve as
useful sources of industrial information for firms that continue to stay in the district. In other
words, these firms' networks extend beyond the district. Such external contacts are important channels for obtaining fresh ideas, especially when managers within the district become more
homogeneous in their mental models over time
(Pouder & St. John, 1996).
Cognitive Dimension
Shared goals. We follow Tsai and Ghoshal
(1998) in using the term shared vision, which embodies the collective goals and aspirations of the members of an intracorporate network.
When a shared vision is present in the network, members have similar perceptions as to how
they should interact with one another. This can
promote mutual understandings and exchanges of ideas and resources. Thus, a shared vision can be viewed as a bonding mechanism that
helps different parts of a network integrate knowledge. When partner firms bring contradicting or in
consistent goals into their strategic alliance, in
terpartner conflict may arise. Conflict among
parties in an interfirm collaboration tends to result in frustration and dissatisfaction (Ander son, 1990). Such a negative atmosphere is not
conducive to the flow of knowledge between the
partners and the alliance. In studying intra- and
interdepartmental conflict within a large utility company, Schnake and Cochran (1985) found that lower levels of goal clarity increased both
types of conflict. For strategic alliances, we also
expect that goal clarity reduces interpartner conflict by facilitating the negotiation and es
tablishment of common goals. When the objec tives and strategies of an alliance are clearly stated, a foundation of common understanding and the means to achieve the collaborative pur pose is established among the partners (Das &
Teng, 1998). For firms in an industrial district to willingly
share knowledge, they must recognize that co
operation and knowledge sharing can enhance their competitive position. In addition, firms
must recognize that combining the economic, cultural, and technological resources of the in dustrial district can lead to joint knowledge cre ation. Thus, firms will share knowledge when an interaction logic is shared across network members (Helmsing, 2001). This logic is derived from the belief that value can be created
through cooperation and knowledge sharing. Joint problem-solving arrangements enrich the
network, because working through problems promotes innovation (Uzzi, 1997).
Shared culture. Although the headquarters of an intracorporate network may try to impose its
corporate culture in all worldwide operations, each operation is geographically embedded in local or national culture (Ghoshal & Bartlett, 1990). For instance, a corporate culture empha sizing participative decision making may not work well in a high power distance culture. The local or national culture needs to be understood and accommodated so that when knowledge is transferred from one member to another, the transfer process will not be hindered by cultural conflicts between concerned members (see Bha
gat, Kedia, Harveston, & Triandis, 2002).
Arguments for and against partner cultural
diversity as an antecedent for alliance learning have been made. Although Parkhe (1991) has
proposed that diversity between the partners in international strategic alliances could lead to
learning, Pitts and Lei (1997) have argued that alliances designed to learn and absorb tacit
158 Academy of Management Review January
knowledge are harder to manage among part ners that come from different cultural contexts
than partners from a similar cultural context.
Phan and Peridis (2000) have proposed that dif ferences between partners support the learning process. The authors' rationale is that attempts to eliminate differences can block second-order
learning processes. We maintain that the over
all effect of cultural diversity should be benefi cial to knowledge transfer.
A major barrier to informal exchange of
knowledge in an industrial district is the risk that the receiver of such knowledge may use it
against the interest of the sender. This risk can
be reduced and, hence, the exchange facilitated if there are implicit industrial norms and rules in the district governing informal know-how
trading among firms such that opportunism is
subject to severe social sanctions. The norms
and rules include a common language for talk
ing about organization and cultural problems and accepted but tacit codes of conduct between
firms (Helmsing, 2001). In an interesting study of
the extensive exchange of proprietary know
how by informal networks of process engineers in the U.S. steel minimill industry, Von Hippel (1987) found such norms.
Relational Dimension: Trust
An intracorporate network is a social structure of coopetition (Tsai, 2002). While intermember
cooperation is encouraged so as to realize econ
omies of scale, intermember competition can
also achieve efficiency (Hill, Hitt, & Hoskisson,
1992). When members compete against one an
other for resources and markets, suspicion may
replace trust in their relationship and, conse
quently, knowledge sharing may be sacrificed.8 It is important that headquarters establish clear
and transparent reward criteria so that the
members concerned will not suspect any under
the-table transactions or favoritism. Clear and
transparent reward criteria will reduce mistrust
among the members. Parkhe's (1993) study of strategic alliance
structuring suggests that opportunism is con
strained and cooperation promoted by the shadow of the future, which can be lengthened by long time horizons, frequent partner interac
tions, and high behavioral transparency. As the fear of opportunism by alliance partners fades because of the development of mutual trust,
partners will be more willing to move forward, even though uncertainty in the relationship may remain (Nooteboom, Berger, & Noorderhaven,
1997). When trust is high, firms may be more
likely to invest resources in learning because of the willingness of their partners to refrain from
instituting specific controls over knowledge spillovers.
In an industrial district, many of the ex
changes between network members are com
mercial transactions. Uzzi and Gillespie (2002)
argue that, in contrast to purely market-based
transactions, commercial transactions embed ded in social ties instill into future exchanges expectations of trust and reciprocity. In turn, re
lationships based on trust and reciprocity are
likely to promote the transfer of distinctive
knowledge and resources. When the relation
ships between industrial network members are
embedded with trust, firms may be more willing to share valuable knowledge and accept the risk
of spillover to competitors (Dyer & Singh, 1998).
Interplay Between Individual and
Organizational Social Capital
We noted earlier that social capital can be created at individual and organizational levels.9 For interorganizational knowledge transfer to
take place in a network, either or both levels of
social capital must be present. The discussion based on Table 1 indicates that the three net
work types display different characteristics with
respect to the two levels of social capital. For
example, for the network ties facet, interper sonal relationships developed from informal so
cial gatherings and meetings are an important characteristic of ties between members in an
8 We do not expect that the suspicion will lead to rampant
opportunism, because the penalties imposed on managers who are caught acting opportunistically may be severe. In
contrast, in strategic alliances, if opportunism is punished, the punishment will usually be at the organizational rather
than individual level. Managers who act opportunistically in
the alliance are normally under the direction of the partner
employing them.
9 In recent years organizational researchers have increas
ingly tried to examine multilevel theoretical perspectives for
concepts such as creativity (Drazin, 1999), learning (Kim,
1993), and trust (Inkpen & Currall, 2002).
2005 Inkpen and Tsang 159
industrial district. In contrast, for alliances, the nature of organizational-level ties has a signif icant impact on interpersonal ties. Based on the
mix of individual and organizational character
istics, the facilitating conditions summarized in
Table 2 can be examined more closely.
Compared with strategic alliances and indus
trial districts, in intracorporate networks orga nizational social capital is readily available between members. Intracorporate network mem
bers are connected within a corporate organiza tional structure that is relatively stable (com
pared to an alliance network or industrial
district). Also, relative to the other network
forms, intracorporate network members are
more likely to work toward a common corporate
goal, share an overarching corporate culture, and trust one another. As a result, some knowl
edge access and transfer will occur by default.
By default, we mean that, by virtue of being part of the network, subsidiary units are entitled to
obtain certain organizational knowledge from the headquarters or other subsidiaries. While there may be personal reasons for in
tracorporate managers not to share knowledge, there will not be the same type of competitive and firm-level barriers to knowledge sharing that exist in alliance networks and industrial districts. However, establishing individual so
cial capital will strengthen the knowledge flow. For instance, suppose an engineer of a subsid
iary is sent to headquarters to learn a new tech
nology. The public good nature of the organiza tional social capital between the subsidiary and
headquarters enables the engineer to access the new technology. Nevertheless, the development of positive interpersonal relationships between the engineer and those teaching the technology should allow for a faster and broader dissemi nation of knowledge. Thus, some of the facilitat
ing conditions in Table 2 are concerned with
developing individual social capital. For in
stance, personnel transfer and low managerial turnover can help individuals establish close,
long-lasting social ties that expedite knowledge transfer between network members.
Unlike an intracorporate network, a strategic alliance is an interfirm relationship with legal and organizational boundaries between firms. Also unlike an intracorporate network, the exis tence of an alliance does not guarantee a flow of
knowledge between partners, who often have a
competitive-collaborative relationship (see
Hamel, 1991). Knowledge may flow very slowly or not at all.10
Properly managing this relationship is critical if the partners seek access to each other's
knowledge. As discussed earlier, the nature of
organizational social capital often sets the tone for individual social capital in an alliance set
ting. Hence, organizational social capital will
likely be the dominant form of social capital necessary for knowledge to flow between alli ance partners. For example, it is not uncommon
for managers who are assigned to alliances to make comments such as "Joe, my counterpart from the alliance partner, is my good friend and I trust him. However, there are things we cannot share without our parents' authorization." Look
ing at Table 2 shows that the facilitating condi tions focus on building up organizational social
capital between alliance partners. In an industrial district, knowledge flows start
on a personal level, because there may not be formal interfirm relationships. When there are
formal relationships, they will tend to be com
mercial transactions, as opposed to the more
structured nature of an alliance relationship. Thus, individual social capital is critical, drives the development of organizational social capi tal, and becomes the focus of the facilitating conditions in Table 2.
In summary, the three network types involve different dynamics between organizational and individual social capital. Such dynamics have
important implications for knowledge transfer between network members and impact the pro
posed facilitating conditions. Organizational so cial capital, for example, takes priority over in dividual social capital in strategic alliances, whereas the reverse is true in industrial dis
10 The existence of an intracorporate network means that
some organizational social capital will exist (by default) and
some knowledge will flow between the members. When an
alliance is formed, there is at least some social capital, because a network tie has been formed. However, the basic
structural tie between the partners (in the form of an alliance
agreement) in the absence of additional social capital means that a large-quantity knowledge flow is unlikely. An
interesting and challenging empirical question involves the
minimum level of organizational and individual social cap ital necessary to effect knowledge transfer. Even more com
plicated is the comparative question, "How much social cap ital is necessary to effect knowledge transfer in an
intracorporate network versus an alliance network versus
an industrial district?"
160 Academy of Management Review January
tricts. For intracorporate networks, we have sug
gested that social capital that impacts knowl
edge transfer will begin at an organizational level and then be enhanced by social capital developments at the individual level.
Network Boundary Conditions
Our analysis helps identify boundary condi tions associated with each network type, with
respect to intermember knowledge transfer. For
example, in his study of knowledge sharing within an intracorporate network, Tsai (2002) maintains that social interaction among organi zational units facilitates intraorganizational knowledge sharing. He argues further that
"knowledge sharing among competing units within the same organization carries synergistic benefits because these units deal with similar resource constraints and market situations"
(2002: 182). Therefore, Tsai proposes that social interaction is more positively associated with
knowledge sharing among units that compete with each other than among units that do not
(i.e., his Hypothesis 4). Tsai's argument is valid in the case of intra
corporate networks, where organizational social
capital is readily available between network members and the development of individual so
cial capital will help knowledge transfer or
sharing. Although network members may com
pete with each other, the risk of knowledge leak
age from one unit to another is not a major concern, because these units work under the same corporate roof.
However, the situation will be different in al liance networks. When alliance partners are
keen competitors, interpartner learning may be come a race (Hamel, 1991). Partners avoid inad
vertently transferring knowledge beyond what is stated in the alliance agreement. Managers
who are appointed to the alliance will be well aware of this risk and more cautious in their social interaction. Hence, we expect exactly the
opposite of Tsai's hypothesis to happen?that is, social interaction will be more positively asso
ciated with knowledge sharing among alliance
partners that do not compete with each other than among partners that do. In fact, when alli ance partners are keen competitors and manag ers strictly follow the rules set by their respec tive parent companies, social interaction may not have any impact on knowledge sharing.
DISCUSSION
In examining various social capital dimen
sions, we address an issue raised by Uzzi and
Gillespie (2002): structural approaches to net works that ignore social qualities inadequately specify how networks function. We also address Kostova and Roth's (2003) call for more research on the outcomes of social capital. This paper shows that each network type has distinct social
capital dimensions. By linking these dimensions to knowledge transfer in networks, a social cap ital outcome, we show that the facilitating con
ditions vary across networks. For effective and efficient knowledge transfer to occur, firms may have to manage and build social capital proac
tively. The conditions identified can be viewed as predictive conditions and provide guidance for firms seeking to exploit network knowledge opportunities.
Implications for Future Research
An examination of the conditions facilitating learning and knowledge transfer (Table 2) re
veals some implications for future research. First and obviously, the sheer number of rela
tionships illustrates the complexity of this area.
The introduction of social capital variables into the analysis of networks and knowledge trans fer adds a level of complexity that has not yet been examined empirically.
Second, virtually all the existing theoretical and empirical studies of interorganizational knowledge transfer are based on a single net work type, without any reference to the bound
ary conditions. The question of how far the re sults of these studies can be generalized from one network type to another rarely has been examined. The distinct facilitating conditions across network types listed in Table 2, and the
preceding discussion of social capital levels,
suggest that generalizability across network
types may be limited and that a contingency approach is appropriate. In other words, pro cesses of interorganizational knowledge trans
fer are affected by the nature of the network type in which the organizations are embedded.
Third, although facilitating conditions are dis tinct across network types, there is value to be
gained by integration and synthesis. The litera ture we examined crossed a diverse terrain, from intracorporate networks to industrial dis
2005 Inkpen and Tsang 161
tricts. In examining the literature streams, which at present are not well integrated or
linked, we realized that all networks are, at their core, about social relationships, and, therefore, social capital dimensions have applicability, re
gardless of the network type. In reviewing the literature on knowledge transfer in organiza tions, Argote and Ingram (2000) argue that social networks play an important role in knowledge transfer and yet related research is inadequate. In this article we address this deficiency and
provide an agenda for future research. Gulati et al. (2000) suggest that incorporating
networks into strategic analysis leads to a more
comprehensive view of the strategic behavior of firms. We agree but would argue that, to truly understand network behavior, researchers should move beyond one-size-fits-all analyses of networks. Osigweh comments that
when concepts are broadened in order to extend
their range of applications, they may be so
broadly defined (or, stretched) that they verge on
being too all-embracing to be meaningful in the realm of empirical observation and professional practice (1989: 582).
The concept of network is one that suffers from
being overstretched. As we have shown, the dy namics of knowledge transfer vary across net work types. We illustrate that social capital di mensions are not uniform in their effects on
knowledge transfer. Rather, they vary across dif ferent types of networks. Network theories that fail to distinguish between network types will be unable to capture the complex variety of fac tors associated with network knowledge pro cesses. These theories need to develop beyond the early, broad theoretical discussions that
were based on a generic type of network (e.g., Jarillo, 1988; Thorelli, 1986) and to examine in de tail the characteristics of different network types.
Social capital is still in the "emerging excite ment" stage of the life cycle typical of an um brella concept and will face validity challenges in its next stage of development (Hirsch & Levin, 1999). The social capital concept has been used
extensively by scholars in discussing interper sonal or interorganizational relationships of a certain type. Yet the concept seldom has been
applied to compare and contrast different types of relationships. By addressing this deficiency, we have shown that the concept, in the form of both individual and organizational social capi tal, is useful for distinguishing between differ
ent network types and that each social capital facet gives rise to different knowledge transfer
facilitating conditions across network types. Through such a compare-and-contrast analysis, we have not only clarified the meanings of the dimensions and facets of social capital but also demonstrated the value added provided by the social capital concept. Further theoretical anal
yses like ours will help bring the concept to the next stage of its life cycle.
In addition to the specific hypotheses sug
gested in Table 2, which provide substantial
scope for new inquiry, several research direc tions can be identified. As we indicated earlier, the locus of network knowledge processes can
be the network or actors, or both. Although we
focused on knowledge transfer by actors that resulted from their participation in the network, there are various interesting questions involv
ing knowledge creation by the network. For ex
ample, if the network enhances its knowledge base, how do individual actors share in the en
hancement, and which social capital factors are most critical in disseminating network knowl
edge? In an industrial district, the network with the most fluid boundaries, how does the network
knowledge base get impacted by the entry and exit of network members? In alliances, the term common benefits is synonymous with network
knowledge acquisition (Khanna, Gulati, & No
hria, 1998). Are common benefits more likely to
emerge under certain network configurations or
conditions? Another important research question involves
the risks of social capital. Although our main thesis is that social capital has an important positive influence on knowledge transfer by net work actors, it must be noted that social capital is not without risks. Uzzi (1997) argues that over embeddedness could inhibit knowledge flow into the network. For example, when firms
within an industrial district establish intense network ties, they tend to pay little attention to the strategies and capabilities of competitors outside the district, resulting in a blind spot situation (Pouder & St. John, 1996). Hansen's
(2002) research on knowledge sharing in intra
corporate networks shows that a focal unit's di rect ties in a knowledge network had pros and cons. The ties provided immediate access to other business units that possessed related
knowledge. However, the ties were costly to maintain. Hansen argues that direct ties are
162 Academy of Management Review January
most effective when they help units deal with dif ficult transfer situations, which probably involve noncodifiable knowledge. When the transfer is not
difficult, the ties are likely to be harmful for unit effectiveness because of their maintenance costs.
We have discussed the three dimensions of social capital independently. In future research scholars should also examine the interaction effects among these dimensions. For instance, network stability (a facet of the structural di
mension) may help develop trust (a facet of the relational dimension). A facilitating condition
may have effects on more than one dimension. Personnel transfer between members of an in
tracorporate network, for example, may contrib ute to establishing social network ties (a facet of the structural dimension), shared culture (a facet of the cognitive dimension), and trust (a facet of the relational dimension).
Finally, although space limitations preclude a
discussion of knowledge types, other research has shown that different knowledge types have different effects on organizational processes
(e.g., Nonaka, 1994). For instance, for effective transfer of tacit knowledge between network
members, individual social capital must be de
veloped, because the transfer normally requires intimate personal interactions. In future re
search scholars may examine how social capi tal dimensions affect the transfer of different
knowledge types.
Limitations
This paper is not without limitations. First, we
discuss only three major network types. Other
important networks and interesting social capi tal relationships exist. Second, our discussion of
major network types is limited, in that it applies to the more typical members of each network
type. There are inevitably exceptions. Third, there are other factors besides social capital factors that impact network knowledge transfer.
Finally, there are multiple facilitating condi tions for each facet of the social capital dimen
sions; we identified what we view as the most
critical ones. For example, a condition we were
unable to explore for industrial districts and
knowledge transfer is institutional thickness in
the form of an interlocking and integrated web of supportive organizations, including firms, lo
cal authorities, development agencies, financial
institutions, local chambers of commerce, trade
associations, and so on (Amin & Thrift, 1995). Thickness involves collective representation and common purpose (Keeble, Lawson, Moore, &
Wilkinson, 1999) and helps industrial district firms develop a common vision.
Conclusion
We believe that network researchers must consider the potential conceptual differences across various network types. In this article we
partially integrate the voluminous network and
organizational knowledge literature and pro vide a common predictive basis for comparing knowledge transfer determinants across differ ent types of networks. When studying network
behaviors, it is thus important to first examine the nature of the network type concerned and how it differs from other types. This practice is in line with the recent call for contextualizing or
ganizational research (Rousseau & Fried, 2001). Contextualization makes theoretical models more accurate and interpretation of empirical results more robust.
REFERENCES
Adler, P. S., & Kwon, S.-W. 2002. Social capital: Prospect for a
new concept. Academy of Management Review, 27:17-40.
Almeida, P., & Kogut, B. 1999. Localization of knowledge and
the mobility of engineers in regional networks. Manage ment Science, 45: 905-917.
Amin, A., & Thrift, N. 1995. Globalization, institutional "thick
ness" and the local economy. In P. Healey, S. Cameron? S. Davoudi, S. Graham, & A. Madani-Pour (Eds.), Man
aging cities: The new urban context: 92-108. Chichester, UK: Wiley.
Anand, V., Glick. W. H., & Manz, C. C. 2002. Thriving on the
knowledge of outsiders: Tapping organizational social
capital. Academy of Management Executive. 16(1): 21-31.
Anderson, E. 1990. Two firms, one frontier: On assessing joint venture performance. Sloan Management Review. 31(2): 19-30.
Argote, L., Beckman, S. L., & Epple, D. 1990. The persistence and transfer of learning in industrial settings. Manage ment Science. 36: 140-154.
Argote, L., & Ingram, P. 2000. Knowledge transfer: A basis for
competitive advantage in firms. Organizational Behav
ior and Human Decision Processes. 82: 150-169.
Ari?o, A., & de la Torre, J. 1998. Learning from failure: To
wards an evolutionary model of collaborative ventures.
Organization Science. 9: 306-325.
Baker, W. E. 2000. Achieving success through social capital:
Tapping the hidden resources in your personal and busi
ness networks. San Francisco: Jossey-Bass.
2005 Inkpen and Tsang 163
Baum, J. A. C, Calabrese, T., & Silverman, B. S. 2000. Don't go it alone: Alliance network composition and startups'
performance in Canadian biotechnology. Strategic
Management Journal. 21: 267-294.
Baum, J. A. C, & Ingram, P. 1998. Survival-enhancing learn
ing in the Manhattan hotel industry, 1898-1980. Manage ment Science. 44: 996-1016.
Belliveau, M. A., O'Reilly. C. A., Ill, & Wade, J. B. 1996. Social
capital at the top: Effects of social similarity and status
on CEO compensation. Academy of Management Jour
nal. 39: 1568-1593.
Bhagat, R. S., Kedia, B. S., Harveston, P. D., & Triandis, H. C.
2002. Cultural variation in the cross-border transfer of
organizational knowledge. An integrative framework.
Academy of Management Review. 27: 204-221.
Blau, P. M. 1964. Exchange and power in social life. New
York: Wiley.
Bolino, M. C Turnley, W. H., & Bloodgood, J. M. 2002. Citi
zenship behavior and the creation of social capital in
organizations. Academy of Management Review. 27:
505-522.
Bourdieu P. 1986. The forms of capital. In J. G. Richardson
(Ed.), Handbook of theory and research for the sociology of education: 241-258. New York: Greenwood.
Brown, J. E., & Hendry, C. 1998. Industrial districts and supply chains as vehicles for managerial and organizational
learning. International Studies of Management & Orga nization. 27(4): 127-157.
Burt, R. S. 1992. Structural holes: The social structure of com
petition. Cambridge, MA: Harvard University Press.
Burt, R. S. 1997. The contingent value of social capital. Ad
ministrative Science Quarterly. 42: 339- 365.
Camagni, R. 1995. Global network and local milieu: Towards
a theory of economic space. In S. Conti, E. J. Malecki, &
P. Oinas (Eds.), The industrial enterprise and its environ
ment: Spatial perspectives: 195-214. Aldershot, UK: Ave
bury.
Capello, R. 1999. Spatial transfer of knowledge in high tech
nology milieux: Learning versus collective learning pro cesses. Regional Studies. 33: 353-365.
Carley, K. 1992. Organizational learning and personnel turn
over. Organization Science. 3: 20-46.
Coleman, J. S. 1988. Social capital in the creation of human
capital. American Journal of Sociology. 94(Supplement): S95-S120.
Das, T. K., & Teng, B.-S. 1998. Between trust and control:
Developing confidence in partner cooperation in alli
ances. Academy of Management Review. 123: 491-513.
Das, T. K., & Teng, B.-S. 2002. Alliance constellations: A social
exchange perspective. Academy of Management Re
view. 27: 445-456.
Dodgson, M. 1993. Learning, trust, and technological collab
oration. Human Relations. 46: 77-95.
Doz, Y. L. 1996. The evolution of cooperation in strategic alliances: Initial conditions or learning processes? Stra
tegic Management Journal. 17(Summer Special Issue):
55-84.
Drazin, R. 1999. Multilevel theorizing about creativity in or
ganizations: A sensemaking perspective. Academy of
Management Review. 24: 286-307.
Dyer, J. H., & Nobeoka, K. 2000. Creating and managing a
high-performance knowledge-sharing network: The
Toyota case. Strategic Management Journal. 21 (Special Issue): 345-367.
Dyer, J. H., & Singh, H. 1998. The relational view: Cooperative
strategy and sources of interorganizational competitive
advantage. Academy of Management Review. 23: 660
679.
Gargiulo, M., & Benassi, M. 2000. Trapped in your own net?
Network cohesion, structural holes, and the adaptation of social capital. Organization Science. 11: 183-196.
Ghoshal, S., & Bartlett, C. A. 1990. The multinational corpo ration as an interorganizational network. Academy of
Management Review. 15: 603-625.
Ghoshal, S., Korine, H., & Szulanski, G. 1994. Interunit com
munication in multinational corporations. Management Science. 40: 96-110.
Grant, R. M. 1996. Toward a knowledge-based theory of the
firm. Strategic Management Journal, 17(Winter Special Issue): 109-122.
Gulati, R. 1995. Does familiarity breed trust? The implica tions of repeated ties for contractual choice in alliances.
Academy of Management Journal. 38: 85-112.
Gulati, R. 1998. Alliances and networks. Strategic Manage ment Journal. 19: 293-317.
Gulati, R., Nohria, N., & Zaheer, A. 2000. Strategic networks.
Strategic Management Journal. 21 (Special Issue): 203
215.
Gupta, A. K., & Govindarajan, V. 1991. Knowledge flows and
the structure of control within multinational corpora tions. Academy of Management Review, 16: 768-792.
Gupta, A. K., & Govindarajan, V. 2000. Knowledge flows
within multinational corporations. Strategic Manage ment Journal, 21: 473-496.
Hamel, G. 1991. Competition for competence and inter
partner learning within international strategic alli
ances. Strategic Management Journal, 12(Summer Spe cial Issue): 83-103.
Hamel, G., Doz, Y. L., & Prahalad, C. K. 1989. Collaborate with
your competitors?and win. Harvard Business Review,
67(1): 133-139.
Hannan, M. T., & Freeman, J. 1977. The population ecology of
organizations. American Journal of Sociology, 82: 929
964.
Hansen, M. T. 2002. Knowledge networks: Explaining effec
tive knowledge sharing in multiunit companies. Orga nization Science. 13: 232-248.
Helmsing, B. 2001. Externalities, learning and governance: New perspectives on local economic development. De
velopment & Change, 32: 277-308.
Hill, C. W., Hitt, M. A., & Hoskisson, R. E. 1992. Cooperative
164 Academy of Management Review January
versus competitive structures in related and unrelated
diversified firms. Organization Science, 3: 501-521.
Hirsch, P. M., & Levin, D. Z. 1999. Umbrella advocates versus
validity police: A life-cycle model. Organization Sci
ence, 10: 199-212.
Inkpen, A. C, & Beamish, P. W. 1997. Knowledge, bargaining
power, and the instability of international joint ven
tures. Academy of Management Review, 22: 177-202.
Inkpen, A. C, & Currall, S. C. 2002. A multilevel approach to
trust in joint ventures. Journal of International Business
Studies, 33: 479-495.
Inkpen, A. C, & Dinur, A. 1998. Knowledge management
processes and international joint ventures. Organiza tion Science, 9: 454-468.
Jarillo, J. C. 1988. On strategic networks. Strategic Manage ment Journal, 9: 31-41.
Kale, P., Singh, H., & Perlmutter, H. 2000. Learning and pro tection of proprietary assets in strategic alliances:
Building relational capital. Strategic Management Jour
nal, 21: 217-237.
Keeble, D., Lawson, C, Moore, B., & Wilkinson, F. 1999. Col
lective learning processes, networking and "institu
tional thickness" in the Cambridge region. Regional Studies, 33: 319-332.
Keeble, D., & Wilkinson, F. 1999. Collective learning and
knowledge development in the evolution of regional clusters of high technology SMEs in Europe. Regional Studies, 33: 295-303.
Khanna, T., Gulati, R., & Nohria, N. 1998. The dynamics of
learning alliances: Competition, cooperation, and rela
tive scope. Strategic Management Journal, 19: 193-210.
Kim, D. H. 1993. The link between individual and organiza tional learning. Sloan Management Review, 35(1): 37-50.
Kogut, B. 1988. Joint ventures: Theoretical and empirical per
spectives. Strategic Management Journal, 9: 319-332.
Koka, B. R., & Prescott, J. E. 2002. Strategic alliances and
social capital: A multidimensional view. Strategic Man
agement Journal. 23: 795-816.
Kostova, T. 1999. Transnational transfer of strategic organi zational practices: A contextual perspective. Academy of Management Review. 24: 306-324.
Kostova, T., & Roth, K. 2003. Social capital in multinational
corporations and a micro-macro model of its formation.
Academy of Management Review. 28: 297-317.
Kotabe, M., Martin, X., & Domoto, H. 2003. Gaining from
vertical partnerships: Knowledge transfer, relationship duration, and supplier performance improvement in the
U.S. and Japanese automotive industries. Strategic Man
agement Journal. 24: 293-316.
Krackhardt, D. 1992. The strength of strong ties. In N. Nohria
& R. G. Eccles (Eds.), Networks and organizations: Struc
ture, form and action: 216-239. Boston: Harvard Business
School Press.
Lane, P. J., & Lubatkin, M. 1998. Relative absorptive capacity and interorganizational learning. Strategic Manage ment Journal. 19: 461-477.
Larson, A. 1992. Network dyads in entrepreneurial settings: A
study of the governance of exchange relationships. Ad
ministrative Science Quarterly. 37: 76-104.
Lawson, C. 1999. Toward a competence theory of the region.
Cambridge Journal of Economics. 23: 151-166.
Lazerson, M. H., & Lorenzoni, G. 1999. The firms that feed
industrial districts: A return to the Italian source. Indus
trial and Corporate Change. 8: 235-266.
Leana, C. R., & Van Buren, H. J., III. 1999. Organizational social capital and employment practices. Academy of
Management Review. 24: 538-555.
Lee, C, Lee, K., & Pennings, J. M. 2001. Internal capabilities, external networks, and performance: A study on tech
nology-based ventures. Strategic Management Journal. 22: 615-640.
Leenders, R. T. A., & Gabbay, S. M. 1999. An agenda for the
future. In R. T. A. Leenders & S. M. Gabbay (Eds.), Cor
porate social capital and liability: 483-494. New York:
Kluwer.
Lyles, M. A. 1988. Learning among joint venture sophisti cated firms. Management International Review. 28(Spe cial Issue): 85-97.
MacKinnon, D., Cumbers, A., & Chapman, K. 2002. Learning, innovation and regional development: A critical ap
praisal of recent debates. Progress in Human Geogra
phy, 26: 293-311.
Makhija, M. V., & Ganesh, U. 1997. The relationship between
control and partner learning in learning-related joint ventures. Organization Science, 8: 508-527.
Marshall, A. 1920. Principles of economics. London: Macmil
lan.
Maskell, P., & Malmberg, A. 1999. Localized learning and
industrial competitiveness. Cambridge Journal of Eco
nomics, 23: 167-185.
Nahapiet, J., & Ghoshal, S. 1998. Social capital, intellectual
capital, and the organizational advantage. Academy of
Management Review, 23: 242-266.
Nishiguchi, T. 1994. Strategic industrial sourcing: The Japa nese advantage. New York: Oxford University Press.
Nonaka, I. 1994. A dynamic theory of organizational knowl
edge creation. Organization Science, 5: 14-37.
Nooteboom, B., Berger, H.. & Noorderhaven, N. G. 1997. Effects
of trust and governance on relational risk. Academy of
Management Journal, 40: 308-338.
Orlikowski, W. J. 2002. Knowing in practice: Enacting a col
lective capability in distributed organizing. Organiza tion Science, 10: 249-273.
Osigweh, C. A. B. 1989. Concept fallibility in organizational science. Academy of Management Review, 14: 579-594.
Paniccia, I. 1998. One, a hundred, thousands of industrial
districts: Organizational variety in local networks of
small and medium-sized enterprises. Organization Studies, 19: 667-699.
Parkhe, A. 1991. Interfirm diversity, organizational learning, and longevity in global strategic alliances. Journal of
International Business Studies, 22: 579-601.
2005 Inkpen and Tsang 165
Parkhe, A. 1993. Strategic alliance structuring: A game the
oretic and transaction cost examination of interfirm co
operation. Academy of Management Journal, 36: 794
829.
Phan, P. H., & Peridis, T. 2000. Knowledge creation in strate
gic alliances: Another look at organizational learning. Asia Pacific Journal of Management, 17: 201-222.
Pitts, R. A., & Lei, D. 1997. Building cooperative advantage:
Managing strategic alliances to promote organizational
learning. Journal of World Business, 32: 203-223.
Podolny, J. M., & Page, K. L. 1998. Network forms of organi zation. Annual Review of Sociology, 24: 57-76.
Portes, A. 1998. Social capital: Its origins and applications in
modern sociology. Annual Review of Sociology, 24: 1-24.
Pouder, R., & St. John, C. H. 1996. Hot spots and blind spots:
Geographical clusters of firms and innovation. Acad
emy of Management Review, 21: 1192-1225.
Powell, W. W. 1990. Neither markets nor hierarchy: Network
forms of organization. Research in Organizational Be
havior. 12: 295-336.
Powell, W. W., Koput, K. W., & Smith-Doerr, L. 1996. Interor
ganizational collaboration and the locus of innovation:
Networks of learning in biotechnology. Administrative
Science Quarterly. 41: 116-145.
Putnam, R. D. 1993. The prosperous community: Social capi tal and public life. American Prospect. 13: 35-42.
Rousseau, D. M., & Fried, Y. 2001. Location, location, location:
Contextualizing organizational research. Journal of Or
ganizational Behavior. 22: 1-13.
Rousseau, D. M., Sitkin, S., Burt, R. S., & Camerer, C. 1998. Not so different after all: A cross-discipline view of trust.
Academy of Management Review. 23: 393-404.
Rowley, T., Behrens, D., & Krackhardt, D. 2000. Redundant
governance structures: An analysis of structural and relational embeddedness in the steel and semiconduc tor industries. Strategic Management Journal. 21: 369 386.
Schaan, J.-L. 1988. How to control a joint venture even as a
minority partner. Journal of General Management. 14(1): 4-16.
Schnake, M. E., & Cochran, D. S. 1985. Effect of two goal setting dimensions on perceived intraorganizational conflict. Group & Organization Studies. 10: 168-183.
Scott, A. 1992. The role of large producers in industrial dis tricts: A case study of high technology systems houses in southern California. Regional Studies. 26: 265-275.
Spender, J.-C. 1989. Industry recipes: The nature and sources
of managerial judgement. Oxford: Blackwell.
Storper, M. 1993. Regional "words" of production: Learning and innovation in the technology districts of France,
Italy and the USA. Regional Studies. 27: 433-455.
Thorelli, H. B. 1986. Networks: Between markets and hierar
chies. Strategic Management Journal. 7: 37-51.
Tsai, W. 2002. Social structure of "cooperation" within a
multiunit organization: Coordination, competition, and
intraorganizational knowledge sharing. Organization Science. 13: 179-190.
Tsai, W., & Ghoshal, S. 1998. Social capital and value cre
ation: The role of intrafirm networks. Academy of Man
agement Journal. 41: 464-476.
Useem, M., & Karabel, J. 1986. Pathways to top corporate
management. American Sociological Review. 44: 184?
200.
Uzzi, B. 1997. Social structure and competition in interfirm
networks: The paradox of embeddedness. Administra
tive Science Quarterly, 42: 35-67.
Uzzi, B., & Gillespie, J. J. 2002. Knowledge spillover in corpo rate financing networks: Embeddedness and the firm's
debt performance. Strategic Management Journal, 23:
595-618.
Von Hippel, E. 1987. Cooperation between rivals: Informal
know-how trading. Research Policy, 16: 291-302.
Von Krogh, G., Roos, J., & Slocum, K. 1994. An essay on
corporate epistemology. Strategic Management Journal,
15(Summer Special Issue): 53-71.
Yan, A. 1998. Structural stability and reconfiguration of in
ternational joint ventures. Journal of International Busi ness Studies, 29: 773-796.
Yan, A., & Zeng, M. 1999. International joint venture instabil
ity: A critique of previous research, a reconceptualiza tion, and directions for future research. Journal of Inter
national Business Studies, 30: 397-414.
Yli-Renko, H., Autio, E., & Sapienza, H. J. 2001. Social capital,
knowledge acquisition, and knowledge exploitation in
young technology-based firms. Strategic Management Journal. 22: 587-613.
Zahra, S. A., Ireland, R. D., & Hitt, M. A. 2000. International
expansion by new venture firms: International diversity, mode of market entry, technological learning, and per formance. Academy of Management Journal. 43: 925-950.
Andrew C. Inkpen is professor of management at Thunderbird, the Garvin School of International Management. He was a visiting professor at Nanyang Business School,
Nanyang Technological University, when this article was written. He received his Ph.D. from The University of Western Ontario. His research interests are in interna tional strategy, strategic alliances and joint ventures, organizational trust, and knowl
edge management.
Eric W. K. Tsang is an assistant professor in the School of Business Administration at
Wayne State University. He received his Ph.D. from the University of Cambridge. His current research interests include organizational learning, strategic alliance, super stitious decision making, and research methods.