Technological Forecasting & Social Change 80 (2013) 1261–1270

Contents lists available at SciVerse ScienceDirect

Technological Forecasting & Social Change

Social capital, absorptive capability, and firm innovation

Sui-Hua Yu ⁎
Department of Accounting, National Chung Hsing University, 250, Kuo Kuang Rd., Taichung 402, Taiwan, ROC

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of this study is to examine the relationship between social capital and firm
Received 2 February 2012 innovation as well as the moderating effects of absorptive capability. In general, network-level
Received in revised form 9 August 2012 technological diversity is utilized to measure a firm's social capital and in this case, an empirical
Accepted 16 December 2012 investigation of 748 high-tech firms in Taiwan indicates that such diversity in a firm's social
Available online 10 January 2013
network has an inverted U-shaped relationship with the firms' innovation performance. In
addition, we find that by increasing absorptive capability, it will systematically increase the slope
Keywords: and amplitude of the positive effects of diversity as well as maximize the overall value of
Social capital network-level technological diversity for firm innovation performance. These results also suggest
Firm innovation
that an overabundance of network-level technological diversity can have a negative impact on
Network technological diversity
firm innovation while a moderate level of such diversity can lead to high levels of innovation
Absorptive capability
performance.
© 2012 Elsevier Inc. All rights reserved.

1. Introduction influence the amount of resources and information a firm can

Intangible assets are crucial for a firm's overall success,
especially in the current knowledge-based economy. By 2002,
75% of a firm's market value is due to its intangible assets,
which shows that effective management of such assets is the
key to achieving superior performance [1]. Discussions in
regard to social capital, one such intangible asset, have recently
gained prominence. Social capital is generally defined as the
resources embedded in a social network that can be mobilized
by a firm in order to increase the success rate for specific actions
[2]. Previous studies have indicated that social networks are
important channels for firms to access external knowledge.
Through such networks, firms can obtain pertinent information
and knowledge that helps them enhance their overall learning
and innovation (e.g. [3–5]). Despite these positive aspects,
there are several limitations in current research that restrict
further understanding of how social capital can improve firm
innovation.
First, few research consider the differences in social net-
works while studying the value of social capital for innovation
(e.g. [6–8]). In fact, the composition of a firm's social network
⁎ Tel.: +886 4 228 40828#675; fax: +886 4 228 58542.
E-mail address: shyu@dragon.nchu.edu.tw.
obtain from its network [9]. Second, the majority of current
research assumes that the relationship between social capital
and innovation performance is linear (e.g. [10–12]). However,
social network theory indicates that a non-linear relationship
between social capital and firm performance is possible. Third,
some recent studies have found that the benefits of social
capital are contingent upon firm characteristics and contextual
factors, which suggests that scholars must consider these
factors when studying the value of social capital [13–16].
Therefore, this study aims to address the above limita-
tions by examining how the composition of a firm's social
network impacts the firm's innovation performance as well
as investigating the moderating effects of absorptive capa-
bility. For this study, we utilized interlocking directorates in
order to establish the boundary for a firm's social network
while using network-level technological diversity to measure
social capital. This particular investigation occurred between
2006 and 2008, and the subjects consisted of 748 high-tech
firms in Taiwan, all of whom believed that innovation was a
vital component of maintaining a competitive position. Ac-
cording to the results, there is an inverted U-shaped relation-
ship between the technological diversity of a firm's social
network and the firm's innovation performance. This shows
that even though network-level technological diversity can

0040-1625/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.techfore.2012.12.005
1262 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270
help increase innovation, an overabundance of such diversity
can adversely affect innovation performance. In addition, we
found that absorptive capability increases the slope of the
positive effects of network-level technological diversity as well
as reduces the negative effects of such diversity. In other words,
firms with greater absorptive capabilities benefit more from
technologically diverse networks.
In addition, this investigation contributes to the literature
and practice in several ways. First, it proves that the diversity of
a firm's social network can influence its innovation perfor-
mance. This means that more attention should be made toward
network composition while studying the benefits of social
capital. In practice, if firms want to increase innovation through
alliance networks, they need to examine the member compo-
sition of the particular network in which they are embedded.
Second, in previous literature, absorptive capability is consid-
ered as an important factor that influences firm innovation
[17,18]. Since a limited number of studies have analyzed
whether a firm's absorptive capability increases the overall
value of its external network, this investigation provides pre-
liminary evidence that shows how a firm's absorptive capabil-
ity does in fact help the firm obtain and absorb external
knowledge. In particular, the greater the diversity a network
provides, the more absorptive capability is required. Based on
this framework, a firm cannot simply rely upon external
networks in order to obtain knowledge; it must also adequately
invest in the development of its own technical capacity.
Many studies suggest that social networks provide channels
for firms to obtain knowledge, resources and technology
as well as promote internal learning in order to help them
innovate [19–21]. However, there is hardly any consensus in
the accompanying literature. In this case, this study also
proposes that the relationship between social network and
firm innovation is non-linear in an attempt to explain such
inconsistencies and help firms realize that only with moderate
investments in social capital can they obtain maximum value.
Finally, most of the current research on the relationship
between social capital and performance has been conducted
based on the individual [22], the project [23], or the industry
itself [24,25]. It is only recently that studies have been
conducted based on the organization or firm [9,26]. There-
fore, the overall analysis in this study is conducted based
on the firm in order to show that specific composition of
interfirm networks are factors in the construction of social
capital.
This paper proceeds as follows. Section 2 presents a liter-
ature review of previous social capital research, and derives
two hypotheses for how network composition influences firm
innovation. Section 3 provides descriptions in regard to the
data and variables employed in the analysis while the empirical
results are discussed in Section 4. Finally, Section 5 provides a
discussion of the implications of this study and its overall
limitations.
2. Theory and hypotheses development
2.1. Definition of social capital
Extensive discussions in regard to social capital began
during the 1980s, and there are multiple definitions of this
abstract concept. For example, according to Bourdieu [27],
social capital is the aggregation of actual and potential
resources within a specific network, where the network is
composed of relationships that involve mutual acquaintance
and mutual recognition. Flap [28,29] defines social capital as
the resources that the ego obtains from stable relationships or
the accessibility and willingness of alternate relationships to
provide such resources. However, most scholars agree that
social capital is an investment of social relationships by an
individual or organization expecting returns from the market
[30–33]. More specifically, social capital is a resource that is
embedded within the social network that can be mobilized by
the actor (e.g., the firm or organization) in order to increase the
success rate for specific actions [2].
According to this definition, social capital has two main
characteristics: 1) it is embedded within the social network;
and 2) it is a resource that can be obtained and used by the
actor. This shows that an inseparable relationship exists be-
tween social networks and the development of social capital.
Network composition refers to the types of actors within a
specific network, which may be identified by their character-
istics, features, or resource endowments [34]. It can also be
described by using the links between groups. Some earlier
studies have analyzed firms' relationships with various orga-
nizations [35–37] while more recent studies have examined
the similarities and differences between linked members
[38,39,9]. Typically, the heterogeneity in network composition
can be described as network diversity. The theoretical
definition of such diversity is the degree to which the system
is composed of uniquely different elements, the frequency
distribution of these elements, and the degree of difference
between them [40]. The diversity of social networks is the
difference in capability, technology, resources, or knowledge
between members of the networks. In this case, the higher the
diversity, the more different types of technology, resources,
and knowledge can be acquired, absorbed, and utilized by firms
within the network [41,16]. There is an aspect of network
diversity called “network technological diversity,” which is the
difference in regard to the types of technology that the
members are attempting to develop [9]. When the technology
is more diverse, the probability of a firm obtaining new
knowledge increases. This shows that they are more likely to
develop innovative solutions to problems and increase inno-
vation output [42–44]. Therefore, since network technological
diversity is a vital characteristic of social networks for high-
tech firms, it is an important point of analysis for this study.
2.2. The relationship between network technological diversity
and firm innovation
According to network theory, different firms have differ-
ent resources and capabilities and thus establishing a linkage
between previously unconnected firms can help in the inte-
gration of diversity or non-overlapping technological knowl-
edge. In other words, firms with diverse ties can obtain diverse
knowledge [39]. Therefore, the higher the network techno-
logical diversity, the greater the chance a firm has in obtaining
abundant technological information and resources as well as
receiving novel information that is vital for firm innovation
[42,43]. In addition, this can also help to increase the diversity
and number of interactions between knowledge and technology,
 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270
which increases the chances of discovering innovative solutions
[45].
According to recombinatory search theory, innovation is a
problem-solving process where firms recombine its current
knowledge, problems, and solutions [45,46] or reconfigures the
linkages in knowledge [47] in order to produce new knowledge.
Such knowledge exchange is vital in knowledge recombina-
tion [48] while the difficulty of knowledge exchange is often
influenced by the characteristics of the knowledge itself. For
example, technological knowledge used in innovation is often
defined by being professional, tacit, and embedded [49,46],
which means that it cannot be obtained through trade in the
market [50]. In this case, interfirm relationships become an
effective search mechanism and medium for knowledge transfer
[51] and firms can increase how often they come into contact
with other firms through such relationships. An increase in such
social interaction reduces the ambiguity of knowledge that a
firm obtains from other firms, thus increasing knowledge trans-
fer and assimilation [52]. In addition, the improvement in in-
centive alignment and monitoring mechanisms are also reasons
that interfirm relationships are more useful in the repeated
transfer of tacit knowledge compared to market transactions
[53].
When a firm obtains technological knowledge through
cooperative relationships, discovers potentially useful elements,
and effectively combines different elements or assimilates this
knowledge, it increases the potential for knowledge creation
[54]. However, when a firm obtains technological knowledge
and capabilities that are similar to what they already possess,
even a recombination can only produce incremental inno-
vation [55,49]. In comparison, if a firm obtains diverse and non-
repetitive knowledge, it can challenge the status quo and
discover new ties between knowledge elements [56] while
developing new concepts and solutions for existing prob-
lems [57]. Thus, firms can generate new combinations of
knowledge and create novel innovations [58,55,59].
By combining network theory and recombinatory search
theory, network technological diversity can help increase cor-
porate innovation performance. However, searching for new
and often unfamiliar knowledge can take time and resources,
and many uncertainties exist [60]. Sampson [61,16] found that
increased diversity led to higher costs for problem-solving and
committing to renegotiation, while other scholars found that
diversity decreased the usability and compilation of knowledge
between firms [17,3,14]. Specifically, since firms have limited
experience in regard to unfamiliar knowledge and cognitive
capacity, when the gap of technological capabilities between
firms grows, their ability to understand the uses of such tech-
nology decreases [45]. This also affects their overall ability to
assimilate and integrate knowledge. In addition, processing
highly diverse knowledge can lead to information overload as
well as uncertainties and diseconomies of scale [62]. Therefore,
increases in network-level technological diversity also increase
the costs for absorption and the usage of such knowledge.
In sum, an increase in network-level technological diversity
can help firms obtain more novel knowledge and provide them
with more opportunities to discover new combinations of
knowledge, which increases their overall chances of innova-
tion. However, excessive diversity will decrease a firm's ability
to process and utilize such knowledge, and the costs associated
with decreased knowledge absorption are far greater than the
1263
benefits of obtaining novel knowledge, which negatively
impacts overall innovation performance. Therefore, this study
proposes that there is an inverted U-shaped relationship
between technological diversity of a firm's social network
and innovation performance, which leads to the following
hypothesis:
H1. The technological diversity in a firm's social network
has an inverted U-shaped relationship with the firm's
innovation performance.
2.3. Definition of absorptive capability
As first discussed by Cohen and Levinthal [63,17], the
concept of absorptive capability is defined as a firm's ability to
improve its understanding, evaluation, assimilation, and appli-
cation of external knowledge through investments in specific
activities. Several studies, built upon Cohen and Levinthal's
[63,17] concept, offered new and expanded definitions. Ac-
cording to Grünfeld [64], absorptive capability is a firm's ability
to learn from other firms while Zahra and George [65] defined it
as the active organizational capability of a firm to obtain,
absorb, transform, and utilize external knowledge. Meanwhile,
Tsai [66] defined a firm's absorptive capability as its ability to
successfully replicate new knowledge. In sum, absorptive capa-
bility is a firm's ability to understand, absorb, and apply new
knowledge obtained from external sources. Moreover, the
majority of previous studies have found that firms with greater
absorptive capacities tend to enhance their learning capabili-
ties, which helps them effectively utilize external knowledge
[67–69].
2.4. The moderating effects of absorptive capability
Existing studies find that learning outcomes differ among
firms that are embedded in the same networks. In addition,
absorptive capability has been found to be an important factor
that contributes toward such an uneven knowledge transfer
process [70]. In this case, the innovation activity involves
searching, processing, and integrating knowledge in order
to create novelty. Based on the literature about absorptive
capability [17], the existing knowledge base increases a firm's
ability to search, assimilate, and utilize new knowledge for
problem-solving activities. Thus, firms with higher levels of
absorptive capability can acquire new knowledge from net-
works in order to enhance their innovation activities and obtain
more effective knowledge transfer [71]. This argument is in
agreement with the literature on resource-based views of firms
and their organizational learning processes [72].
There are reasons to expect that greater absorptive capa-
bility enables greater utilization of embedded knowledge
from external networks and as a result, it promotes superior
innovative outcomes [73–75]. Cohen and Levinthal [17] believe
that external knowledge sources are the key to innovation, but
whether a firm could absorb and utilize this knowledge is
based on its experiences and capabilities. The social network of
a firm allows the firm to obtain information through external
knowledge [76,77] and therefore it saves both time and money
for experiments on future innovation [73,74]. A firm's internal
development capabilities enable it to evaluate, understand, and
assimilate the new external knowledge that it obtains from
1264 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270
social networks [73,78]. Therefore, when a firm includes a high
level of absorptive capability, the value created from the
information and knowledge obtained from the social network
is also higher. In other words, firm absorptive capability and
knowledge from external networks complement one another.
More specifically, when firms access diverse knowledge, the
absorptive capability of a firm impacts how much the firm can
improve their innovation through the search, transformation,
and utilization of such knowledge [79]. Innovation is the novel
recombination of knowledge that firms possess and the new
knowledge that they obtain [75,45]. When a firm includes
higher absorptive capability, the search for knowledge be-
comes faster, and the use of diverse technological knowledge
increases overall innovation performance [80]. On the other
hand, when a firm does not have enough capability, the ben-
efits of diverse knowledge become limited. Therefore, this
study suggests that when absorptive capability is higher, firms
are more likely to benefit from diverse technological knowl-
edge in social networks. This leads us to the second hypothesis:
H2. Since absorptive capability moderates the curvilinear
relationship between network-level technological diversity
and innovation performance, increasing absorptive capability
will: (a) increase the slope of the positive effects of diversity;
(b) increase the amplitude of the effects of diversity; and
(c) increase the value of diversity that maximizes innovation
performance.
3. Methodology
3.1. Sample and data collection
The subjects for this study included 748 electronics
industrial firms that were listed in the Taiwan Stock Exchange
Corporation (TWSE). This study identified whether a firm was
in the electronics industry based on the first two digits of the
firm index compiled by the TWSE. During the three-year
research period from 2006 to 2008, a total of 2244 observations
were obtained for the research sample. After deleting the firms
with missing values and those without filed patents during the
research period, the sample included a total of 1235 firm-year
observations. Through the reference by firm index, we col-
lected the number of patents, which were used to measure
innovation performance, and the patent categories for network
diversity, from the R.O.C. Patent Information Search System of
the Taiwan Intellectual Property Office. In regard to patent
categories, we used the International Patent Classifications,
which was divided into eight individual sections (A–H). In
addition, we obtained other variables from the Taiwan Econom-
ic Journal (TEJ) database and the Market Observation Post
System. Finally, in order to define the boundary of social net-
works for sample firms, this study collected the list of directors
from each firm's annual reports.
Finally, this study obtained data particularly from the 2006–
2008 periods based on the following three reasons. First, high
tech firms are generally characterized by a strong emphasis on
innovation and a relatively short product life cycle [81]. Many
previous studies have examined the determinants of firm
innovation using one to three years as sample periods (e.g.
[82,71,83,84]). Such time lengths are considered appropriate
for studying technological innovations of firms [85]. Second,
this study collected the latest patent data available considering
the constant development in the high-tech environment. Since
the duration of patent investigation generally ranges from a
couple of months to three years, this study collected patent
data for each firm based on the date of filing and the number of
patents that were ultimately approved within the subsequent
three-year period as the measure of innovation performance.
In this case, 2008 was the latest year that patent information
was available for this study.
3.2. Dependent variables
Following previous research, we used the number of
approved patents as the measure of a firm's innovation per-
formance as well as the dates they were filed and the number
of patents that were ultimately approved. However, the date
on which the patents were approved can be an inaccurate
portrayal of innovation performance due to time delay. In
addition, since the effect of innovation performance may
extend well into the following year, we used the cumulative
number of approved patents from year t to year t+ 1 in order
to measure a firm's innovation performance.
3.3. Independent variables
3.3.1. Boundaries of social networks
The first step when analyzing social networks is to
establish the boundaries. Generally, there are two ways to
set such boundaries: by geography or by function [86]. When
establishing boundaries by geography, the network bound-
aries are set based on geographical factors such as if the
location is in rural, industrial, or metropolitan regions [87].
When setting boundaries by function, firms are divided ac-
cording to if they rely upon one another or have similar goals
such as interlocking directorates [88], strategic alliances [3],
or participation in common events. For this study, we chose
to establish boundaries by function, especially in regard to
whether the organizations had interlocking directorates. Many
scholars believe that interlocking directorates are the channels
for firms to trade knowledge and strategies and thus, they are
important elements in the formation of social networks for
firms [89,90]. In addition, due to the connections between
boards of directors, these organizations can share resources
and strategies, which can produce the same overall effect as
collaborative networks [91,92]. On the other hand, social
networks can be analyzed on two levels: whole networks and
ego networks [93]. Whole networks are similar to a “bird's eye
view” of the social structure where the focus is on the nodes of
the network instead of the surroundings of any specific node. In
this regard, this particular study focuses on ego networks. In
order to perform this research, we initially found a focal firm,
and then linked that firm to other firms in order to study the
overall impact of the relationship network on the focal firm.
The relationships between the focal firm and other firms are
defined by whether they have interlocking directorates.
3.3.2. Network technological diversity
In previous studies, technological diversity was measured
using several different methods. For example, both Sampson
[16] and Wang [94] studied the distribution of patent categories
within a network while Ahuja [3] focused on the distribution of
 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270 1265

the number of patents within countries. In addition, Talke et al.
[59] researched the diversity of high level management teams.
However, since the majority of previous studies only measured
technological diversity based on the distribution of patents, we
therefore follow the method by Sampson [16] who utilized the
patent categories as a measure for patent distribution. More
specifically, this study focuses on patent combinations within
a three-year timeframe, and uses multi-dimensional vectors
h i
Fi ¼ F1i …Fsi , where Fsi represents the fraction of firm i's
measured by calculating the difference between the year that
the firm was established and year t.
3.4.4. Market power
Since market power can increase product prices and
motivate firms to innovate, it needs to be controlled. Therefore,
this study utilizes Luo's [98] measure for market power, which
is the proportion of a particular firm's sales to the total sales in
the industry.

patents in category s, to capture the distribution of firms in

different patent categories. We then use the following equation
to calculate technological diversity:
′ example, for the year 2006, YR1 = 1, otherwise YR1 = 0. For the
Fi F j
Technological diversity ¼ 1−rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi ; i ≠ j:
 ′  ′ 
Fi F i Fj F j
When the network is larger than the two firms, the firms
are then paired and a calculation using the equation above is
performed for each pair in order to obtain the average of the
resultant values. In this case, the values for network tech-
nological diversity are between 0 and 1, with 1 being the
highest level of diversity and 0 being the lowest.
3.3.3. Absorptive capability
Previous studies have indicated that absorptive capability is
a byproduct of a firm's investments in research and develop-
ment (R&D) and it may also be improved through technolog-
ical training [7,95]. Following previous studies, this study
utilizes the percentage of R&D expenditures (R&D expense / net
sales ∗ 100) in order to measure a firm's absorptive capability.
3.4. Control variables
3.4.5. Year dummies
Since this study uses information from a three-year period,
we added two-year dummies (YR1 and YR2) as controls. For
year 2007, YR2 = 1, otherwise YR2 = 0.
3.4.6. Industry dummies
Since firms differ greatly even within the electronics
industry, this study first divided the industry into eight
subsectors: semiconductors, electronics and peripherals, photo-
electronics, communication networks, electronics and compo-
nents, electronic channels, information services, and other
electronics. Then we established seven dummy variables to
control for the effect of industry dummies: (1) for firms in the
semiconductor industry, IND1 = 1, otherwise IND1 = 0; (2) for
firms in the computer and peripherals industry, IND2 = 1,
otherwise IND2 =0; (3) for firms in the photo-electronics
industry, IND3 = 1, otherwise IND3 = 0; (4) for firms in the
communication networks industry, IND4 = 1, otherwise
IND4 = 0; (5) for firms in the electronics and components
industry, IND5 = 1, otherwise IND5 = 0; (6) for firms in the
electronic channels industry, IND6 = 1, otherwise IND6 = 0;
and (7) for firms in the information services industry, IND7 = 1,
otherwise IND7 = 0.

3.4.1. Network size 3.5. Empirical models

Network size is simply the number of firms in a particular
network. In resource-based theory, a network is a place for In order to examine the impact of network diversity on
learning and resource exchange. Therefore, as the network size innovation performance, we developed the following model
increases, the amount of information and accompanying re- (M1):
sources also increases [18,96]. However, according to transac-
2
tion cost theory, when the number of partners or the members INNit ¼ α0 þ β1 NDIVit þ β2 NDIVit þ β3 NSIZEit þ β4 SIZEit
of the business relationship increases, the complexity of the
þ β5 ABSit þ β6 MKTit þβ7 AGEit þ β8 YR1it
network grows exponentially. As a result, this greatly increases
7
the non-uniformity of the information and the complexity of þ β9 YR2it þ γj ∑j¼1 INDijt þ εit ðM1Þ
management as well as the coordination and management of
costs [97]. In this case, the expansion of the network has a where:
negative impact on innovation performance and it must be
controlled. Following the research by Thorgren [96], this study INNit number of approved patents for firm i in year t
uses the number of firms in a firm's network in order to NDIVit technological diversity of the network for firm i in
measure size of the network. year t
NSIZEit size of the network for firm i in year t
3.4.2. Firm size SIZEit size of firm i in year t
Previous studies have found that the size of a firm had a ABSit absorptive capability of firm i in year t
substantial impact on innovation performance [53]. There- MKTit market power of firm i in year t
fore, this study utilizes the natural log of sales revenue for AGEit age of firm i in year t
firm i in the year t, in order to control for firm size. YR1it dummy variable for firm i in year t
YR2it dummy variable for firm i in year t
3.4.3. Firm age
Since a firm's experience affects its innovation performance, ∑7j¼1 INDijt = dummy variable for firm i as a part of j industry
we use the firm's age to control for business experience. This is in year t, where: Ind1 is the semiconductors industry; Ind2 is
1266 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270

Table 1 firms are part of the electronics industry, there is a substantial

Descriptive statistics. difference in innovation performance. In addition, the standard
Variables N Min Max Mean Std. dev
deviation is 0.086 for network diversity, which means that
heterogeneity in network composition exists. Absorptive capa-
Innovation performance 1235 0 389 9.859 24.868
bility ranged from 0 to 90, which shows that there is a large
Network technological 1235 0 0.943 0.086 0.174
diversity difference between firms in how much they invest in R&D,
Network size 1235 1 14 2.284 2.211 especially in regard to absorptive capability. In terms of control
Absorptive capability 1235 0 90 4.991 6.197 variables, the greatest value for network size is 14, which
Market power 1235 0.000056 0.790 0.013 0.048
means that the largest network for any firm includes a total of
Firm size 1235 4.939 9.168 6.516 0.672
Firm age 1235 0.25 58 17.88 9.021 14 firms. The standard deviations for market power, firm size,
and firm age are all greater than 0, and there are significant
differences between the firms that need to be controlled.
the computer and peripherals industry; Ind3 is the As seen in the correlation matrix of Table 2 below, there is a
photoelectronics industry; Ind4 is the communi- positive correlation between network diversity (NDIV) and
cation networks industry; Ind5 is the electronics and innovation performance (INN). In other words, the composi-
components industry; Ind6 is the electronic chan- tion of social networks helps to promote innovation in firms,
nels industry; and Ind7 is the information services which is the expected outcome. However, the correlation
industry. between absorptive capability (ABS) and innovation perfor-
mance (INN) is inconsistent, which means that the impact of
In order to investigate how absorptive capability moder- absorptive capability on innovation performance may depend
ates the relationship between network technological diver- upon the knowledge obtained from external social networks.
sity and innovation performance, we added two interaction
terms to the model. They are the cross products of network 4.2. The relationship between network technological diversity
technological diversity and absorptive capability and the and firm innovation
cross products of squared network technological diversity
and absorptive capability, respectively. The following model The results of the regression analysis in regard to the
(M2) is thus developed: influence of network technological diversity on innovation
performance are shown in Table 3. The first column utilizes
2
INNit ¼ α0 þ β1 NDIVit þ β2 NDIVit þ β3 NDIVit  ABSit ðM2Þ innovation performance for the current period as the de-
2
pendent variable, and the second column uses cumulative
þ β4 NDIVit  ABSit þ β5 NSIZEit þ β6 SIZEit þ β7 ABSit innovation performance as the dependent variable. For ex-
þ β8 MKTit þβ9 AGEit þ β10 YR1it þ β11 YR2it ample, in the first column, the adjusted R 2 is 0.435 for Model
7 1 and the F value is 60.446, which indicates a suitable model
þ γj ∑j¼1 INDijt þ εit
fit. In terms of the impact of the variables, the coefficient for
network technological diversity is positive and significant
where: The definitions of the variables are the same as (coefficient = 24.982, t = 2.169) while the coefficient for
Eq. (M1). squared network technological diversity is negative and sig-
nificant (coefficient = − 31.605, t = − 2.037). In other words,
4. Results the relationship between network technological diversity
and innovation performance is in the form of an inverted
4.1. Descriptive statistics U-shape, which supports our first hypothesis (H1).
These results show that innovation performance initially
As seen in the descriptive statistics of Table 1 below, the increases with a rise in network technological diversity. How-
firm with the highest innovation performance produced 389 ever, when network technological diversity reaches a certain
patents while the one with the least created none. The standard point, innovation performance declines with the rise of network
deviation is 24.868, which shows that even though all of the technological diversity. This shows that moderate levels of

Table 2
Correlation matrix.

Variables INN NDIV NDIV2 NSIZE ABS MKT SIZE AGE

INN 1 0.231⁎⁎⁎ 0.231⁎⁎⁎ 0.259⁎⁎⁎ 0.057⁎⁎ 0.313⁎⁎⁎ 0.379⁎⁎⁎ −0.027

NDIV 0.084⁎⁎⁎ 1 1 0.910⁎⁎⁎ −0.037 0.364⁎⁎⁎ 0.423⁎⁎⁎ −0.019
NDIV2 0.017 0.922⁎⁎⁎ 1 0.910⁎⁎⁎ −0.037 0.364⁎⁎⁎ 0.423⁎⁎ −0.019
NSIZE 0.223⁎⁎⁎ 0.483⁎⁎⁎ 0.237⁎⁎⁎ 1 −0.040 0.410⁎⁎⁎ 0.461⁎⁎⁎ −0.018
ABS −0.037 −0.052⁎ −0.024 −0.075⁎⁎⁎ 1 −0.368⁎⁎⁎ −0.364⁎⁎⁎ −0.275⁎⁎⁎
MKT 0.625⁎⁎⁎ 0.073⁎⁎⁎ 0.002 0.178⁎⁎⁎ −0.104⁎⁎⁎ 1 0.870⁎⁎⁎ 0.144⁎⁎⁎
SIZE 0.442⁎⁎⁎ 0.200⁎⁎⁎ 0.064⁎⁎ 0.430⁎⁎⁎ −0.332⁎⁎⁎ 0.555⁎⁎⁎ 1 0.154⁎⁎⁎
AGE −0.007 −0.037 −0.041 −0.039 −0.230⁎⁎⁎ 0.056⁎⁎ 0.132⁎⁎⁎ 1

Spearman correlation coefficients are listed in the upper triangle; Pearson correlation coefficients are listed in the lower triangle.
⁎⁎⁎ 1% significance level.
⁎⁎ 5% significance level.
⁎ 10% significance level.
 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270 1267

Table 3
Regression results: the relationship between network technological diversity and innovation performance.

Dependent variable Innovation performance Cumulative innovation performance

Independent variable Coefficients t value Coefficients t value

Constant −5.331 −0.592 −58.729⁎⁎⁎ −4.282

Network technological diversity 24.982⁎⁎ 2.169 31.805⁎⁎ 1.851
Network technological diversity2 −31.605⁎⁎ −2.037 −42.433⁎⁎ −1.835
Network size −0.429 −1.146 −0.872 −1.562
Firm size 1.180 0.842 10.369⁎⁎⁎ 4.835
Absorptive capability 0.311⁎⁎⁎ 2.989 0.633⁎⁎⁎ 4.102
Market power 740.888⁎⁎⁎ 20.653 764.877⁎⁎⁎ 12.961
Firm age −0.032 −0.502 −0.037 −0.400
Year dummy1 6.612⁎⁎⁎ 4.982 9.296⁎⁎⁎ 4.721
Year dummy2 2.154 1.603 1.724 0.864
Industry dummy1 −2.186⁎ −1.002 −7.431⁎⁎ −2.294
Industry dummy2 −2.104 −1.005 −2.527 −0.813
Industry dummy3 −2.152 −1.034 −2.884 −0.935
Industry dummy4 −11.021⁎⁎⁎ −4.655 −15.474⁎⁎⁎ −4.398
Industry dummy5 −0.926 −0.472 −0.343 −0.118
Industry dummy6 −21.180⁎⁎⁎ −4.660 −27.653⁎⁎⁎ −4.090
Industry dummy7 −23.947⁎⁎⁎ −5.525 −29.854⁎⁎⁎ −4.620
N = 1235 Adj. R2 = 0.435 Adj. R2 = 0.338
F = 60.446⁎⁎⁎ F = 40.397⁎⁎⁎
P b 0.0001 P b 0.0001
⁎⁎⁎ 1% significance level.
⁎⁎ 5% significance level.
⁎ 10% significance level.

network technological diversity can produce the best perfor-
mance in firms, while too much network technological diversity
negatively impacts firm innovation. The results of the second
column in Table 3 are similar to the first column. The coefficient
for network technological diversity is positive and significant
(coefficient=31.805, t=1.851) while the coefficient for squared
network technological diversity is negative and significant
(coefficient=−42.433, t=−1.835). In other words, there is a
nonlinear relationship between network technological diver-
sity and innovation performance, which also supports our first
hypothesis (H1).
4.3. The moderating effects of absorptive capability
Table 4 shows the moderating effects of absorptive capability
on the relationship between network technological diversity and

Table 4
Regression results: the moderating effect of absorptive capability.

Dependent variable Innovation performance Cumulative innovation performance

Independent variable Coefficients t value Coefficients t value

Constant −5.085 −0.566 −58.009⁎⁎⁎ −4.226

Network technological diversity 34.852⁎⁎⁎ 2.963 42.335⁎⁎ 2.409
Network technological diversity2 −45.182⁎⁎⁎ −2.842 −57.018⁎⁎ −2.403
Network technological diversity ∗ absorptive capability 10.760⁎⁎⁎ 3.795 11.771⁎⁎⁎ 2.785
Network technological diversity2 ∗ absorptive capability −9.575⁎⁎⁎ −3.230 −10.588⁎⁎ −2.396
Network size −0.526 −1.411 −0.970⁎ −1.739
Firm size 1.090 0.780 10.198⁎⁎⁎ 4.755
Absorptive capability 0.441⁎⁎⁎ 3.984 0.772⁎⁎⁎ 4.687
Market power 755.830⁎⁎⁎ 21.049 784.104⁎⁎⁎ 13.227
Firm age −0.046 −0.728 −0.053 −0.565
Year dummy1 6.576⁎⁎⁎ 4.981 9.257⁎⁎⁎ 4.712
Year dummy2 2.177 1.629 1.752 0.880
Industry dummy1 −2.853 −1.310 −8.143⁎⁎ −2.513
Industry dummy2 −1.926 −0.923 −2.320 −0.747
Industry dummy3 −1.963 −0.948 −2.674 −0.868
Industry dummy4 −11.213⁎⁎⁎ −4.761 −15.697⁎⁎⁎ −4.472
Industry dummy5 −0.855 −0.438 −0.287 −0.099
Industry dummy6 −21.084⁎⁎⁎ −4.663 −27.600⁎⁎⁎ −4.092
Industry dummy7 −23.359⁎⁎⁎ −5.412 −29.298⁎⁎⁎ −4.541
N = 1235 Adj. R2 = 0.441 Adj. R2 = 0.341
F= 55.144⁎⁎⁎ F= 36.533⁎⁎⁎
P b 0.0001 P b 0.0001
⁎⁎⁎ 1% significance level.
⁎⁎ 5% significance level.
⁎ 10% significance level.
1268 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270
firm innovation. The first column includes the results of Model 2
on the independent variables of innovation performance for the
current period while the second column shows the results of
Model 2 on the independent variables of cumulative innovation
performance. According to these results, the coefficient for
network technological diversity is positive and significant
(coefficient=34.852, t=2.963) and the coefficient for squared
network technological diversity is negative and significant
(coefficient=−45.182, t=−2.842), which corresponds to the
results in Table 3. In addition, the coefficient for the cross product
of absorptive capability and network technological diversity is
positive and significant (coefficient=10.760, t=3.795), while
the coefficient for the cross product of absorptive capability and
squared network technological diversity is negative and signif-
icant (coefficient=−9.575, t=−3.230). In other words, the
higher the absorptive capability, the more network technological
diversity positively impacts innovation. In addition, increasing
absorptive capability also increases the amplitude of the impact
and the value of network technological diversity, which
maximizes innovation performance. This supports the second
hypothesis (H2).
The results of the second column in Table 4 are similar to
the first, where the coefficient for network technological
diversity is positive and significant (coefficient= 42.335, t =
2.409) while the coefficient for squared network technological
diversity is negative and significant (coefficient = −57.018,
t = −2.403). In this case, an inverted U-shaped relationship
exists between network technological diversity and innovation
in the following period. The coefficient for the cross products
of absorptive capability and network technological diversity is
positive and significant (coefficient =11.771, t= 2.785) and
the coefficient for the cross products of absorptive capability
and squared network technological diversity is negative and
significant (coefficient= −10.588, t = −2.396). This means
that absorptive capability helps to increase the positive effects
of network diversity on innovation, the amplitude of the im-
pact of network diversity, and the value of network diversity
that maximizes innovation performance. This also supports our
second hypothesis (H2).
5. Discussion and conclusion
Since innovation has gradually moved from closed to open,
external knowledge, particularly those embedded in a firm's
social network, has become an important source for innovation
performance. This research suggests that the composition of a
firm's external network and its internal capabilities are critical
for determining the level of innovation performance.
A firm's external knowledge access is characterized by its
network diversity. The result of this research indicates that a
firm's innovation performance is significantly increased by the
diversity in its external networks, which provides opportuni-
ties to obtain abundant technological knowledge, to access
novel information, and to create innovative outcomes. How-
ever, the results also indicate that the relationship between
network technological diversity and innovation performance is
in the form of an inverted U-shape, which means that the
benefits of excessive diversity may outweigh its costs and
moderate network diversity can lead to significant innovation
performance. Although a diverse network can help firms gain a
wide array of informational benefits, absorbing and utilizing
highly diverse knowledge may cause information overload and
lead toward high costs of such knowledge transfer. The results
of this research demonstrate the importance of gaining access
to knowledge through social networks and, simultaneously
contribute toward the literature on networks and innovations
[18,66,73].
In addition, a firm's internal absorptive capability deter-
mines the extent to which it can absorb, transform, and utilize
new knowledge from other firms [17]. Investing in such
capability also allows a firm to effectively increase its learning
capabilities, and assimilate and apply external knowledge for
its own use [66]. Besides, this research demonstrates that
absorptive capability significantly moderates the relationship
between network technological diversity and firm innovation.
This result suggests that high absorptive capability is associated
with increased absorption and utilization of knowledge diverse
networks, increased amplitude of the effects of diversity on
innovation, and increased value of diversity that maximizes
innovation performance. In other words, there is a compli-
mentary relationship between a firm's absorptive capability
and its external networks, which also shows that improving
innovation performance is one of the most important objec-
tives for firms [99]. Furthermore, previous research has focused
on the direct effects of social networks on firm innovation
without addressing whether the effects might be dependent
upon the extent to which a firm can absorb knowledge
[9,16,100]. In this case, a firm may be able to access knowledge
from its networks but it may not have sufficient capability to
absorb such knowledge. Hence, the higher a firm's overall
absorptive capability, the more it can access knowledge from
external sources. This suggests that a firm must invest signif-
icantly in its absorptive capability, especially when they are
embedded in a more diverse network.
Finally, this research contributes to the growing body of
research on networks in several ways. First, it indicates that
network composition plays an important role in determining
the value of social networks. Second, this research utilizes
network analysis in order to indicate a method of exploring
the relational profiles of firms. Although a few studies have
examined attributes of networks and their influences on
business performance, most of their analyses were based on
the individual [22], the project [23], or the unit level [66]. In
this case, this research examined network composition based
on the firm-level. Third, this research verifies that investing
in absorptive capability while expanding network links is
critical to the success of firms, especially when acquiring new
knowledge that eventually leads to superior innovation. In
this regard, a firm's network diversity reveals its relative
strength when gaining access to novel knowledge, whereas a
firm's absorptive capability reveals its ability to absorb and
utilize such knowledge. The present study shows a nonlinear
relationship between network diversity and firm innovation,
and confirms the moderating role of absorptive capability in
such a relationship. The results also suggest that both the
influence of network composition and absorptive capability
should be studied simultaneously in order to better under-
stand how much a firm can benefit from its social network.
This study has provided several important management
insights regarding how social networks can become an
important source for a firm's innovation performance. In
addition, our research also implied that there is a non-linear
 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270
relationship between social capital and innovation. There-
fore, firms should find the optimal level of network diversity
to reap the greatest benefits for innovation performance.
However, this research also includes two overall limitations.
First, since this research focused only on electronics manu-
facturers, perhaps future research will examine how social
capital can improve innovation in other industries such as
the service industry. Second, subject to data availability, we
were unable to observe how knowledge is transferred among
individual firms and therefore we cannot provide direct
evidence of such transfer processes and mechanisms. In this
case, perhaps future studies can focus on these individual
processes.
Acknowledgments
I would like to acknowledge the National Science Council
of the Republic of China, Taiwan, for financially supporting
this research under contract NSC 99-2410-H-005-020.
References
[1] R.S. Kaplan, D.P. Norton, Strategy Maps: Converting Intangible Assets
into Tangible Outcomes, Harvard Business School Press, Boston, MA,
2003.
[2] N. Lin, Social Capital: A Theory of Social Structure and Action,
Cambridge University Press, 2001.
[3] G. Ahuja, Collaboration networks, structural holes, and innovation: a
longitudinal study, Admin. Sci. Q. 45 (3) (2000) 425–455.
[4] J.H. Dyer, K. Nobeoka, Creating and managing a high-performance
knowledge-sharing network: the Toyota case, Strateg. Manag. J. 21
(3) (2000) 345–367.
[5] A.C. Inkpen, W.K. Tsang, Social capital, networks, and knowledge
transfer, Acad. Manag. Rev. 30 (1) (2005) 146–165.
[6] B.N. Anand, T. Khanna, Do firms learn to create value? The case of
alliances, Strateg. Manag. J. 21 (2000) 295–315.
[7] T.H. Hsu, W. Fang, Intellectual capital and new product development
performance: the mediating role of organizational learning capability,
Technol. Forecast. Soc. Change 76 (2009) 664–677.
[8] R. Landry, N. Amara, M. Lamari, Does social capital determine
innovation? To what extent? Technol. Forecast. Soc. Change 69 (2002)
681–701.
[9] C.C. Phelps, A longitudinal study of the influence of alliance network
structure and composition on firm exploratory innovation, Acad.
Manag. J. 53 (4) (2010) 890–913.
[10] B.W. Lin, P.C. Li, J.S. Chen, Social capital, capabilities, and entrepre-
neurial strategies: a study of Taiwanese high-tech new ventures,
Technol. Forecast. Soc. Change 73 (2006) 168–181.
[11] C.L. Luk, O.H. Yau, L.Y.M. Sin, A.C.B. Tse, R.P.M. Chow, J.S.Y. Lee, The
effects of social capital and organizational innovativeness in different
institutional contexts, J. Int. Bus. Stud. 39 (2008) 589–612.
[12] C. Steinfield, A. Scupola, C. Lopez-Nicolas, Social capital, ICT use and
company performance: findings from the Medicon Valley Biotech
cluster, Technol. Forecast. Soc. Change 77 (2010) 1156–1166.
[13] C. Haeussler, M. Higgins, Control right and the performance of
strategic alliances, 2009. (Electronic copy available at http://ssrn.
com/abstract=1424073).
[14] P.J. Lane, M. Lubatkin, Relative absorptive capacity and inter-
organizational learning, Strateg. Manag. J. 19 (5) (1998) 461–477.
[15] S.K. Muthusamy, M.A. White, Does power sharing matter? The role of
power and influence in alliance performance, J. Bus. Res. 59 (7)
(2006) 811–819.
[16] R.C. Sampson, R&D alliances and firm performance: the impact of
technological diversity and alliance organization on innovation, Acad.
Manag. J. 50 (2) (2007) 364–386.
[17] W.M. Cohen, D.A. Levinthal, Absorptive capacity: a new perspective
on learning and innovation, Admin. Sci. Q. 35 (1) (1990) 128–152.
[18] M.P. Knudsen, The relative importance of interfirm relationships and
knowledge transfer for new product development success, J. Prod.
Innov. Manag. 24 (2) (2007) 117–138.
[19] R.S. Burt, The network structure of social capital, Res. Organ. Behav. 22
(2000) 345–423.
1269
[20] A. Goerzen, P.W. Beamish, The effect of alliance network diversity on
multinational enterprise performance, Strateg. Manag. J. 26 (4) (2005)
333–354.
[21] P.H. Soh, The role of networking alliances in information acquisition
and its implications for new product performance, J. Bus. Ventur. 18
(2003) 727–744.
[22] C. Galunic, P. Moran, Social capital and productive exchange:
structural and relational embeddedness and managerial performance
link, INSEAD, Research Report, 2000.
[23] R. Hansen, Alliances create opportunity, Ind. Eng. 31 (1999) 50.
[24] T. Rowley, D. Behrens, D. Krackhardt, Redundant governance structures:
an analysis of structural and relational embeddedness in the steel and
semiconductor industries, Strateg. Manag. J. 21 (March 2000) 369–386.
[25] B. Uzzi, Embeddedness in the making of financial capital: how social
relations and networks benefit firms seeking financing, Am. Sociol.
Rev. 64 (1999) 481–505.
[26] J. Shaner, M. Maznevski, The relationship between networks, institu-
tional development, and performance in foreign investments, Strateg.
Manag. J. 32 (2011) 556–568.
[27] P. Bourdieu, Le capital social: notes provisoires, Actes Rech. Sci. Soc. 3
(1980) 2–3.
[28] H.D. Flap, Social capital in the reproduction of inequality, a review,
Comp. Soc. Fam. Health Educ. 20 (1991) 6179–6202.
[29] H.D. Flap, Creation and returns of social capital, Rev. Tocqueville 20
(1999) 5–26.
[30] N.W. Biggart, R.P. Castanias, Collateralized social relations: the social
in economic calculation, Am. J. Econ. Sociol. 60 (2001) 471–500.
[31] R.S. Burt, Structural Holes, Harvard University Press, Boston, 1992.
[32] J.S. Coleman, Social capital in the creation of human capital, Am. J.
Sociol. 94 (1988) 95–120, (supplement).
[33] J.S. Coleman, Foundations of Social Theory, Harvard University Press,
Cambridge, MA, 1990.
[34] S. Wasserman, K. Faust, Social Network Analysis: Methods and
Applications, 1st ed. Cambridge University Press, 1994.
[35] M. Kalwani, N. Narayandas, Long-term manufacturer–supplier relation-
ships: do they pay off for supplier firms? J. Mark. 59 (1995) 1–16.
[36] P. Kotler, G. Armstrong, Marketing Management, Prentice-Hall,
Englewood Cliffs, NJ, 2004.
[37] X. Luo, D.A. Griffith, S.S. Liu, Y.Z. Shi, The effects of customer relationships
and social capital on firm performance: a Chinese business illustration,
J. Int. Mark. 12 (2004) 25–45.
[38] J.C. Baum, T. Calabrese, B.S. Silverman, Don't go it alone: alliance network
composition and startups' performance in Canadian biotechnology,
Strateg. Manag. J. 21 (3) (2000) 267–294.
[39] A. Goerzen, Alliance networks and firm performance: the impact of
repeated partnerships, Strateg. Manag. J. 28 (2007) 487–509.
[40] A. Stirling, A general framework for analyzing diversity in science,
technology and society, J. R. Soc. Interface 4 (2007) 707–719.
[41] D. Hambrick, T. Cho, M.J. Chen, The influence of top management
team heterogeneity on firms' competitive moves, Admin. Sci. Q. 41
(4) (1996) 659–684.
[42] H.R. Greve, Exploration and exploitation in product innovation, Ind.
Corp. Change 16 (5) (2007) 945–975.
[43] J.J.P. Jansen, F.A.J. Van Den Bosch, H.W. Volberda, Exploratory innovation,
exploitative innovation, and performance: effects of organizational
antecedents and environmental moderators, Manag. Sci. 52 (11) (2006)
1661–1674.
[44] K.Y. Williams, C.A. O'Reilly, Demography and diversity in organization: a
review of 40 years of research, Res. Organ. Behav. 20 (1998) 77–140.
[45] L. Fleming, O. Sorenson, Technology as a complex adaptive system:
evidence from patent data, Res. Policy 30 (7) (2001) 1019–1039.
[46] R.R. Nelson, S.G. Winter, An Evolutionary Theory of Economic Change,
Harvard University Press, Cambridge, MA, 1982.
[47] R.M. Henderson, K.B. Clark, Architectural innovation: the reconfigu-
ration of existing product technologies and the failure of established
firms, Admin. Sci. Q. 35 (1990) 9–30.
[48] J. Nahapiet, S. Ghoshal, Social capital, intellectual capital, and the
organizational advantage, Acad. Manag. Rev. 23 (1998) 242–266.
[49] G. Dosi, Sources, procedures, and microeconomic effects of innova-
tion, J. Econ. Lit. 26 (1988) 1120–1171.
[50] D.J. Teece, Technological and organizational factors in the theory of
multinational enterprise, in: M. Casson (Ed.), Growth of International
Business, Allen and Unwin, London, 1983, pp. 51–62.
[51] P. Ingram, Interorganization learning, in: J. Baum (Ed.), Companion to
Organizations, Blackwell, 2002, pp. 642–663.
[52] R.J. Jesen, G. Szulanski, Template use and effectiveness of knowledge
transfer, Manag. Sci. 53 (2007) 1716–1730.
[53] D. Teece, Competition, cooperation, and innovation: organizational
arrangements for regimes of rapid technological progress, J. Econ.
Behav. Organ. 18 (1) (1992) 1–25.
1270 S.-H. Yu / Technological Forecasting & Social Change 80 (2013) 1261–1270
[54] D. Galunic, S. Rondan, Resource recombination in the firm: knowledge
structures and the potential for Schumpeterian innovation, Strateg.
Manag. J. 19 (1998) 1193–1201.
[55] P.G. Audia, J.A. Goncalo, Past success and creativity over time: a study
of investors in the hard disk drive industry, Manag. Sci. 53 (2007)
1–15.
[56] K.D. Simonton, Creativity from a historiometric perspective, in: R.J.
Sternberg (Ed.), Handbook of Creativity, Cambridge University Press,
New York, 1999, pp. 116–133.
[57] A. Hargadon, R.I. Sutton, Technology brokering and innovation in a
product development firm, Admin. Sci. Q. 42 (1997) 716–749.
[58] G. Ahuja, C.M. Lampert, Entrepreneurship in the large corporation: a
longitudinal study of how established firms create breakthrough
inventions, Strateg. Manag. J. 22 (2001) 521–543.
[59] K. Talke, S. Salomo, K. Rost, How top management team diversity
affects innovativeness and performance via the strategic choice to
focus on innovation fields, Res. Policy 39 (7) (2010) 907–918.
[60] J.G. March, Exploration and exploitation in organizational learning,
Organ. Sci. 2 (1991) 71–87.
[61] R. Sampson, Organizational choice in R&D alliances: knowledge-
based and transaction cost perspectives, Manag. Decis. Econ. 25 (2004)
421–436.
[62] L. Fleming, Recombinant uncertainty in technological search, Manag.
Sci. 47 (1) (2001) 117–132.
[63] W.M. Cohen, D.A. Levinthal, Innovation and learning: the two faces of
R&D, Econ. J. 99 (1989) 569–596.
[64] L.A. Grünfeld, Meet me halfway but don't rush: absorptive capacity
and strategic R&D investment revisited, Int. J. Ind. Organ. 21 (2003)
1091–1109.
[65] S.A. Zahra, G. George, Absorptive capacity: a review, reconceptualization,
and extension, Acad. Manag. Rev. 17 (2) (2002) 185–203.
[66] W. Tsai, Knowledge transfer in intraorganizational networks: effects
network position and absorptive capacity on business unit innovation
and performance, Acad. Manag. J. 44 (2001) 996–1004.
[67] D.C. Mowery, J.C. Oxley, Inward technology transfer and competitive-
ness: the role of national innovation systems, Camb. J. Econ. 19 (1)
(1995) 67–93.
[68] S.H. Liao, C.C. Wu, D.C. Hu, K.A. Tsui, Relationships between
knowledge acquisition, absorptive capacity and innovation capability:
an empirical study on Taiwan's financial and manufacturing indus-
tries, J. Inf. Sci. 36 (1) (2010) 19–35.
[69] X. Deng, W.J. Doll, M. Cao, Exploring the absorptive capacity to
innovation/productivity link for individual engineers engaged in IT
enabled work, Inf. Manag. 45 (2) (2008) 75–87.
[70] P.H. Soh, E.B. Roberts, Technology alliances and networks: an external
link to research capability, IEEE Trans. Eng. Manag. 52 (4) (2005)
419–428.
[71] C. Lee, K. Lee, J.M. Pennings, Internal capabilities, external networks,
and performance: a study on technology-based ventures, Strateg.
Manag. J. 22 (2001) 612–640.
[72] M. Zollo, S. Winter, Deliberate learning and the evolution of dynamic
capabilities, Organ. Sci. 13 (2002) 339–352.
[73] W.W. Powell, K.W. Koput, L. Smith-Doerr, Interorganizational collab-
oration and the locus of innovation: networks of learning in
biotechnology, Admin. Sci. Q. 41 (1) (1996) 116–145.
[74] A. Arora, A. Gambardella, Evaluating technological information and
utilizing it: scientific knowledge, technological capability, and external
linkages in biotechnology, J. Econ. Behav. Organ. 24 (1994) 91–114.
[75] K.R. Fabrizio, Absorptive capacity and the search for innovation, Res.
Policy 38 (2) (2009) 255–267.
[76] R. Nelson, The role of knowledge in R&D efficiency, Q. J. Econ. 97 (3)
(1982) 453–470.
[77] L. Fleming, O. Sorenson, Science as a map in technological search,
Strateg. Manag. J. 25 (2004) 909–928.
[78] U. Lichtenthaler, Absorptive capacity, environmental turbulence, and
the complementarity of organizational learning processes, Acad.
Manag. J. 52 (4) (2009) 822–846.
[79] B. Kogut, U. Zander, Knowledge of the firm, combinative capabilities,
and the replication of technology, Organ. Sci. 3 (1992) 383–397.
[80] E.O. Laumann, Community structure as interorganizational linkages,
Annu. Rev. Sociol. 4 (1) (1978) 455–484.
[81] G.T. Milkovich, Compensation systems in high tech companies, in: C.S.
Anderson, A. Kleingartner (Eds.), Human Resource Management in High
Technology Firms, Lexington Books, Lexington, MA, 1987, pp. 103–114.
[82] Y. Caloghirou, I. Kastelli, A. Tsakanikas, Internal capabilities and
external knowledge sources: complements or substitutes for innova-
tive performance? Technovation 24 (2004) 29–39.
[83] K.G. Smith, C.J. Collins, K.D. Clark, Existing knowledge, knowledge
creation capability, and the rate of new product introduction in
high-technology firms, Acad. Manag. J. 40 (2005) 346–357.
[84] A. Zaheer, G.G. Bell, Benefiting from network position: firm capabil-
ities, structural holes, and performance, Strateg. Manag. J. 26 (2005)
809–825.
[85] Z.L. He, P.K. Wong, Exploration vs. exploitation: an empirical test of
the ambidexterity hypothesis, Organ. Sci. 15 (4) (2004) 481–494.
[86] I. Spergel, Interaction between community structure, delinquency,
and social policy in the inner city, in: M. Klein (Ed.), Juvenile Justice
System, Sage, Beverly Hills, 1976, pp. 55–100.
[87] H. Turk, Organizations in Modern Life, Jossey-Bass, San Francisco, 1977.
[88] J. Wincent, S. Anokhin, Does network board capital matter? A study of
innovative performance in strategic SME networks, J. Bus. Res. 63 (3)
(2010) 265–275.
[89] S.B. Hagan, M.B. Green, Corporate knowledge transfer via interlocking
directorates: a network analysis approach, Geoforum 35 (1) (2004)
127–139.
[90] P.R. Haunschild, C.M. Beckman, When do interlocks matter?: alternative
sources of information and interlock influence, Admin. Sci. Q. 43 (1998)
815–844.
[91] M.S. Mizruchi, What do interlocks do? An analysis, critique, and
assessment of research on interlocking directorates, Annu. Rev. Sociol.
22 (1996) 271–298.
[92] J. Scott, Social Network Analysis: A Handbook, Sage, London, 1991.
[93] A. Marin, B. Wellman, Social network analysis: an introduction, in:
Peter Carrington, John Scott (Eds.), Handbook of Social Network
Analysis, SAGE, Los Angeles, CA, 2009.
[94] H. Wang, Is firm-specific innovation associated with greater value
appropriation? The roles of environmental dynamism and techno-
logical diversity, Res. Policy 39 (2010) 141–154.
[95] N. Murovec, I. Prodan, Absorptive capacity, its determinants, and
influence on innovation output: cross-cultural validation of the
structural model, Technovation 29 (2009) 859–872.
[96] S. Thorgren, Designing interorganizational networks for innovation:
an empirical examination of network configuration, formation and
governance, J. Eng. Technol. Manag. 26 (3) (2009) 148–166.
[97] P.W. Beamish, A. Kachra, Number of partners and JV performance,
J. World Bus. 39 (2) (2004) 107–120.
[98] Y. Luo, Partner selection and venturing success: the case of joint ventures
with firms in the People's Republic of China, Organ. Sci. 8 (6) (1997)
648–662.
[99] D.C. Mowery, J.E. Oxley, B.S. Silverman, Strategic alliances and
interfirm knowledge transfer, Strateg. Manag. J. 17 (1996) 77–91.
[100] W. Tsai, S. Ghoshal, Social capital and value creation: the role of
intrafirm networks, Acad. Manag. J. 41 (1998) 464–476.
Sui-Hua Yu is an associate professor in the Department of Accounting at
National Chung Hsing University, Taiwan. She has published in the areas of
business and management accounting research. Her main research interests
are intellectual capital, social network and innovation management.