Strategic Management Journal

Strat. Mgmt. J., 26: 121–140 (2005)

Published online 28 October 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/smj.436

DOES INTERNATIONAL RESEARCH AND

DEVELOPMENT INCREASE PATENT OUTPUT? AN
ANALYSIS OF JAPANESE PHARMACEUTICAL FIRMS
JOAN PENNER-HAHN1 * and J. MYLES SHAVER2
1
School of Business Administration, Wayne State University, Detroit, Michigan,
U.S.A.
2
Carlson School of Management, University of Minnesota, Minneapolis, Minnesota,
U.S.A.

Internationalizing research and development is often advocated as a strategy for fostering the
development of technological capabilities. Although firms conduct international R&D to tap
into knowledge bases that reside in foreign countries, we argue that in order to benefit from
international R&D investments firms must already possess research capabilities in underlying or
complementary technologies. We examine the international R&D expansion activities, research
capabilities, and patent output of 65 Japanese pharmaceutical firms from 1980 to 1991. We find
that firms benefit from international R&D only when they possess existing research capabilities in
the underlying technologies. In addition to refining our understanding of when international R&D
enhances firm innovation, our results integrate asset-seeking and asset-based theories of foreign
direct investment. Internationalizing R&D to tap into foreign knowledge bases is consistent with
asset-seeking theories of foreign direct investment, while the contingent nature by which firms
benefit from international R&D is consistent with asset-based theories of foreign direct investment
and the notion of absorptive capacity. Copyright  2004 John Wiley & Sons, Ltd.

INTRODUCTION nature, which is further utilized by international

The internationalization of research and devel-
opment (R&D) is increasingly recognized as an
important firm strategy in many industries. Firms
internationalize their R&D activities to gain access
to knowledge and capabilities available in other
countries. Thus, internationalizing R&D is an
example of the type of foreign direct investment
(FDI) that several authors have termed asset-
seeking (Kogut, 1991; Wesson, 1999). This is
in contrast to the more traditional view of FDI
as an asset-based activity (Buckley and Casson,
1976; Caves, 1982; Dunning, 1998). Here, a firm
possesses some advantage, generally intangible in
Keywords: international R&D; patents; pharmaceutical
research; absorptive capacity; foreign direct investment
*Correspondence to: Joan Penner-Hahn, School of Business
Administration, Wayne State University, 5201 Cass Avenue,
Detroit, MI 48202-3030, U.S.A. E-mail: jdph@wayne.edu
expansion.
Internationalization of R&D has become impor-
tant in recent years in response to the perceived
increase in technological sophistication through-
out the world and to the existence of specific
expertise in particular countries or regions. Com-
panies reach outside of their domestic bound-
aries to acquire technologies and technological
skills (e.g., Nelson, 1993). Much of the cur-
rent literature advocates internationalizing R&D
in order to acquire new skills and technologies
(e.g., DeMeyer, 1992; Komaran and Goodman,
1993). However, we know little about if, or
when, firms that internationalize their R&D activ-
ities enhance their technological capabilities. Our
contribution stems from the examination of this
issue.
We argue that the establishment of international
R&D activities is often not sufficient for a firm

Copyright  2004 John Wiley & Sons, Ltd. Received 8 May 2000
Final revision received 7 July 2004
122 J. Penner-Hahn and J. M. Shaver
to acquire desired skills and technologies. Just as
entering a country that has a large consumer mar-
ket is no guarantee for successful manufacturing-
based FDI, entering a market that has a unique
knowledge base is no guarantee for successful
R&D FDI. Consistent with the notions of asset-
based FDI as well as that of absorptive capacity
(Cohen and Levinthal, 1990), we propose that the
acquisition of skills and technologies is contingent
on the investing firm possessing underlying tech-
nological capabilities that foster the acquisition of
skills and technology.
There are two types of capabilities that we
believe aid firms in successfully acquiring new
skills and technologies through international R&D.
First, firms with strong existing research activities
in the underlying technologies will be better able
to absorb and put to use the skills and technolo-
gies they are exposed to in their international R&D
activities. Second, firms that possess research skills
in technologies complementary to the skills and
technologies they are exposed to in their interna-
tional R&D activities will be better able to adapt
the new technologies to their use. By comple-
mentary technologies, we mean technologies that
are necessary for the commercial development of
another technology or science base as in the sense
proposed by Teece (1986).
To test our arguments, we investigate the inter-
national R&D activities and pharmaceutical patent
output of firms in the Japanese pharmaceutical
industry for the years 1980–91. We measure firms’
research capabilities in the underlying technolo-
gies by previous pharmaceutical patents. Because
most of the international R&D activities estab-
lished by these firms were focused on biotech-
nology, we measure firms’ research capabilities
in complementary technologies as previous fer-
mentation patents. As we discuss shortly, fer-
mentation technology is complementary to many
biotechnology applications because it is the pro-
cess by which many biotechnology products are
produced.
We find that international R&D increases phar-
maceutical patent output. However, this is only for
firms with existing capabilities in the underlying
technologies. Firms without existing capabilities
in the underlying technologies do not benefit from
international R&D. We do not find evidence that
firms with capabilities in complementary technolo-
gies benefit more from international R&D com-
pared to firms lacking these capabilities. Moreover,
Copyright  2004 John Wiley & Sons, Ltd.
we find that firms with greater existing capabilities
in the underlying technologies have higher future
pharmaceutical patent output. We also find in many
specifications that firms with greater capabilities
in complementary technologies have higher future
pharmaceutical patent output.
Our findings have important implications for
both managers and researchers. We shed light on
what conditions must hold for international R&D
to facilitate firm innovation. Moreover, we show
that successful asset-seeking FDI, such as inter-
nationalizing R&D, occurs when firms possess
certain capabilities. This insight fosters an inte-
gration of asset-seeking and asset-based theories
of FDI.
In the next section, we discuss previous research
that addresses the internationalization of R&D and
present our hypotheses of when international R&D
increases firm innovation. The third section pro-
vides background information on the industry con-
text of our study. In the fourth section, we describe
our methodology and in the fifth section we dis-
cuss the findings of our study. We conclude with
implications.
BACKGROUND AND HYPOTHESES
International R&D: Background
Internationalization of R&D has generated con-
siderable interest in academic and business cir-
cles. The primary question addressed by the exist-
ing literature is ‘why do firms undertake interna-
tional R&D?’ Previous authors have pointed to the
importance of accessing skills and capabilities that
reside overseas as the underlying motivation for
internationalizing R&D (e.g., Buckley and Casson,
1976; De Meyer and Mizushima, 1989; Komaran
and Goodman, 1993). This type of activity has
been called asset-seeking FDI by some (Wes-
son, 1999) and capability-seeking by others (e.g.,
Anand and Delios, 2002). This rationale for inter-
nationalizing contrasts with the asset-based ratio-
nales for internationalization that were proposed
by many FDI theorists (Caves 1971, 1982; Buck-
ley and Casson, 1976). In asset-based FDI, the
firm possesses some advantage, generally intan-
gible in nature, that is further utilized by interna-
tional expansion.
Asset-seeking FDI has become more prevalent
with the realization that the skills and capabilities
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
being sought by firms might be uniquely avail-
able in a foreign location. Although technological
capability is diffusing and equilibrating through-
out the world, there exist pockets of expertise that
develop due to peculiarities of specific ‘national
innovation systems’ (Nelson, 1993). Foreign firms
may find it necessary to establish R&D activities
in these locations to tap into sources of technol-
ogy that diffuse slowly across national boundaries
(Kogut, 1991). Technological knowledge is often
tacit and requires frequent interaction for trans-
fer (Kogut and Zander, 1992). Thus, proximity
is necessary for acquiring such localized knowl-
edge.
Indeed, Kuemmerle (1997) found that 45 percent
of the foreign R&D laboratories in his sample
were established for ‘home-base-augmenting’ pur-
poses. That is, the laboratories were established
for the purpose of tapping knowledge from com-
petitors and institutions in other lands. Empiri-
cal evidence also documents that foreign investors
are able to tap into foreign knowledge bases and
supports these arguments. For example, Almeida
(1996) shows that foreign semiconductor firms
tap into local knowledge in the United States.
Cantwell (1995) shows that technological leaders
are becoming increasingly geographically special-
ized with respect to their technological activity.
Jaffe et al. (1993) find that inventors are more
likely to cite patents of other local inventors. Simi-
larly, firms might conduct foreign R&D in order to
be near the users of the technology. Foreign cus-
tomers are often the most sophisticated users of a
technology and therefore important contributors to
future technology development. For example, Dow
Chemical established a furniture R&D lab in Italy
because Italian furniture makers are the most inno-
vative in their demands for adhesives, finishes, and
other materials.
Although there has been a great deal of discus-
sion of the role of asset-seeking FDI in expand-
ing the capabilities of firms, there have been few
empirical studies of the results of this type of FDI.
This study attempts to rectify this situation. In
order to assess the impact of international R&D
on innovation it is important to first recognize that
firm capability is also an important determinant of
firm innovation. Therefore, we turn to the existing
literature to assess how capabilities, specifically
underlying and complementary technology capa-
bilities, affect firms innovation. With this building
block we then hypothesize how such capabilities,
Copyright  2004 John Wiley & Sons, Ltd.
123
when combined with international R&D, would
result in even greater innovation.
Research capabilities and patents
We expect that firms with greater research capa-
bilities in underlying technologies will exhibit
greater innovative output. As Scherer and Ross
(1990) observed, ‘Technical innovations do not fall
like manna from heaven. They require effort—the
creative labor of invention, development, testing
and introduction into the stream of economic life.’
As Hall, Griliches, and Hausman (1986: 265) state:
‘The annual research and development expendi-
tures of a firm are considered to be investments
which add to the firm’s stock of knowledge.’ Firms
undertake R&D activities, in large part, to create
innovations that will ultimately provide new prod-
ucts and therefore profits. While most innovations
may be serendipitous, firms may have a compar-
ative advantage in generating inventions because
they ‘are more likely to put together the criti-
cal combination of a fertile mind, a challenging
problem, and the will to solve it’ (Scherer and
Ross, 1990).
R&D is one of the firm-specific assets that is
the result of a cumulative pattern of activity (Dier-
ickx and Cool, 1989). In addition, Dierickx and
Cool point out the asset mass efficiencies nature of
R&D; that is, that increments to an existing stock
are facilitated by possessing high levels of that
stock. The organizational learning literature has
also emphasized the experiential nature of inno-
vation and that an organization’s past can influ-
ence its future capabilities for renewal and change
(Mezias and Glynn, 1993).
Economists have long explored the outcome of
corporate R&D activities through empirical studies
(Mansfield, 1962, 1965; Terleckyj, 1980). Mans-
field (1980) found that firms with large amounts
of basic R&D had relatively high rates of pro-
ductivity increase, although he finds that long-
term research rather than basic research specif-
ically may explain most of this growth. Subse-
quently, Griliches (1986) demonstrated that R&D
contributed positively to productivity growth and
that basic research is an important part of pro-
ductivity. Pakes (1985) and Griliches (1980) have
results which suggest that patents are an output of
R&D rather than an input.
In addition, firms that are more skilled in partic-
ular research areas will be more likely to discover
Strat. Mgmt. J., 26: 121–140 (2005)
124 J. Penner-Hahn and J. M. Shaver
new innovations in those same areas. In a study of
the pharmaceutical industry, Henderson and Cock-
burn (1996) show that a large proportion of the
variance in research productivity is due to firm
fixed effects. Moreover, Henderson and Cockburn
(1994) show that part of this fixed effect is related
to more fine-grained measures of research compe-
tence. They also find that accumulated knowledge,
measured by the accumulated stock of patents,
accounts for 20 percent of the variance in future
patenting within particular programs. Accordingly,
we hypothesize greater pharmaceutical research
capabilities will result in greater pharmaceutical
patent output. Although, as we have discussed pre-
viously, this is not a novel hypothesis, it is neces-
sary groundwork for the subsequent hypotheses.
Hypothesis 1: Firms that have greater underly-
ing technological capabilities will subsequently
generate greater innovative output.
Following from the same reasoning, we expect
firms that possess complementary research capa-
bilities to have greater innovation. Complementary
capabilities are forms of know-how or technologies
that might be used in conjunction with the under-
lying technologies to create new products (Teece,
1986). Generally, complementary capabilities are
downstream from the innovative technology; that
is, they assist in commercializing the new. Here
we are hypothesizing that there might be groups
of technologies which work together to create
new products. Firms that possess complementary
research capabilities will be able to use these capa-
bilities in combination with other research skills
to create new innovations. This is similar to the
argument that asset stocks can be interconnected
(Dierickx and Cool, 1989). That is, accumulating
increments in one area may depend on the level
of another area. For example, in a study of R&D
in the petroleum industry, Helfat (1997) finds that
greater amounts of complementary technological
knowledge in oil refining influences the firms’
efforts to alter their stock of knowledge in syn-
thetic fuels.
Therefore, we expect that firms with greater
complementary research capabilities will exhibit
greater innovative output. We hypothesize:
Hypothesis 2: Firms that possess greater com-
plementary research capabilities will subseque-
ntly generate greater innovative output.
Copyright  2004 John Wiley & Sons, Ltd.
International R&D and patents
Turning to the more novel hypotheses in our study,
we expect that international R&D will increase
firm innovation. However, we expect that this
effect will only be pronounced for firms with exist-
ing capabilities in underlying or complementary
technologies. The following paragraphs detail this
argument.
As we stated previously, the motivation of firms
to undertake international R&D is to access knowl-
edge that is geographically bound in a foreign loca-
tion. Because there is tacitness (Polyani, 1964) and
stickiness (Szulanski, 1996) associated with the
absorption of scientific knowledge, firms cannot
absorb scientific knowledge from afar by read-
ing scientific papers, attending conferences and
citing patents. To the extent that firms can tap
into this knowledge and such knowledge is ben-
eficial in fostering innovation, internationalizing
R&D should enhance firms’ innovation.
Nevertheless, despite the potential benefits of
international R&D the difficulties of international
R&D have long been recognized (Fayerweather,
1969). Such difficulties include the costs associated
with communicating across distance and culture.
Moreover, complications stemming from the tac-
itness and transferability of knowledge are often
exacerbated in an international context (Teece,
1976).
Because both potential benefits and substantial
difficulties exist in effectively conducting interna-
tional R&D, we expect that the mere act of inter-
nationalizing R&D will often not be sufficient to
achieve increased innovative outcomes. We believe
that there are necessary preconditions for suc-
cessfully undertaking international R&D activities.
Specifically, we expect that, in order to be suc-
cessful, firms must possess (a) existing research
capabilities that are akin to the skills they seek in
foreign nations, or (b) research capabilities that are
complementary to the skills they seek in foreign
nations.
Several arguments in the R&D literature ground
our expectation that the possession of research
capabilities positively affects the effectiveness of
international R&D. Cohen and Levinthal note that
by developing an absorptive capacity in a particu-
lar area ‘a firm may more readily accumulate what
additional knowledge it needs . . . to exploit any
critical external knowledge that may become avail-
able’(Cohen and Levinthal, 1990: 136). They argue
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
that firms with increased absorptive capacity will
tend to be more proactive, exploiting opportunities
present in the environment, independent of current
performance. Moreover, Cohen and Levinthal sug-
gest that absorptive capacity is largely a function
of the firm’s level of prior knowledge (i.e., their
existing research capabilities). Namely, firms with
existing research capabilities will have greater abil-
ity to absorb external knowledge than firms lacking
such capabilities. Consistent with these arguments,
Pisano (1990) finds that firms that have established
a capability in a particular research skill are better
able to absorb the information gained from external
research activity in that area.
Moreover, because greater absorptive capacity
enables firms to better interpret and assess knowl-
edge from outside of the company, we expect
these abilities also to allow firms to then bet-
ter communicate and transfer knowledge that they
are able to interpret and assess within the com-
pany. Therefore, firms with research capabilities
are more likely able to transfer information that
they access within the company compared to com-
panies lacking such capabilities.
Additionally, increased research capabilities will
often improve a firm’s attractiveness to research
scientists or R&D partners in the host country. This
is because a foreign firm with research capabilities
brings more to a venture and enhances the like-
lihood that individual scientists or partners inno-
vate, compared to a firm that lacks these capabil-
ities. For example, Ahuja (2000) shows that firms
with greater technical capital are able to enter into
greater numbers of linkages. The greater attractive-
ness to local research scientists and firms increases
the degree to which international R&D activities
are able to tap into the local knowledge base.
As a result, we expect that firms with greater
research capabilities ex ante should experience
greater benefits and mitigate some of the compli-
cations from international R&D compared to firms
lacking these capabilities. Therefore, we expect
that the international R&D activities of firms with
existing research capabilities will be more success-
ful than those lacking such capabilities.
Hypothesis 3: Firms that internationalize their
R&D and possess underlying research capabil-
ities will subsequently generate greater innova-
tive output.
Copyright  2004 John Wiley & Sons, Ltd.
125
Similarly, firms that possess specific complemen-
tary technical skills and internationalize will be
more effective at absorbing the knowledge that is
available in the foreign location. The additional
insight granted by the in-depth knowledge of an
important related field will allow those firms to
increase their absorptive capacity and better uti-
lize the information created in the new location.
Firms with relevant complementary technical skills
will be better able to overcome the difficulties of
tacitness and transferability of knowledge experi-
enced by firms internationalizing their activities.
This is a more related application of the arguments
outlined above for how underlying research capa-
bilities and international R&D affect innovation.
We thus hypothesize:
Hypothesis 4: Firms that internationalize their
R&D and possess complementary research
capabilities will subsequently generate greater
innovative output.
Although grounded in the R&D literature, these
hypotheses are consistent with previous research of
international expansion success. Mitchell, Shaver,
and Yeung, (1992) and Morck and Yeung (1992)
find that international expansion activities in man-
ufacturing contexts tend to be successful when the
expanding firms possess existing skills or capa-
bilities. Moreover, Shaver, Mitchell, and Yeung
(1997) show that firms must possess a threshold
level of knowledge in order to learn about activi-
ties in a foreign market.
INDUSTRY CONTEXT
The Japanese pharmaceutical industry, which we
employ as the empirical setting, provides an excel-
lent context for testing hypotheses related to the
internationalization of R&D for four reasons. First,
the pharmaceutical industry is a high-technology
industry that depends heavily on the outcome of
R&D activities. Japanese pharmaceutical firms, on
average, spent 12 percent of sales on R&D dur-
ing the study period (JPMA, 1992). In addition,
the Japanese domestic pharmaceutical market is
the second largest in the world after the United
States.
Second, the Japanese pharmaceutical industry is
sufficiently populated to provide a viable sample
Strat. Mgmt. J., 26: 121–140 (2005)
126 J. Penner-Hahn and J. M. Shaver
that offers the research design advantages of focus-
ing on one industry and firms from one nation. The
top four domestic firms accounted for 25 percent
of sales in the Japanese market in 1991. The
top 30 domestic firms (by sales) accounted for
70 percent of sales in 1980 and 85 percent of sales
in 1991. Moreover, the introduction of biotechnol-
ogy resulted in a number of firms entering the
industry. Because many of these entrants have
international R&D activities, an advantage of our
sample is that there exists substantial variance in
the research skills of firms that do and do not con-
duct international R&D.
Third, competition in the Japanese pharmaceuti-
cal market increased due to changes in the regula-
tory environment. Government restrictions on the
entry of foreign firms were lifted in the mid-1970s.
In addition, the government introduced price con-
trols for pharmaceuticals starting in the 1980s in
order to decrease national expenditures on health
(Reich, 1990). The price controls were designed to
encourage pharmaceutical innovation by allowing
newer drugs to receive higher prices. These two
factors started to place pressure on pharmaceutical
firms’ profits. In turn, the need for innovative prod-
ucts in order to obtain higher prices increased the
motivation for Japanese firms to undertake R&D.
Finally, there has been a change in the technol-
ogy necessary for pharmaceutical discovery that
requires a response from firms involved in the
industry: the development of biotechnology. Sci-
entists working in Great Britain provided the intel-
lectual development of molecular biology (Nelson,
1993). Much of the subsequent work in biotech-
nology has been performed in the United States
(National Research Council, 1992). Both countries
continue to be leaders in the field of biotechnology.
Thus the expertise in biotechnology has developed
in the United States and Western Europe. Cantwell
(1992) documents that Japan has a revealed tech-
nological disadvantage in pharmaceutical research
when compared to the United States, the United
Kingdom, Germany, and Switzerland, which he
identifies as the powerhouses of pharmaceutical
research. Therefore, the geographic concentration
of advantages in pharmaceutical and biotechnol-
ogy research outside of Japan provides an obstacle
for Japanese wishing to obtain competence in new
pharmaceutical creation. It is for this reason that
many Japanese companies have looked outside of
Japan for pharmaceutical R&D.
Copyright  2004 John Wiley & Sons, Ltd.
METHOD
Sample
Our sample includes 65 participants in the Japanese
ethical pharmaceutical industry over the period
1980–91. The sample includes the 30 largest
Japanese pharmaceutical firms by sales and 35
entrants to the Japanese pharmaceutical industry
over that time period. We excluded the Japanese
subsidiaries of European and U.S. pharmaceutical
firms from the sample.
The entrants were selected on the basis of indus-
try analyses by Yano Keizai Research Institute,
Toyo Keizai, the Pharmaceutical Industry Forum
(PIF) that indicated that these firms had either
initiated pharmaceutical research or had pharma-
ceutical products for sale. The entrants included
textile, food, beverage, chemical, and steel firms.
These firms were not entrants in the sense of being
start-ups but were established firms that diversified
into the pharmaceutical industry. As is common in
the Japanese context, the firms entered the phar-
maceutical industry by establishing pharmaceutical
subsidiaries. Most of these firms started partici-
pation in the pharmaceutical industry during the
1970s, yet had not introduced products until the
latter part of our sample period.
The year 1980 is an appropriate starting point
for considering the international activities of these
firms because this is the beginning of a period
of real change for the Japanese pharmaceutical
industry. The development of biotechnology, the
changes in the government’s payment policy, and
increased foreign competition were all affecting
the Japanese pharmaceutical industry during this
period. Moreover, we found no evidence of inter-
national R&D prior to this year.
Joan Penner-Hahn initially developed our data
set in her dissertation (1995). Although the data
that we employ in this study come from publicly
available sources, Penner-Hahn and colleagues
also interviewed Japanese pharmaceutical execu-
tives in 15 companies (the interview team consis-
tent of two to four members who reviewed the
interview notes, reaching consensus on their con-
tent). In the context of this study, the primary
purpose of the interviews was to verify the accu-
racy of the publicly available data regarding phar-
maceutical and international R&D activities. The
interviews also provided an opportunity to develop
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
a more complete understanding of the issues con-
fronting the pharmaceutical firms, which aids in
formulating our tests and assessing the results.
Approach
To test the hypotheses, we examine how interna-
tional R&D activity affects firm-level innovative
output. Before we define the variables, we high-
light the merits of our approach.
We focus on the firm’s overall innovative output
because the impact of international R&D activi-
ties is often not isolated within a particular foreign
R&D lab or location. For example, from the inter-
views many of the research managers indicated
that they planned to send domestic R&D person-
nel overseas and rotate them back into domestic
R&D labs. The goal was to diffuse the knowledge
obtained abroad throughout a firm’s research activ-
ities. Similarly, several managers mentioned estab-
lishing dual tracks of research: one at the foreign
facility and one at a domestic facility. Some of the
firms also held formal meetings of scientific staffs
in order to share research findings. For example,
one firm alternated scientific meeting sites between
its Japanese lab and its foreign lab in order to better
acquaint the scientists with the activities occurring
at each site.
In these cases, innovations and other benefits
from international R&D are not necessarily mani-
fest or captured in the foreign facility but at other
research facilities throughout the company. If we
were only to focus on research output that was
deemed to have occurred in a foreign facility, then
our measure would miss important ways in which
international R&D efforts were being managed to
increase innovation. This is why we focus on firm-
level innovative output rather than the output of the
specific foreign R&D activities.1
In addition, by measuring overall firm innova-
tive output, we are able to include in our com-
parisons firms that have and do not have inter-
national operations. We need this variance in the
sample to properly test Hypotheses 3 and 4. Should
we restrict the sample to firms with international
R&D activities and find that underlying research
capability or complementary research capabilities
1
Focusing on firm-level innovative output raises the concern
of controlling for other firm characteristics that would affect
the dependent variable. We discuss our efforts to do so in the
following section.
Copyright  2004 John Wiley & Sons, Ltd.
127
increase the innovative output of firms, we can-
not rule out that underlying research capability or
complementary research capabilities per se affect
patent output. Therefore, we must examine firms
with and without international R&D activities to
effectively test our hypotheses.
Variable definitions
Dependent variable
The dependent variable that we employ to mea-
sure the construct of innovative output is the count
of U.S. drug patents granted to a firm in a given
year, which we label DRUG PATENT COUNT.
This is appropriate because our focus is pharma-
ceutical firms that we expect to be innovative in
drug related areas. We identify drug patents as any
patent by the firm that is in the patent classes 424
or 514. Examiners at the U.S. Patent and Trade-
mark Office (USPTO) assign a patent class and
subclass based on the information provided in the
patent application. Patent classes reflect the tech-
nological and functional principals of an applica-
tion rather than products or industries. Classes 424
and 514 are identified by the USPTO as ‘Drug,
Bio-affecting and Body Treating Compositions.’
This patent class includes any molecules that are
perceived to have drug applications. It is important
to note that every patent has multiple class assign-
ments and that these two categories are not limited
to a particular formulation of drug. Patent class
data have been previously used to assess firms’
technical skills. For example, Jaffe (1986) uses
patent class data to characterize the technological
positions of firms. Moreover, Griliches suggests
that ‘it is possible to use a firm’s distribution of
patenting by field to infer its position in “techno-
logical space” . . .’ (Griliches, 1990: 1702).
Patent counts have been used to measure inno-
vative output in a number of previous studies
(e.g., Bound et al., 1984; Henderson and Cock-
burn, 1996). However, there are four issues with
respect to our use of patent count that warrant
comment. First, the propensity to patent is not
necessarily constant across firms, especially when
looking across nations. However, as Cantwell
(1989) notes, the variations between firm patenting
levels seem to reflect systematic industry-specific
and country-specific differences. Because we focus
on Japanese pharmaceutical firms, our design holds
constant industry and nationality.
Strat. Mgmt. J., 26: 121–140 (2005)
128 J. Penner-Hahn and J. M. Shaver
Second, not all patented inventions result in
innovations, as defined as new products. Once
again, the design choice of the pharmaceutical
industry mitigates this concern. In a study of phar-
maceutical firms, Comanor and Scherer (1969)
found a significant correlation between the num-
ber of patents and new products introduced. Hen-
derson and Cockburn (1996) report preliminary
results that indicate that Investigational New Drug
Applications are highly correlated with ‘important’
patents. Moreover, looking across many indus-
tries, Acs and Audretsch (1989) also found a high
degree of similarity between patents and innova-
tive activity.
Third, not all innovations are patented. Again,
our choice of industry mitigates this concern. The
pharmaceutical industry provides a setting where
patents offer reasonable protection of proprietary
knowledge. In his survey of companies about the
effectiveness of patents in protecting rights, Levin
(1986) reported that only the pharmaceutical and
chemical industries found patents to be an effec-
tive means of protecting competitive advantages
of new technology. Moreover, the pharmaceutical
industry is one in which patents play an extremely
important role in protecting the intellectual capi-
tal of firms (Henderson and Cockburn, 1996). In
this industry, more so than in many others, firms
are likely to apply for patents when there is a
development that may lead to a future drug. Phar-
maceutical firm executives, with whom we con-
versed, noted that they tend to apply for patents
within 2–3 months of discovery of a promising
compound.
Fourth, we use U.S. patents to measure the
technical skills of Japanese firms. Foreign firms,
with and without operations in the United States,
patent in the U.S. system. Japanese firms had the
largest percentage of U.S. patents granted, after
U.S. firms, in 1986 (Wineberg, 1988). Because
the United States represents the largest single mar-
ket in the world for pharmaceutical products, for-
eign firms often seek to protect their intellec-
tual property rights with U.S. patents. In addi-
tion, the Japanese system is very slow to grant
patents, 5–7 years as opposed to 2–3 in the United
States, which might provide even more incentive
for Japanese firms to patent their drug discov-
eries in the United States (Dunphy, 1988). One
firm in our sample had over 450 U.S. patents
granted during the period studied. Moreover, pre-
vious research has utilized U.S. patent counts to
Copyright  2004 John Wiley & Sons, Ltd.
measure revealed technological advantage across
countries (Cantwell, 1989).
We gathered patent data from the CASSIS
Database of the Patent and Trademark Office of the
U.S. Department of Commerce. We downloaded
patent information for each company in the sample
and used a FORTRAN program to extract patent
class information.
Independent variables
The variable INTERNATIONAL records whether
or not a firm had an international R&D activity
within a given year. It takes the value of one for
firms that engaged in international R&D activity,
zero otherwise.2
These data were gathered from an extensive
search of multiple secondary sources. We included
the following types of activities in our defini-
tion of international R&D: sponsored, collabora-
tive, and controlled (Penner-Hahn, 1998). Spon-
sored research activities are those in which a firm
funds research projects focused on the discovery
of a specific product or phenomenon at a for-
eign university or R&D lab. Sponsored research
involves the dissemination of knowledge back to
the sponsoring firm and, in some cases, participa-
tion by a limited number of firm personnel in the
foreign laboratories as researchers. Moreover, the
sponsoring firm often receives rights to products
developed as a result of the research. Collabora-
tive research projects involve participation of a
firm’s employees in the foreign research activity,
either through relocation and rotation of teams of
employees to the foreign site or through under-
taking dual tracks of research at home and abroad.
Collaborative research projects are commonly joint
ventures or alliances. Controlled research activities
are those for which the firm establishes ownership
either through the acquisition of a foreign facil-
ity or the establishment of a new facility. In these
instances, the firm employs foreign researchers to
work in these facilities and often transfers scien-
tists from its domestic operation to the foreign
2
There might exist a time lag from the establishment interna-
tional R&D activities until these activities affect firm innovative
output. Our results do not materially change if we redefine
INTERNATIONAL so that it takes the value 1 only after firms
have international operations for at least 1 year, 2 years, or
3 years. Because we cannot justify the use of one of these lag
structures over the other, we choose to present the results based
on whether a firm has international R&D activities in a given
year.
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
site. The INTERNATIONAL variable makes no
distinction between the type of international activ-
ity undertaken, merely that at least one of three
existed during the particular year. Although a more
refined measure of the intensity of the international
activity, such as the amount of expenditure, would
be preferable, it is simply not available. Firms are
generally loath to release such level of detail about
their activities.
We excluded two types of activities from our
definition of international R&D. We excluded
patent licensing because the primary reason for
such licensing is to sell a specific product in a
foreign market, rather than to undertake R&D
activities with the licensor. We also excluded clin-
ical development facilities from our definition.
Although a few firms established clinical devel-
opment facilities during the study period, these
activities were directed toward market adaptation
and acceptance of existing research rather than the
discovery and development of science. Therefore,
we did not code either of these activities as inter-
national R&D for the purposes of this study.
We used interviews with 15 of the sample firms
to verify the accuracy of the secondary data.
The interviews were conducted in Japan during
May and June of 1993. In most cases, the com-
pany employees interviewed were members of the
strategy staff or the equivalent, generally a vice-
presidential level employee or above. Prior to the
interviews, the firms were sent summaries of their
foreign R&D activity based on the secondary data
collection. The interviewers verified the accuracy
of the activity reports and obtained information
on any missing activities. Only in very isolated
instances were any errors found. Moreover, such
‘errors’ were often the companies updating their
current international R&D strategies. Given this
confirmation of secondary sources’ accuracy, we
relied on the public sources for the remaining sam-
ple firms.
Of the 65 firms in our sample, 36 undertook
international R&D activities at some period dur-
ing the sampling frame. Consistent with our priors
and our discussions with the Japanese pharmaceu-
tical executives, international R&D appeared to
focus on sourcing knowledge that resided within
countries with revealed technological advantage
in pharmaceutical research. Of the 36 firms, 31
had investments in the United States; the remain-
ing five firms had investments in Germany (two),
the United Kingdom (one), Canada (one), and
Copyright  2004 John Wiley & Sons, Ltd.
129
Israel (one). As previously mentioned, Cantwell
(1992) shows that the United States, Germany, and
the United Kingdom have a revealed technological
advantage in pharmaceuticals.
We measure firms’ underlying research capa-
bilities by the count of drug patents in the 3-
year period prior to a focal year (i.e., the sum of
the three lagged years of the dependent variable).
We label this variable DRUG PATENT STOCK.
Our aim in selecting the 3-year period is to cap-
ture recent research skills. Moreover, by focus-
ing on 3 years, we lessen the chance that a ran-
dom, extreme event in 1 year disproportionately
influences our measure of pharmaceutical research
capabilities. Nevertheless, because the choice of a
3-year period is arbitrary, we re-examined our tests
by redefining this variable with periods as short as
1 year and as long as 5 years. Our results were
not sensitive to the length of period over which
we measured this variable.
An alternative way to measure such capabilities
is to measure R&D spending or R&D intensity.
We prefer DRUG PATENT STOCK because it is
a focused measure of pharmaceutical capabilities.
Due to data constraints, measures of R&D spend-
ing or R&D intensity are available only at the firm
level, not the pharmaceutical business level.3
We measure firms’ complementary research
capabilities by assessing the firms’ fermentation
research skills. Through discussions with execu-
tives in the pharmaceutical industry, we identified
fermentation research skills as a complementary
technological area for pharmaceutical research.4
Understanding of fermentation processes obtained
through research would assist firms in the cre-
ation of biotechnology based molecules and drugs,
necessary in today’s competitive pharmaceutical
3
We were able to compare the correlation between DRUG
PATENT STOCK and R&D for 2 years of our sample and found
them correlated (r = 0.61).
4
Fermentation processes are fundamental to the production of
many biotechnology-based drugs. As Grace (1997: 78) writes,
‘[t]o date, microorganisms (i.e., fermentation) are still the main
tools bioengineers use to turn out pharmaceutical products.’ Esti-
mates are that almost 50 percent of biotechnological products
will be produced by fermentation processes and further, that half
of all drugs will be biotechnology based in 10–20 years (Fox,
2001). Because much of the international R&D in our sample
was focused on biotechnology, ex ante we expect fermentation
to be a complementary skill that would influence firms’ success
in their drug research. An additional rationale for choosing fer-
mentation as the complementary skill is that many of the firms in
this industry had particular expertise in fermentation processes
through the production of antibiotics, yogurt, soy sauce, and even
beer.
Strat. Mgmt. J., 26: 121–140 (2005)
130 J. Penner-Hahn and J. M. Shaver

market. Fermentation research skills were mea-
sured by the count of firm patents in class 435 of
the USPTO classification scheme, for the 3-year
period prior to a focal year. We label this vari-
able FERMENTATION PATENT STOCK. Patent
class 435, which is the Molecular Biology and Bio-
chemistry class, is of particular interest because it
includes all patents related to fermentation. How-
ever, we should note that this patent class con-
tains more than just fermentation patents, which
might or might not be useful in drug discovery.
Once again, because the length of time period over
which we measure this count is arbitrary, we also
examined alternative period lengths. We examined
periods from 1 to 5 years and found no material
difference in the results that we present.
Controls
Other firm characteristics and strategies influence
pharmaceutical patent output such as international
experience, culture of innovativeness, size, or orga-
nization structure. To control for these effects we
take advantage of the panel data design and in
many specifications include firm fixed effects or
random effects. The advantage of employing firm
fixed effects or random effects is that they control
for many firm characteristics and strategies that
affect pharmaceutical patent output, which might
or might not be observable and measurable to the
researcher.
In specifications where we do not use fixed
effects, we can examine differences in patent out-
put by firms that are entrants to the pharmaceutical
industry. To do so we create a dummy variable
labeled ENTRANT that takes the value one if a
firm is an entrant to the Japanese pharmaceutical
industry, zero otherwise.
In order to control for firm size, we gathered
data on firm sales. Sales data that were restricted to
pharmaceutical sales were not available; therefore,
these data capture total firm sales. The variable
SALES measures firm sales in trillions of Japanese
yen, inflation adjusted to 1980. We were unable to
collect reliable sales data on six firms over the
entire panel. In addition, in four cases companies
changed their year-end for reporting during the
sample period. Here, we dropped the transition
year from the data and included the comparable
years before and after the change. Finally, we
were not able to collect reliable sales data for six
other firm-year observations (pertaining to three
firms). Therefore, our usable sample when we
include the variable SALES drops by 64 firm-year
observations to 521.
Table 1 reports descriptive statistics for the vari-
ables. We have a balanced panel of 65 firms over
12 years (1980–91)—with the exception of miss-
ing data for SALES. Because the patent stocks are
calculated as the cumulative 3-year lag of the drug
and fermentation patent counts, the usable sample
for the statistical analyses is a balanced panel of
9 years (1983–91).
ESTIMATION AND RESULTS
The empirical models that we estimate, in general,
take the following form:
DRUG PATENT COUNT
= f (INTERNATIONAL,
DRUG PATENT STOCK, FERMENTATION
PATENT STOCK, CONTROLS)

Table 1. Descriptive statistics (n = 585)a

Variable Correlation
Mean S.D. Min. Max. 1. 2. 3. 4. 5.

1. DRUG PATENT COUNTt 3.23 5.23 0.00 35.0 1

2. INTERNATIONALt 0.33 0.47 0.00 1.0 0.37 1
3. DRUG PATENT STOCK(t−1,t−3) 8.34 13.55 0.00 89.0 0.89 0.35 1
4. FERMENTATION PATENT STOCK(t−1,t−3) 2.94 5.15 0.00 30.0 0.38 0.19 0.39 1
5. ENTRANT 0.54 0.50 0.00 1.0 −0.41 −0.18 −0.42 −0.06 1
6. SALESt b (yen, trillions) 2.24 2.41 0.04 13.8 −0.11 0.10 −0.12 0.04 0.44

a
The usable base sample is a 9-year balanced panel of 65 firms.
b
The usable sample size for SALES is 521.

Copyright  2004 John Wiley & Sons, Ltd. Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents? 131

Because of the limited nature of the dependent
variable (it is a count), we employ non-linear
estimators including Poisson and negative bino-
mial regression models. The use of such meth-
ods is common in studies of patent output (e.g.,
Graves and Langowitz, 1993; Hausman, Hall,
and Griliches, 1984; Henderson and Cockburn,
1996). Poisson regression models are formulated
on the assumption that the dependent variable
is drawn from a Poisson distribution, which has
equal mean and variance. Should this assump-
tion not hold (i.e., the data are overdispersed),
the likelihood function from which the estimates
are derived is misspecified. Under this condi-
tion, a common approach is to employ negative-
binomial regression analysis, which allows the
mean and variance of the Poisson process to
vary by introducing an individual unobserved
disturbance. In our analysis, we employ both
Poisson and negative binomial regression mod-
els.
We first examine Hypotheses 1 and 2 in Table 2.
The specification in Model 1 presents the results
of the Poisson regression model and the speci-
fication in Model 2 presents the results of the
negative binomial regression model. Because we
have nine observations for each firm, we recognize
that observations are not independent within firms.
To account for this, we utilize the Huber–White
sandwich estimator for the variance and allow for
the possibility that observations are not indepen-
dent within firms but independent across firms (see
Cameron and Trivedi, 1998).
The test of overdispersion in these data is
significant, as shown in Model 2. Therefore, we
focus our discussion on the results from the neg-
ative binomial estimates. Nevertheless, the pattern
of results is similar across the negative binominal

Table 2. Analyses of firm drug patents Dependent variable: DRUG PATENT COUNTt ; (t-values in parentheses)
[marginal effects in brackets]

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Poisson Negative Negative Negative Negative Negative
binomial binomial binomial binomial with binomial with
with random with random random effects fixed effects
effects effects Pharm. firms only

INTERNATIONALt 0.41∗∗∗ 0.34∗∗ 0.42∗∗∗ 0.42∗∗∗ 0.40∗∗∗ 0.49∗∗∗

(2.72) (2.16) (4.16) (3.91) (3.75) (4.84)
[0.94] [0.67] [1.76] [1.70] [2.67] [0.83]
DRUG PATENT 0.03∗∗∗ 0.05∗∗∗ 0.01∗∗∗ 0.01∗∗∗ 0.01∗∗∗ 0.01∗∗
STOCK(t−1,t−3) (7.86) (3.04) (3.8) (3.17) (3.97) (1.69)
[0.06] [0.10] [0.05] [0.05] [0.10] [0.01]
FERMENTATION 0.02∗∗∗ 0.02 0.02∗∗ 0.02∗∗ −0.01 0.04∗∗∗
PATENT
STOCK(t−1,t−3)
(2.52) (1.08) (1.94) (1.87) (0.82) (4.86)
[0.04] [0.03] [0.10] [0.09] [−0.07] [0.07]
ENTRANT −0.84∗∗∗ −0.72∗∗∗ −1.19∗∗∗ −1.11∗∗∗
(3.27) (3.38) (4.96) (3.92)
[−1.90] [−1.42] [−4.64] [−4.12]
SALESt −0.13∗∗ 0.46∗∗∗
(1.93) (4.09)
[−0.49] [2.90]
Intercept 0.79∗∗∗ 0.40∗∗ 1.67∗∗∗ 1.91∗∗∗ 1.01∗∗∗
(6.35) (2.02) (6.25) (6.22) (3.57)
[1.67] [0.76] [6.48] [7.12] [6.53]
χ 2 for covariates 321.46∗∗∗ 60.40∗∗∗ 99.65∗∗∗ 94.57∗∗∗ 114.14∗∗∗ 37.55∗∗∗
(d.f.) (4) (4) (4) (5) (4) (3)
χ 2 test of overdispersion 344.02∗∗∗ 128.78∗∗∗ 95.37∗∗∗ 14.54∗∗∗ 290.84∗∗∗
(col. 2) and of panel
model (other cols.)
n 585 585 585 521 231 522

∗ ∗∗ ∗∗∗
p < 0.1; p < 0.05; p < 0.01; one-tailed tests

Copyright  2004 John Wiley & Sons, Ltd. Strat. Mgmt. J., 26: 121–140 (2005)
132 J. Penner-Hahn and J. M. Shaver
and Poisson specifications with the exception of
the variable FERMENTATION PATENT STOCK.
Turning to the negative binomial results in Model
2, the positive and significant effect of DRUG
PATENT STOCK indicates that firms receive more
drug patents if they have greater drug patents
over the previous 3 years. This is consistent with
Hypothesis 1. In order to assess the magnitude of
this effect, we examine the marginal effect of this
variable on the dependent variable. Because the
negative binomial model is a non-linear estimator,
the coefficient estimate is not the marginal effect.
Rather, the marginal effect equals eβX b. Therefore,
the magnitude of the marginal effect is contingent
on the values of the independent variables. We
present the marginal effect, evaluated at the mean
of the independent variables, in square brackets in
the table. The marginal effect of DRUG PATENT
STOCK is 0.1. This indicates that firms with 10
drug patents in the previous three years will have
one more patent in the current year.
The coefficient estimate of FERMENTATION
PATENT STOCK is non-significant in Model 2.
Therefore, we do not find support for Hypothesis
2 in this specification. As we noted previously,
this variable was significant in Model 1. However,
given that the data are overdispersed, we favor the
interpretation of the negative binomial estimates.
The coefficient estimate of INTERNATIONAL
is positive and significant in Model 2. The marginal
effect of 0.67 suggests that firms increase their
patent output once they undertake international
R&D.5 We find evidence that many firms increase
their drug patenting when they conduct interna-
tional R&D. We did not hypothesize an effect
for this variable in the entire sample because we
hypothesize that the effect of INTERNATIONAL
is contingent on a firm’s research capabilities.
Finally, we find that ENTRANT is negative and
highly significant. The size of the coefficient esti-
mate indicates that firms that entered the pharma-
ceutical industry have about 1.4 less patents in any
given year than pharmaceutical firms.6
5
Because INTERNATIONAL is a dummy variable, we calcu-
late its marginal effect in the following manner: eβX1 − eβX0 .
Where b is the vector of coefficient estimates, eβX1 is evaluated
at INTERNATIONAL = 1 and at the mean of the other inde-
pendent variables; eβX0 is evaluated at INTERNATIONAL = 0
and at the mean of the other independent variables.
6
In the acquisition context, which is different than the situation
that we examine, Hitt et al. (1991) show that acquisitions by
diversifying acquirers have a negative effect on patent intensity.
Copyright  2004 John Wiley & Sons, Ltd.
As previously mentioned, we wish to take advan-
tage of our panel data structure and better isolate
the impact of our hypothesized effects from poten-
tial sources of unobserved firm heterogeneity. This
is an important consideration because the firms
choose to conduct international R&D (e.g., Shaver,
1998). The use of panel data techniques such as
random effects or fixed effects mitigates this prob-
lem in that it is one way to control for constant firm
effects that are not included in the set of indepen-
dent variables.
Model 3 of Table 2 presents the negative bino-
mial specification with firm random effects. The
test of the inclusion of the random effects is highly
significant, indicating that there exist unobserved
firm effects. The statistical significance level of all
variables increases with the addition of the random
effect. As a result, the variable FERMENTATION
PATENT STOCK becomes statistically significant
as predicted by Hypothesis 2.
There are some noticeable changes in the mag-
nitude of the marginal effects in the random
effects specification.7 First, although the variable
increases in statistical significance, the marginal
effect of DRUG PATENT STOCK is half of what
it was in the model without the random effects. In
contrast, the marginal effect of INTERNATIONAL
increases. In this specification, a firm that conducts
international R&D in a year generates 1.76 more
patents than a firm without international R&D.
Finally, the marginal effect of ENTRANT also
decreases to −4.64.8
Models 4 through 6 of Table 2 further examine
the sensitivity of these results across various spec-
ifications. In Model 4 we include SALES as an
independent variable. As previously discussed, the
sample size drops to 521 for this specification due
to missing data. In Model 4, all of the variables
of interest remain with similar levels of statisti-
cal significance and marginal effect. This gives
us confidence that the previously discussed results
were not driven by changes in firm sales. How-
ever, unexpectedly, SALES takes a negative and
significant effect. The marginal effect is such that
7
We evaluate the marginal effects at the mean of the independent
variables and at the mean random effect.
8
ENTRANT represents a group fixed effect because we expect
differences in the DRUG PATENT COUNT of firms that
are diversifying entrants to the pharmaceutical industry. The
random-effects model then accounts for the possibility of firm
heterogeneity beyond this group effect. Namely, within the
groups of entrants and non-entrants there exist firm differences
that we capture by a random effect.
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
a trillion yen increase in sales decreases patents
by one-tenth. This effect, while statistically signif-
icant, is not large given that the average value of
SALES is 2.2 trillion yen.
Our interpretation of the negative effect is that
the following two factors are working in tandem.
First, many of the entrants to the pharmaceu-
tical business are large diversified firms, which
have most of their sales in non-pharmaceutical
lines of business. The positive correlation between
SALES and ENTRANT supports this claim. Sec-
ond, should diversifying entrants put more re-
sources and effort into pharmaceutical research
when their other businesses are deteriorating, then
sales would be negatively correlated with drug
patents.
To further assess this interpretation, we restrict
the sample to the pharmaceutical firms (i.e., we
removed firms where ENTRANT equaled 1 from
the sample), which reduces the sample size from
521 to 231. We present these results in Model
5. The effect of SALES is positive and highly
significant, as we would expect. This lends sup-
port to our interpretation of the negative effect
of SALES in the entire sample. The magnitude
of the marginal effect is that a one trillion yen
increase in sales results in almost one-half addi-
tional patent, which is not a large effect. Turn-
ing to the other coefficient estimates, there are
three noteworthy differences between the results
in Model 5 and Model 4. First, the marginal effect
of INTERNATIONAL is 2.67 for the restricted
sample of pharmaceutical firms vs. the value of
1.70 in the full sample. This result foreshadows
the support we find for Hypothesis 3 in later anal-
ysis. Consistent with Hypothesis 3 we expect that
the pharmaceutical firms that comprise the subsam-
ple have greater ability to interpret and incorporate
international research and hence will more likely
benefit from such. Second, the marginal effect of
DRUG PATENT STOCK is larger (0.1 vs. 0.05).
This might indicate that past research leads to a
greater trajectory into future patenting for phar-
maceutical firms vs. entrants. Third, the effect of
FERMENTATION PATENT STOCK loses signifi-
cance. This could indicate that the pharmaceutical
firms do not have extensive biotech skills or that
biotech skills are not the basis of their drug dis-
covery.
Our final sensitivity analysis of the specifica-
tion in Model 3 is to replace the random effect
with firm fixed effects, which we present in Model
Copyright  2004 John Wiley & Sons, Ltd.
133
6.9 The sample size drops in the fixed-effects
analysis because seven companies had zero val-
ues of DRUG PATENT COUNT over the entire
sample and meaningful fixed effects cannot be
estimated for these firms given the model speci-
fication. In addition, the effect of ENTRANT is
subsumed within the firm fixed effect and, there-
fore, is not included in the specification. When
comparing these results to those in Model 3, all
variables of interest remain statistically significant.
In order to calculate marginal effects, we evalu-
ate the marginal effects at the average fixed effect
of the firms that comprise the sample. There are
some changes in the level of the marginal effects.
The marginal effects of INTERNATIONAL and
DRUG PATENT STOCK decrease, while the effect
of FERMENTATION PATENT STOCK increases.
However, these are not as readily comparable to
those in Model 3 because we have removed firms
from the sample that had zero patents for the entire
panel. Thus, we removed a very selected portion
of the sample in this analysis. Nevertheless, even
by removing a very selected portion of the sample,
we find that the results are very consistent between
the random and fixed-effect specifications.
To summarize the results from Table 2, we
find support for Hypothesis 1. Firms with greater
DRUG PATENT STOCK generate more patents.
We find mixed support for Hypothesis 2. We find
some evidence that FERMENTATION PATENT
STOCK increases patent output; however, this
finding is sensitive to specification. This could
be attributable to the fact that the patent class
of Molecular Biology and Biochemistry contains
patents other than fermentation. Finally, we find
that across specifications firms that conduct inter-
national R&D have greater patent output. We
turn to testing Hypotheses 3 and 4, which argues
that this effect is contingent on a firms’ DRUG
PATENT STOCK and FERMENTATION PATENT
STOCK.
In order to test Hypothesis 3, we split the
sample into observations where DRUG PATENT
9
We estimate this specification by including individual firm
dummy variables for each firm (e.g., Greene, 2001). We do
so for two reasons. First, the commonly used negative bino-
mial estimator with fixed effects (Hausman et al., 1984) does
not necessarily control for all stable firm effects (Allison and
Waterman, 2001). Second, the estimator conditions out the firm
effects. Because marginal effects are contingent on where a firm
lies upon the curve, conditioning out the firm effects make it
difficult to assess the marginal effect of the covariates (Greene,
2000).
Strat. Mgmt. J., 26: 121–140 (2005)
134 J. Penner-Hahn and J. M. Shaver

STOCK is zero and where it is greater than zero.10
Hypothesis 3 predicts that INTERNATIONAL will
have a significant positive effect in the subsam-
ple where DRUG PATENT STOCK is greater
than zero. Likewise, we do not expect much, if
any, effect of INTERNATIONAL in the subsample
where DRUG PATENT STOCK is zero. We choose
zero as the value of DRUG PATENT STOCK to
split the sample because we believe that there are
important differences in the research capabilities
of firms that have pharmaceutical patents over a
period of time vs. those that have none.
We examine this hypothesis using two of the
specifications that we previously described. First,
we use the negative binomial specification without
firm effects, but with adjusted variances to account
for the fact that each firm has 9 years of data
in the panel. This parallels the specification from
Model 2 of Table 2. Table 3 presents these results.
Second, we examine the negative binomial spec-
ification with firm random effects. This specifi-
cation parallels the specification of Model 3 in
Table 2. Table 4 presents these results. We pre-
fer these specifications because they allow us to
include the entire sample. Nevertheless, we find
very similar results with consistent interpretations
for the other specifications that we presented in
Table 2.11 The advantage of the specification in
Table 4 over Table 3 is that it employs the random
effect in order to better control for unobserved firm

10
The sample is split by observation, not firm. This means that a
firm (but not the same firm-year observation) can appear in both
subsamples if the value of DRUG PATENT STOCK is zero in
some years and non-zero in others.
11
Restricting the sample to the pharmaceutical firms (i.e., Model
5 of Table 2) was not possible in the forthcoming analysis due to
the very small sample size for the set of firms with zero values
of DRUG PATENT STOCK.

Table 3. Analyses of firm drug patents: negative binomial regression models. Dependent Variable: DRUG PATENT
COUNTt (t-values in parentheses) [marginal effects in square brackets] {variance of marginal effects in curly brackets}

Model 1 Model 2 Model 3 Model 4

DRUG PATENT DRUG PATENT FERM. PATENT FERM. PATENT
STOCK = 0 STOCK >0 STOCK = 0 STOCK >0

INTERNATIONALt 0.57 0.32∗∗∗ 0.65∗∗∗ 0.23∗

(1.16) (2.62) (2.8) (1.36)
[0.11] [1.12] [0.65] [0.66]
{0.12} {0.44} {0.30} {0.50}
t-test that the marginal 2.35∗∗∗ 0.02
effect of
INTERNATIONAL in
odd model is ≤ even
model
DRUG PATENT 0.04∗∗∗ 0.12∗∗∗ 0.04∗∗∗
STOCK(t−1,t−3) (4.21) (4.7) (3.39)
[0.13] [0.10] [0.12]
FERMENTATION PATENT −0.01 0.01 0.01
STOCK(t−1,t−3) (0.07) (1.05) (0.52)
[0.00] [0.03] [0.02]
ENTRANT −1.32∗∗∗ −0.33∗∗ −0.90∗∗∗ −0.60∗∗∗
(3.07) (1.95) (2.39) (2.75)
[−0.41] [−1.06] [−0.80] [−1.73]
Intercept −0.69 0.69∗∗∗ −0.31 0.72∗∗∗
(0.34) (5.33) (1.20) (3.99)
[−0.11] [2.33] [−0.25] [2.03]
χ 2 for covariates 9.84∗∗ 73.58∗∗∗ 80.00∗∗∗ 51.26∗∗∗
(d.f.) (3) (4) (3) (4)
χ 2 test of over-dispersion 4.11∗∗ 191.73∗∗∗ 51.47∗∗∗ 198.05∗∗∗
(1 d.f.)
n 178 407 240 345

∗ ∗∗ ∗∗∗
p < 0.1; p < 0.05; p < 0.01; one-tailed tests

Copyright  2004 John Wiley & Sons, Ltd. Strat. Mgmt. J., 26: 121–140 (2005)

heterogeneity. The advantage of the specification
in Table 3 over Table 4 is that the marginal effects
are more easily comparable across models. This is
because the marginal effects in Table 4 are evalu-
ated at the mean random effect within each model.
However, the mean random effect might not be
equal across models.
There are two factors we examine in making
our assessment of whether INTERNATIONAL has
a greater effect for firms with existing levels of
patent stock. First, we examine the significance
level of the coefficient estimate of INTERNA-
TIONAL in the two subsamples. Models 1 and 2 of
Tables 3 and 4 present these results. We find that
INTERNATIONAL has a positive and significant
effect when DRUG PATENT STOCK is greater
than zero. The effect of INTERNATIONAL is not
significant when DRUG PATENT STOCK equals
zero. This lends support to our hypothesis.
We then examine the marginal effect of INTER-
NATIONAL across the two subsamples and test
if the marginal effect is greater in the subsam-
ple where DRUG PATENT STOCK is greater
than zero. We focus on the marginal effect rather
than the coefficient estimate because of the non-
linearity of the estimator. Namely, if the obser-
vations in one model lie on a different part of
the estimated curve, then comparing the coef-
ficient estimates will be misleading. Turn first
to Table 3. The marginal effect of INTERNA-
TIONAL when DRUG PATENT STOCK equals
zero is 0.11. However, the marginal effect of
INTERNATIONAL when DRUG PATENT
STOCK is greater than zero is 1.12. Below the
marginal effect, we present the variance of the
marginal effect.12 With these data, we then conduct
a t-test of the marginal effects across the two mod-
els. The t-test is highly significant, indicating that
the marginal effect of INTERNATIONAL for firms
with drug patents from the previous 3 years is
greater than those without. This supports Hypoth-
esis 3.
Table 4 adds the random effect to the specifica-
tions in Table 3. Once again the effect of INTER-
NATIONAL is significantly greater than zero for
the set of firms where DRUG PATENT STOCK is
greater than zero and not significantly greater than
12
The variance of the marginal effect is calculated as G VG,
where G is the matrix of first derivatives of the marginal effect
by β (see footnote 5) and V is the variance–covariance matrix
of the estimates of β.)
Copyright  2004 John Wiley & Sons, Ltd.
Does International R&D Increase Patents? 135
zero for the set of firms where DRUG PATENT
STOCK equals zero. The marginal effects are also
significantly different (1.12 vs. 0.34, p < 0.08);
however, the level of significance is lower than
in the previous table. Two points are worth noting
here. First, the marginal effects are calculated at the
average random effect within each model, which
might not be equal. In particular, we would expect
that companies with DRUG PATENT STOCK
greater than zero will have more patents in a given
year based on the results in Table 2 and thus the
average firm effect should be larger. Therefore, the
test in Table 4 might exhibit a conservative bias.
Second, the test of including the random effect in
the specification is non-significant in Model 1 yet
it is in Model 2. This indicates that the specifi-
cation in Model 1 of Table 3 is preferred to the
specification in Model 1 of Table 4. If we compare
the random effect of Model 1 Table 3 to Model 2
Table 4 (0.11 vs. 1.12), the difference in marginal
effect is highly significant (p < 0.01). Overall, we
find very consistent support for our hypothesis that
firms benefit from international R&D only when
they possess skills in the underlying technologies.
We follow a similar procedure to test Hypothesis
4. We create two subsamples—one where FER-
MENTATION PATENT STOCK is zero and one
where it is greater than zero—and examine the
estimates of INTERNATIONAL across the mod-
els. Models 3 and 4 of Tables 3 and 4 present
these estimates. In all models, the coefficient esti-
mates of INTERNATIONAL are positive and sig-
nificantly different from zero (although marginally
so in Table 3 model 4). Moreover, the difference
in the marginal effects across models are non-
significant. The marginal effects in Table 3 are
0.65 vs. 0.66 and in Table 4 are 1.01 vs. 1.82.
Because there is no difference in the marginal
effect, we do not find support for Hypothesis 4.
To summarize the results from Tables 3 and 4,
we find that international R&D increases patent
count for firms that have drug patents in the pre-
vious 3 years. Otherwise, there is no effect. How-
ever, the positive effect of international R&D on
patent count is not contingent on firms having fer-
mentation patents in the previous 3 years.
Further robustness checks

We performed a number of additional robustness
checks to further verify our interpretation of the
results that we report. First, we examined whether
Strat. Mgmt. J., 26: 121–140 (2005)
136 J. Penner-Hahn and J. M. Shaver

Table 4. Analyses of firm drug patents: negative binomial random-effects regression models Dependent variable:
DRUG PATENT COUNTt (t-values in parentheses) [marginal effects in square brackets] {variance of marginal effects
in curly brackets}

Model 1 Model 2 Model 3 Model 4

DRUG PATENT DRUG PATENT FERM. PATENT FERM. PATENT
STOCK = 0 STOCK > 0 STOCK = 0 STOCK > 0

INTERNATIONALt 0.48 0.37∗∗∗ 0.65∗∗∗ 0.28∗∗∗

(1.02) (3.69) (3.5) (2.38)
[0.34] [1.12] [1.01] [1.82]
{0.47} {0.30} {0.57} {0.95}
t-test that the marginal effect 1.39∗ 0.73
of INTERNATIONAL in
odd model is ≤ even model
DRUG PATENT 0.02∗∗∗ 0.06∗∗∗ 0.02∗∗∗
STOCK(t−1,t−3) (4.29) (3.24) (3.97)
[0.13] [0.08] [0.10]
FERMENTATION PATENT 0.02 0.02∗ 0
STOCK(t−1,t−3) (0.25) (1.48) (0.32)
[0.01] [0.03] [0.03]
ENTRANT −1.25∗∗∗ −0.65∗∗∗ −1.42∗∗∗ −0.89∗∗∗
(2.54) (3.33) (4.46) (3.51)
[−0.75] [−1.10] [−1.82] [−5.63]
Intercept 0.50 1.60∗∗∗ 0.65 2.00∗∗∗
(0.56) (5.71) (1.18) (5.53)
[0.30] [2.33] [0.83] [12.67]
χ 2 covariates 6.55∗ 74.22∗∗∗ 84.53∗∗∗ 50.96∗∗∗
(d.f.) (3) (4) (3) (4)
χ 2 test of random effect 0.01 47.56∗∗∗ 2.58∗ 64.44∗∗∗
(1 d.f.)
n 178 407 240 345

∗ ∗∗ ∗∗∗
p < 0.1; p < 0.05; p < 0.01; one-tailed tests

the results of INTERNATIONAL in Model 1 of
Tables 3 and 4 might be driven by a lack of obser-
vations where INTERNATIONAL takes the value
of one. In Model 1 INTERNATIONAL takes the
value of one for 19 percent of the observations.
Second, we experimented with different cut-off
points for splitting the subsamples in Tables 3
and 4, which use a cut-off of zero. In examining
cut-offs of one to four patents, we found results
that were consistent with those presented in the
table.
Finally, because most of the international expan-
sion activity of these firms is into the United
States, there is the possibility that INTERNA-
TIONAL captures an increased propensity of firms
to patent in the United States once they locate
there and does not capture increased innovative
output. There are three reasons why we expect
that this is not driving the results. First, as we
previously discussed, many of the foreign firms in
Copyright  2004 John Wiley & Sons, Ltd.
this sample (and outside of it) extensively patent
in the United States when they do not have opera-
tions there. Second, to the extent that operating in
the United States per se increases the propensity
to patent in the United States, we would expect
INTERNATIONAL to increase equally in both
subsamples. This is not consistent with our results.
Nevertheless, there is a possibility that this alter-
native explanation would be most pronounced for
firms with drug patent stocks greater than zero.
To assess this, we reran the analysis by examining
only the international expansion activities of the
five firms that conducted international R&D out-
side of the United States. Although the number of
observations where INTERNATIONAL equals one
becomes very small, we find results that are quali-
tatively similar to the ones that we present. The
coefficient estimate of INTERNATIONAL was
negative for the subsample where DRUG PATENT
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
STOCK equals zero and positive for the subsam-
ple where DRUG PATENT STOCK was greater
than zero. Although neither coefficient estimate
was significantly zero, they were significantly dif-
ferent from each other. Again, this suggests that,
as hypothesized, international benefits firms with
greater absorptive capacity.
In summary, the empirical results are consis-
tent with our predictions regarding firms’ previ-
ous innovative activities. We find that firms with
greater underlying research capabilities are more
innovative. We also find that firms with underly-
ing research capabilities benefit most from interna-
tional R&D activities. Therefore, it is the combina-
tion of international R&D activities with existing
skills in underlying research that increases innova-
tive output.
DISCUSSION
Our focal finding is that Japanese firms that com-
menced international R&D, on average, increased
patent output. However, this effect was only signif-
icant for firms that had existing capabilities in the
underlying technology prior to expanding interna-
tionally. This finding is important for the following
reasons.
First, in line with many recent prescriptions, it
provides empirical evidence that there exist bene-
fits from conducting international R&D. This helps
augment the findings of Almeida (1996). How-
ever, because we examine all Japanese entrants
that expanded overseas, we are able to identify an
important contingency with respect to when inter-
national R&D is beneficial. Namely, only firms
that have skills in the underlying or underlying
technology benefit from international R&D.
Second, our results suggest that tapping into for-
eign knowledge bases is not a panacea, especially
for firms that are lagging in technological skill.
Rather, our results suggest that such firms are the
ones that do not benefit from such activities. This is
important because international R&D is an effort
that is costly in terms of capital, scientific, and
managerial resources. We do not want to rule out
the possibility that technologically lagging firms
can benefit from conducting international R&D.
However, our results show that this will likely
require more refined or nuanced strategies. Iden-
tifying such strategies is beyond the scope of this
Copyright  2004 John Wiley & Sons, Ltd.
137
investigation. While we do not rule out the possi-
bility that such strategies exist, overall we do not
find firms that lack technological skills employing
effective strategies.
Finally, our findings highlight that conducting
international R&D and investing in R&D at home
are complements rather than substitutes. It is the
firms that have technological capabilities that most
greatly benefit from international R&D. This in
turn, can start a virtuous circle whereby this greater
knowledge base can further foster benefits from
increased international R&D.
Overall we find that international R&D can play
an important role in firms’ innovation. However,
this role should be coordinated with its underly-
ing efforts in R&D at home. Moreover, due to the
contingent nature of when this effort is successful,
we highlight the importance of managerial strate-
gic decision making with respect to successfully
engaging in international R&D.
CONCLUSION
We find that international R&D activities increase
the patent output of Japanese pharmaceutical firms.
However, it is firms with existing pharmaceutical
research capabilities that benefit most from inter-
national R&D. While there are many benefits that
firms can obtain from international R&D activities,
there also exist many complications that they incur
when making and managing such investments. For
this reason, only under certain conditions are net
benefits from international R&D to be realized.
This finding corroborates previous research show-
ing that not all firms enjoy success with their
international expansion efforts even when there
exist advantages to having international operations
(Mitchell et al., 1992).
We find that in order for firms’ patent output to
benefit from international R&D, they must possess
research capabilities in addition to overseas R&D
activities. This is consistent with our argument
that knowledge-based skills and capabilities are
not simply acquired and transferred back to the
parent. Rather, firms must have existing research
capabilities in order to fully benefit from engaging
in international R&D. This finding is consistent
with the notion of absorptive capacity as described
by Cohen and Levinthal (1990).
Our results have important implications for rec-
onciling theories of asset-seeking FDI (e.g., Kogut
Strat. Mgmt. J., 26: 121–140 (2005)
138 J. Penner-Hahn and J. M. Shaver
and Chang, 1991) with theories of asset-based FDI
such as internalization (e.g., Buckley and Cas-
son, 1976; Dunning, 1977). Asset-seeking theories
argue that firms make direct foreign investments
in order to obtain assets that they do not currently
possess. Asset-based theories, on the other hand,
argue that firms possess proprietary assets that they
choose to internalize by making direct investment
overseas.
Our research suggests that while the Japanese
pharmaceutical firms expanded in order to gain
access to skills and technology not resident in
Japan, those firms with pharmaceutical research
capabilities were best able to enhance their patent
output. Therefore, it appears that both asset-seeking
and asset-based factors are important in the success
of international R&D. Just as entering a country
that has a large consumer market is no guarantee
for successful asset-based FDI, entering a market
that has a unique knowledge base is no guarantee
of successful asset-seeking FDI.
Many previous studies have shown the impor-
tance of R&D when assessing which firms expand
overseas (e.g., Hennart and Park, 1994) and when
international expansion is successful (e.g., Morck
and Yeung, 1991, 1992). In this paper we identify
when international R&D activities enhance firms’
innovation. Continuing research into what deter-
mines successful international expansion, espe-
cially expansion activities that focus on the acqui-
sition of knowledge within geographical locations,
is warranted.
ACKNOWLEDGEMENTS
We appreciate helpful comments from the anony-
mous referees, Richard Arend, Joel Baum, José
Campa, John Cantwell, Wilbur Chung, Bill Greene,
Xavier Martin, Will Mitchell, Tom Pugel, Chris
Tucci, seminar participants at HEC, INSEAD, the
Universidad Complutense de Madrid, the Univer-
sity of Toronto, and at the Academy of Inter-
national Business Annual Meetings in Vienna.
We also wish to thank Hiro Ishida and Anne
Parmigiani for research assistance. Myles Shaver’s
research was supported by a grant from the Ten-
neco Fund Program at the Stern School of Busi-
ness, New York University. Joan Penner-Hahn’s
research was supported by the Japan Technology
Management Program at the University of Michi-
gan.
Copyright  2004 John Wiley & Sons, Ltd.
REFERENCES
Acs ZJ, Audretsch D. 1989. Patents as a measure of
innovation. Kyklos 42(2): 171–180.
Ahuja G. 2000. The duality of collaboration: inducements
and opportunities in the formation of interfirm
linkages. Strategic Management Journal , Special
Issue 21(3): 317–343.
Allison PD, Waterman R. 2001. Fixed-effects negative
binomial regression models. Working paper, Univer-
sity of Pennsylvania.
Almeida P. 1996. Knowledge sourcing by foreign
multinationals: patent citation analysis in the
U.S. semiconductor industry. Strategic Management
Journal , Winter Special Issue 17(3): 155–165.
Anand J, Delios A. 2002. Absolute and relative resources
as determinants of international acquisitions. Strategic
Management Journal 23(2): 119–134.
Bound J, Cummins J, Griliches Z, Hall BH, Jaffe A.
1984. Who does R&D and who patents? In R&D
Patents and Productivity, Griliches Z (ed). University
of Chicago Press: Chicago, IL; 21–54.
Buckley PJ, Casson MC. 1976. The Economic Theory of
the Multinational Enterprise. Macmillan: London.
Cameron AC, Trivedi PK. 1998. Regression Analysis of
Count Data. Cambridge University Press: Cambridge,
U.K.
Cantwell J. 1989. Technological Innovation and Multi-
national Corporations. Basil Blackwell: Cambridge,
MA.
Cantwell J. 1992. The internationalisation of technologi-
cal activity and its implications for competitiveness. In
Technology Management and International Business,
Granstrand O, Hakanson L, Sjolander S (eds). Wiley:
New York; 25–95.
Cantwell J. 1995. The globalisation of technology: what
remains of the product cycle model? Cambridge
Journal of Economics 19: 155–174.
Caves RE. 1971. International corporations: the indus-
trial economics of foreign investment. Economica
38(February): 1–27.
Caves RE. 1982. Multinational Enterprise and Economic
Analysis. Cambridge University Press: Cambridge,
U.K.
Cohen WM, Levinthal DA. 1990. Absorptive capacity:
a new perspective on learning and innovation.
Administrative Science Quarterly 35(1): 128–152.
Comanor WS, Scherer FM. 1969. Patent statistics as a
measure of technical change. Journal of Political
Economy 77(3): 392–398.
Dierickx I, Cool K. 1989. Asset stock accumulation and
sustainability of competitive advantage. Management
Science 35(12): 1504.
DeMeyer A, Mizushima A. 1989. Global R&D manage-
ment. R&D Management 19(2): 135–146.
DeMeyer A. 1992. Management of international R&D
operations. In Technology Management and Interna-
tional Business, Granstrand O, Hakanson L, Sjolan-
der S (eds). Wiley: New York.
Dunning J. 1977. Trade, location of economic activity and
MNE: a search for an eclectic approach. In The Inter-
national Allocation of Economic Activity, Ohlin B,
Strat. Mgmt. J., 26: 121–140 (2005)
 Does International R&D Increase Patents?
Hesselborn PO, Wijkman PM (eds). Macmillan: Lon-
don; 395–418.
Dunning J. 1998. Location and the multinational
enterprise: a neglected factor. Journal of International
Business 29(1): 45–67.
Dunphy JF. 1988. Japan’s patent system comes under fire.
Chemical Week 143(4): 26–28.
Fayerweather J. 1969. International Business Manage-
ment: A Conceptual Framework . McGraw-Hill Series
in International Business. McGraw-Hill: New York.
Fox S. 2001. Business models: contract biopharma-
ceutical manufacturing. Chemical Market Reporter
260(16): 20–22.
Grace E. 1997. Biotechnology Unzipped: Promises and
Realities. Joseph Henry: Washington, DC.
Graves SB, Langowitz NS. 1993. Innovative productivity
and returns to scale in the pharmaceutical industry.
Strategic Management Journal 14(8): 593–605.
Greene W. 2000. Econometric Analysis (4th edn).
Prentice-Hall: Upper Saddle River, NJ.
Greene W. 2001. Fixed and random effects in nonlinear
models. Working paper, Stern School of Business,
New York University.
Griliches Z. 1980. R & D and the productivity slowdown.
American Economic Review 70(2): 343–348.
Griliches Z. 1986. Productivity, R and D, and basic
research at the firm level in the 1970’s. American
Economic Review 76(1): 141–154.
Griliches Z. 1990. Patent statistics as economic indi-
cators: a survey. Journal of Economic Literature
28(December): 1661–1707.
Hall BH, Griliches Z, Hausman J. 1986. Patents and R
and D: is there a lag? International Economic Review
27(2): 265–283.
Hausman J, Hall BH, Griliches Z. 1984. Econometric
models for count data with an application to
the patent–R&D relationship. Econometrica 52(4):
909–938.
Helfat C. 1997. Know-how and asset complementarity
and dynamic capability accumulation: the case of
R&D. Strategic Management Journal 18(5): 339–360.
Henderson R, Cockburn I. 1994. Measuring competence?
Exploring firm effects in pharmaceutical research.
Strategic Management Journal , Winter Special Issue
15: 63–84.
Henderson R, Cockburn I. 1996. Scale, scope, and
spillovers: the determinants of research productivity
in drug discovery. RAND Journal of Economics 27:
32–59.
Hennart J-F, Park Y-R. 1994. Location, governance,
and strategic determinants of Japanese manufacturing
investment in the United States. Strategic Management
Journal 15(6): 419–436.
Hitt MA, Hoskisson RE, Ireland RD, Harrison JS. 1991.
Effects of acquisitions on R&D inputs and outputs.
Academy of Management Journal 34(3): 693–706.
Jaffe AB, Trajtenberg M, Henderson R. 1993. Geo-
graphic localization of knowledge spillovers as evi-
denced by patent citations. Quarterly Journal of Eco-
nomics 108(3): 577.
Jaffe AB. 1986. Technological opportunity and spillovers
of R&D: evidence from firm’s patents, profits, and
Copyright  2004 John Wiley & Sons, Ltd.
139
market value. American Economic Review 76(5):
984–1001.
JPMA (Japanese Pharmaceutical Manufacturers Associa-
tion). Data Book 1992 . Tokyo, Japan.
Kogut B. 1991. Country capabilities and the permeability
of borders. Strategic Management Journal , Summer
Special Issue 12: 33–47.
Kogut B, Chang SJ. 1991. Technological capabilities and
Japanese foreign direct investment in the United
States. Review of Economics & Statistics 73(3):
401–413.
Kogut B, Zander U. 1992. Knowledge of the firm,
combinative capabilities, and the replication of
technology. Organization Science 3: 383–397.
Komaran RV, Goodman RA 1993. Acquisition of domes-
tic and foreign technologies by American multina-
tional corporations. In Academy of Management Pro-
ceedings, Atlanta, GA.
Kuemmerle W. 1997. Building effective R&D capabili-
ties abroad. Harvard Business Review 75(2): 61–70.
Levin RC. 1986. A new look at the patent system.
American Economic Review 76(2): 200.
Mansfield E. 1962. Entry, Gibrat’s law, innovation, and
the growth of firms. American Economic Review
52(5): 1023–1051.
Mansfield E. 1965. Rates of return from industrial
research and development. American Economic
Review 55(1/2): 310–322.
Mansfield E. 1980. Basic research and productivity
increase in manufacturing. American Economic
Review 70(5): 863–873.
Mezias SJ, Glynn MA. 1993. The three faces of corporate
renewal: institution, revolution, and evolution. Strate-
gic Management Journal 14(2): 77–161.
Mitchell W, Shaver JM, Yeung B. 1992. Getting there in
a global industry: impacts on performance of changing
international presence. Strategic Management Journal
13(6): 419–432.
Morck R, Yeung, B. 1991. Why investors value
multinationality. Journal of Business 64(2): 165–188.
Morck R, Yeung B. 1992. Internalization: an event study
test. Journal of International Economics 33: 41–56.
National Research Council. 1992. U.S.–Japan Technol-
ogy Linkages in Biotechnology: Challenges for the
1990’s. National Academy of Sciences.
Nelson RR (ed). 1993. National Innovation Systems.
Oxford University Press: New York.
Pakes A. 1985. On patents, R & D, and the stock market
rate of return. Journal of Political Economy 93(2):
390–409.
Penner-Hahn JD. 1995. The internationalization of
research and development: a firm level study. PhD
dissertation, University of Michigan, Ann Arbor, MI.
Penner-Hahn JD. 1998. Firm and environmental influ-
ences on the mode and sequence of foreign research
and development activities. Strategic Management
Journal 19(2): 149–168.
Pisano GP. 1990. The R&D boundaries of the firm: an
empirical analysis. Administrative Science Quarterly
35(1): 153–176.
Polyani M. 1964. Personal Knowledge. Harper & Row:
New York.
Strat. Mgmt. J., 26: 121–140 (2005)
140 J. Penner-Hahn and J. M. Shaver
Reich MR. 1990. Why the Japanese don’t export more
pharmaceuticals: health policy as industrial policy.
California Management Review , Winter: 124–150.
Scherer FM, Ross D. 1990. Industrial Market Structure
and Economic Performance. Houghton Mifflin:
Boston, MA.
Shaver JM. 1998. Accounting for endogeneity when
assessing strategy performance: does entry mode
choice affect FDI survival? Management Science
44(4): 571–585.
Shaver JM, Mitchell W, Yeung B. 1997. The effect of
own-firm and other-firm experience on foreign direct
investment survival in the United States, 1987–1992.
Strategic Management Journal 18(10): 811–824.
Szulanski G. 1996. Exploring internal stickiness: imped-
iments to the transfer of best practice within the firm.
Copyright  2004 John Wiley & Sons, Ltd.
Strategic Management Journal , Winter Special Issue
17: 27–43.
Teece DJ. 1976. Technology transfer by multinational
firms: the resource cost of transferring technological
know-how. Economic Journal 87: 242–261.
Teece DJ. 1986. Profiting from technological innovation.
Research Policy 15: 285–306.
Terleckyj NE. 1980. What do R&D numbers tell us about
technological change? American Economic Review
70(2): 55–61.
Wesson T. 1999. A model of asset-seeking foreign direct
investment driven by demand conditions. Canadian
Journal of Administrative Sciences 16(1): 1–10.
Wineberg A. 1988. The Japanese patent system: a non-
tariff barrier to foreign businesses? Journal of World
Trade 22(1): 11–22.
Strat. Mgmt. J., 26: 121–140 (2005)