Strategic Entrepreneurship Journal

Strat. Entrepreneurship J., 4: 106–127 (2010)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/sej.86

A STORY OF BREAKTHROUGH VERSUS

INCREMENTAL INNOVATION: CORPORATE
ENTREPRENEURSHIP IN THE GLOBAL
PHARMACEUTICAL INDUSTRY
DENISE DUNLAP-HINKLER,1* MASAAKI KOTABE,2 and RAM MUDAMBI2,3
1
College of Business Administration, Northeastern University, Boston,
Massachusetts, U.S.A.
2
Fox School of Business, Temple University, Philadelphia, Pennsylvania, U.S.A.
3
Center for Strategic Management and Globalization, Copenhagen Business
School, Copenhagen, Denmark

Breakthrough innovations are difficult to create, yet they are critical to long-term competitive
advantage. This highlights the considerable opportunities and risks that face corporate entre-
preneurs. We study the complex explorative and exploitative entrepreneurial processes of
multinational firms operating in the global pharmaceutical industry. We analyze over 1,500
new drug approvals by the U.S. Food and Drug Administration (FDA). We find that a suc-
cessful track record in breakthrough innovation significantly increases the likelihood of a
current breakthrough, while achievements in nongeneric incremental innovation do not have
a significant effect. A strong foundation in generic incremental innovation hinders break-
through performance. Thus, incremental innovation processes appear to be heterogeneous.
Products that emerge from joint ventures and alliances are more likely to be breakthroughs.
Foreign subsidiary participation in innovation processes did not significantly inhibit break-
throughs. These suggestive findings support the decentralization literature that highlights the
benefits associated with exploiting knowledge from foreign centers of excellence. Contrary to
the literature arguing that younger firms tend to have greater advantages in exploration, we
do not find firm age to be a significant predictor of the likelihood of breakthrough innovation.
Copyright © 2010 Strategic Management Society.

THE CORPORATE innovation research is vast and there is a large body

ENTREPRENEURSHIP PERSPECTIVE
According to Snow (2007), innovation (i.e., the
introduction of a new product or service) and entre-
preneurship (i.e., the founding of a new business)
are virtually one and the same. The history of
Keywords: corporate entrepreneurship; breakthrough and
incremental innovation; open innovation; organizational
ambidexterity; foreign subsidiaries; strategic alliances
*Correspondence to: Denise Dunlap-Hinkler, College of
Business Administration, Northeastern University, 319
Hayden Hall, Boston, MA 02115-5000,
E-mail: d.dunlaphinkler@neu.edu
of literature on the subject. For many scholars,
Schumpeter is the seminal thinker for the literature
on innovation (Schumpeter, 1934, 1939, 1942). He
provided one of the first counterperspectives to the
mainstream viewpoint held by neoclassical econo-
mists and historians of science and history that inno-
vation is not a black box exogenous activity that
could be understood only after the serendipitous
event had occurred. While there is currently no dom-
inant theory on innovation, there is agreement that
it is a complex, difficult to measure construct
U.S.A. (Fischer, 2001; Tidd, 2001) that involves newness
to some degree to either the adopting unit (Rogers,

Copyright © 2010 Strategic Management Society

 Breakthrough versus Incremental Innovation
1983) or the marketplace, sector, or industry (Lawson
and Samson, 2001) and has a positive effect on firm
performance (e.g., Zahra, 1991; Zahra and Coven,
1995). According to DiMasi (2000), the impact on
performance can be profoundly long term. The phar-
maceutical firms in their study maintained relatively
stable leadership positions for innovative output
over a 35-year time span.
Scholars contend there is still much that is not
understood about how innovative opportunity is
created within the firm (e.g., Leiblein, 2007) and,
thus, researchers continue to study the nuances of
innovation from new perspectives. Accordingly, a
number of theoretical perspectives have been used
to examine firm innovation processes, including
cognitive theory, dynamic capabilities, institutional
theory, market orientation, resource-based view,
sociotechnical approaches, transaction cost econom-
ics, and so on. We integrate and review the litera-
tures from knowledge management, learning
organizational theory, and organizational ambidex-
terity to develop an integrative framework for cor-
porate entrepreneurship. We attempt to provide a
richer awareness of how international opportunity-
and advantage-seeking activities (Ireland, Hitt, and
Simon, 2003) affect the entrepreneurial outcomes of
breakthrough and incremental innovations.
Scholars of corporate entrepreneurship research
traditionally have focused on ways in which firms
can create positive changes within their organiza-
tions (Narayanan, Yang, and Zahra, 2009). Scholars
and practitioners have argued that corporate entre-
preneurs must not innovate just occasionally, but
often quickly and efficiently to ensure future revenue
growth generated from customers purchasing new
and improved products and services. Cooper (2007:
145) states that ‘the behavior of entrepreneurs and
the influences upon that behavior are clearly at the
heart of . . . entrepreneurship.’ To this end, academ-
ics have determined that there are core elements,
such as firm differences, competitive environment,
strategy, task complexity, and management style
that affect the entrepreneurial processes and
innovative outcomes across firms.
The two key phenomena that best define the pro-
cesses surrounding corporate entrepreneurship are:
(1) the birth of new businesses or internal venturing;
and (2) the transformation of organizations through
renewed patterns of resource deployment (Guth and
Ginsberg, 1990). Within the wider context of corpo-
rate entrepreneurship, corporate venturing focuses
on the first component of Guth and Ginsberg’s
Copyright © 2010 Strategic Management Society
107
(1990) definition, with added attention to the pro-
cesses that firms use both internally (i.e., new inno-
vative businesses) and externally (i.e., licensing and
strategic alliances) to create new opportunities
within existing firm portfolios. Most of the research
in this domain has focused on the parent organiza-
tion rather than the venture unit or the new venture
itself (Narayanan et al., 2009). We examine the
impact that the firm’s foreign subsidiary activities
(emanating from Greenfield operations versus
acquired units) and strategic alliance relationships
have on its entrepreneurial outcomes.
Corporate entrepreneurship research is becoming
increasingly important, as firms that were once not
thought of as being entrepreneurial must become so
if they want to prosper in the global marketplace.
With regard to the second component of Guth and
Ginsberg’s (1990) definition, an entrepreneurial firm
possessing this dominant logic must embrace the
uncertainty surrounding the innovation and diffu-
sion process (Kuratko and Audretsch, 2009) and, at
the same time, enhance their responsiveness to
change, increase their willingness to take risks, and
engage in innovative decision making (Phan et al.,
2009). One of the significant challenges facing
scholars in the field is that there needs to be a better
understanding of the heterogeneity of corporate
entrepreneurship activities (e.g., markets, products,
established versus start-ups) from a broader life
cycle perspective (Phan et al., 2009).
In this article, we study how corporate entrepre-
neurship changes over time by examining a total of
1,496 commercialized innovations (both break-
through and incremental) approved by the U.S. Food
and Drug Administration (FDA) for the years 1993
to 2002. We incorporate product innovations prior
to 1993 to capture past organizational learning. Our
sample consists of public firms and includes smaller,
start-up firms, medium-sized firms, and larger,
mature firms. The 98 firms in our study ranged from
being in their first year of business to being 163
years old and had net sales from as little as $0 to
$52.2 billion. Recognizing that firms’ home bases
may differ in terms of environmental inertia, insti-
tutional rigidities, and opportunities for organiza-
tional learning, we control for national country
differences for the major global pharmaceutical
players in our study (e.g., the U.S., the United
Kingdom, and Switzerland). Further, we recorded
the number of new efficacy supplements (i.e., this
refers to the means whereby firms exploit their exist-
ing innovation portfolios) generated from a firm’s
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
108 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
Breakthrough innovation
H3
H1 (-)
(+)
Acquired
Pre-existing stock
of breakthrough
subsidiaries
innovations
H1
(-)
H3
(+)
Figure 1. Conceptual model of breakthrough versus incremental innovation in the global pharmaceutical industry
core commercialized drugs (1992 to 2002) up until
the year 2008 for both public and private firms.
According to a recent report in Businessweek,
Mandel (2009: 34) finds that the last decade of
American innovative spirit has not been ‘an era of
rapid innovation . . . (but) . . . an era of innovation
interrupted . . . (leading to) . . . today’s financial
crisis.’ In the health care field alone, there have been
estimated to be far too many clinical test break-
through pipelines that were once promising but
ended up as failures (e.g., cancer treatments, cloning,
and, gene therapy) even with strong in-house global
R&D facilities. For instance, in 2007, Pfizer
announced plans to close five facilities (e.g., research,
manufacturing) that employed 10,000 workers due
to: (1) an insufficient number of breakthroughs in
the pipeline; and (2) a risk of losing almost 41
percent of its sales revenue to generic competition
following the loss of patent protection for its two
most profitable drugs, Zoloft and Norvasc (Smith,
2007). Not surprisingly, in 2007 the United States
Census Bureau, recorded a U.S. trade deficit of $53
billion in high-tech areas (such as life sciences) com-
pared to the $30 billion trade surplus it recorded in
1998 (Mandel, 2009).
We recognize that the announcement of such
novel innovations may not only push the firm out of
its established knowledge platform, but may also
have a profound impact on its ability to generate
significant future profits. Given the magnitude that
breakthrough innovations have on the economic
Copyright © 2010 Strategic Management Society
H4
(+)
Knowledge H2
sourcing (-)
H5
(+)
Pre-existing stock
Foreign of incremental
JVs and
alliances innovations
Greenfield H5 H2
(-) (+)
H4
(+)
Incremental innovation
growth of a nation and the long-term survival of a
firm in its industry, it is crucial to understand the key
factors that underpin successful corporate entrepre-
neurship. Our study focuses on the global pharma-
ceutical industry where the discovery, development,
and commercialization of new knowledge are par-
ticularly important for the delivery of innovative
new products to the marketplace. We develop an
integrative framework of corporate entrepreneurship
and study the firm’s ability to simultaneously manage
two very different types of innovation processes—
breakthrough (i.e., exploration of new innovations)
and incremental (i.e., exploitation of existing inno-
vations) (see Figure 1).
Organizational ambidexterity and
corporate entrepreneurship
One of the most critical insights of Nelson and
Winter’s seminal study (1982) was that they shed
light on how organizations and their deliberate prob-
lem-solving efforts to innovate define the introduc-
tion of newness to a society. Their approach
concentrated on historical technoeconomic change
and how the very basic attributes of newness con-
tinue to captivate the attention of consumers, corpo-
rate innovators, investors, business practitioners,
and management scholars. An important question
arises with regard to the definition of innovation and
newness. Scholars have suggested that breakthrough
and incremental innovation labels for both process
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
and product innovations are not necessarily tangible,
concrete categories; they specify the extremes of
the continuum of technological change. Typically,
breakthrough innovations start the cycle of techno-
logical change (e.g., polio vaccine, personal comput-
ers). Over time, incremental innovations in the form
of new features, extensions, variations, or comple-
ments to an existing product line (e.g., needleless
vaccine delivery systems, laptop computers) build
on the dominant designs created by breakthrough
innovations. Incremental process innovations fre-
quently involve innovations in production efficiency
and/or product quality (Kuratko and Welsch, 2001;
Tushman and Nadler, 1986; Utterback and
Abernathy, 1975). According to Anderson and
Tushman (1991), the length of time needed to estab-
lish a dominant design depends on whether the inno-
vation enhances or destroys current knowledge.
Typically, a longer time period is required for inno-
vations that destroy current industry know-how.
From a punctuated equilibrium perspective, tech-
nological progress evolves through long periods of
relatively minor changes punctuated with short, dra-
matic events of breakthrough change (Romanelli
and Tushman, 1994). These short periods of radical
change are referred to as technology discontinuities,
which create price and/or performance improve-
ments relative to current technologies (Tushman
and Anderson, 1986). According to Tushman and
O’Reilly (1996), periods of discontinuous change
are most likely to materialize when there are major
downtrends resulting from changes in the firm’s
external environment (e.g., regulatory, technologi-
cal, political, and economic changes). March’s
(1991) seminal insights suggest that the internal
tensions between these two organizational modes
of innovation learning activities—breakthrough
(i.e., exploration) versus incremental (i.e., exploita-
tion)—are distinctly and fundamentally different
from one another.
While the traditional exploration-exploitation
framework views breakthrough and incremental
innovations as being on two ends of the same con-
tinuum, scholars’ most recent focus has been on
firms that can and indeed thrive in dynamically
changing environments through simultaneously bal-
ancing and reconciling these complementary, yet
competing, tensions (e.g., Rothaermel and Deeds,
2004). Traditionally, Schumpeter and neo-
Schumpeterian scholars have highlighted how cre-
ative destruction is critical to earning above average
rents in the endogenous economic system. Others in
Copyright © 2010 Strategic Management Society
109
the literature have related creative destruction to
breakthrough inventions (Ahuja and Lampert, 2001),
breakthrough innovations (Mote, Boylan, and Rice,
2001), discontinuous innovations (Abernathy and
Utterback, 1978), pioneering products (Ali, 1994),
and disruptive innovations (Christensen, 1997).
Contemporary Schumpeterian scholars have given
closer attention to the theoretical and empirical com-
plexities associated with integrating both break-
through innovations or exploration activities and
incremental innovations or exploitation activities
inside the firm (Raisch and Birkinshaw, 2008).
Firms that are able to pursue exploitation and
exploration activities in parallel are more likely able
to achieve long-term gains.
By not becoming too trapped in a never ending
spiral of searching for breakthrough opportunities
emanating from potentially unrewarding trends in
the external environment (Levinthal and March,
1993; Tushman and O’Reilly, 1996), firms are able
to reap returns from existing knowledge capabilities
(Palmer and Brookes, 2002). Nokia Corporation, the
foremost mobile phone supplier is an example of a
firm that was able to successfully and simultane-
ously combine both exploration and exploitation
activities in a favorable way (Mudambi, 2008). The
company comprises highly vertically integrated
divisions—Nokia Networks and Nokia Mobile
Phones—that focus on incremental innovation. In
addition it has a separate Nokia Ventures organiza-
tion that can work with internal resources and alli-
ance partners on breakthrough innovations, free
from interference from the operating units. This
enables Nokia to ‘constantly explore new ways of
learning while simultaneously exploiting traditional
learning methods’ (Masalin, 2003: 68). To thwart
core rigidities and dependence on highly specific
settings (Leonard-Barton, 1992), firms need to
actively create both breakthrough and incremental
innovations to sustain competitive advantage. Our
model builds on the strategic alliance literature
that has found that firms are better able to manage
these paradoxical demands through partnering
(Rothaermel and Boeker, 2008; Rothaermel and
Deeds, 2004).
Firm size and corporate entrepreneurship
All firms engaging in entrepreneurship must pur-
posefully implement both opportunity-seeking and
advantage-seeking activities to create new wealth
and competitiveness (Ireland et al., 2003). Ketchen,
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
110 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
Ireland, and Snow (2007) find that the entrepreneur-
ial process continues to be mysterious and that some
firms, regardless of size, are more dependable in
their ability to produce high rates of innovations. So
far, a common theme in the literature is that firm size
is an important factor in the breakthrough innovation
development process. During the 1950s and 1960s,
an era dominated by economies of scale and big
labor unions, large firms were regarded not only as
having greater productive efficiency than small
firms, but also as being the engines of innovative
progress.
Schumpeter (1942) warned that innovation in
large corporations was being reduced to routine pro-
cedures in which teams of specialists were trained
to produce technological changes that followed pre-
dictable knowledge conditions. The dominant logic
of the day created a lack of internal flexibility, which
further created a complex system of self-sustaining,
reliable routines that were narrow in scope and did
not instigate organizational conflict (Mote et al.,
2001). These deliberate and rather mechanical sets
of organizational routines through bureaucratic
control structures failed to encourage the exploration
of radical breakthrough opportunities (Nelson and
Winter, 1982). Large firms, in essence, became
trapped by their familiarity, maturity, and need to
search for new incremental solutions to well-known
market domains (Ahuja and Lampert, 2001).
Since the mid-1970s, however, through the
dynamic process of deconcentration and decentral-
ization, the trend of creating larger enterprises has
been reversing. Average firm size has decreased and
the share of sales from entrepreneurial firms—
defined as smaller, younger, family-owned, and/or
new start-up businesses—has grown (Audretsch,
1995). The literature supports the notion that
although small firms are limited by their production,
marketing, financial, and human resources, they are
more likely to search for and produce more novel
innovations than larger firms (Almeida and Kogut,
1997; Hannan and Freeman, 1984; Rothaermel and
Boeker, 2008). While some suggest that fostering
innovation inside large firms cannot coexist within
most bureaucratic structures, others have focused on
describing how firms can develop successful corpo-
rate entrepreneurial processes that can exist within
routinized frameworks (e.g., Ahuja and Lampert,
2001).
The past few decades have witnessed an increas-
ing emphasis on traditional learning theories (Cyert
and March, 1963) that have attempted to understand
Copyright © 2010 Strategic Management Society
how organizations of all sizes can gain knowledge
from past innovative behaviors to stimulate future
entrepreneurial growth. Starting from Peterson and
Berger’s (1972) seminal work, researchers have
focused on identifying the organizational and envi-
ronmental factors that influence corporate entrepre-
neurship and innovation. The desire to pursue
innovative thinking inside corporate enterprises first
arose during the entrepreneurial economy of the
1980s (Kuratko and Welsch, 2001). Scholars sought
to understand the process that established firms used
when entering into new business ventures (e.g.,
Burgleman, 1983).
Earlier researchers, such as Miller (1983), estab-
lished the theoretical framework and research meth-
odology to examine the main linkages between
environmental, strategic, and organizational
variables and a firm’s entrepreneurial activities.
Management experts at that time suggested that
innovation was the vehicle through which entrepre-
neurs create the potential for wealth-producing
resources (e.g., Drucker, 1985). Gifford Pinchott
(1985) first developed the concept of intrapreneur-
ship to describe the entrepreneurial activities inside
large corporate structures. Currently, corporate
intrapreneurship is viewed in the literature as a nec-
essary component of corporate entrepreneurship and
refers to the firm’s change agents that challenge the
status quo and constantly initiate new innovative
thinking (Kuratko and Audretsch, 2009).
According to Birkinshaw (1997), corporate entre-
preneurial behavior that leads to the birth of new
businesses (Guth and Ginsberg, 1990) can best be
described as a focused entrepreneurial strategy.
Managers at 3M have successfully pursued this type
of strategy (Birkinshaw, 1997). By following a lead
user process strategy, the managers at 3M were able
to set forth a new strategic goal in the mid-1990s
that challenged the company to produce 30 percent
of its product line from products that did not exist
in the company’s portfolio of products four years
ago (von Hippel, Thomke, and Sonnack, 1999). Cor-
porate entrepreneurship requires management to
encourage all employees to pursue the development
of new opportunities. According to Phan et al., 2009,
corporate entrepreneurship succinctly refers to
renewal based on two phenomena: (1) corporate
venturing (Narayanan et al., 2009); and (2) strategic
entrepreneurship (Kuratko and Audretsch, 2009).
Given that today’s high-tech economy supports a
greater number of sophisticated competitors than
ever before, managers recognize that they can no
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
longer derive value by focusing on known strategic
competencies (Vale and Addison, 2002). To cope
successfully within these ever-changing environ-
ments, firms must be open-minded enough to con-
tinually search for new avenues to develop new
products for new markets (Ireland, Covin, and
Kuratko, 2009; Kuratko and Audretsch, 2009). In
many cases, directing a flow of entrepreneurial
events—defined as any unplanned combination of
economic resources initiated by the vague outlook
of temporary monopoly profits (Binks and Vale,
1990)—within the firm is being viewed as a strategic
competency. According to Phan et al. (2009), it is
important to note that small and large firms face the
same risks as they attempt to exploit and create new
knowledge. From a learning organizational perspec-
tive, large and small firms engaged in corporate
entrepreneurship often find it advantageous to join
forces and share complementary opportunity-
seeking and advantage-seeking knowledge compe-
tencies. Specifically, large firms may have to think
small and small firms may have the think big to
instigate real change (Ireland et al., 2009).
THEORY DEVELOPMENT
AND HYPOTHESES
We have argued that corporate entrepreneurs, regard-
less of size, need to actively and simultaneously
pursue exploitative (i.e., incremental) and explor-
ative (i.e., breakthrough) innovation opportunities
that meet market needs. Now, we integrate these
concepts from a knowledge management and orga-
nizational learning perspective and formulate our
hypotheses.
The role of accumulated past knowledge in
corporate entrepreneurship
From a learning perspective, organizational routines
typically guide and constrain processes by which
new knowledge is assimilated and organized (Nelson
and Winter, 1982). Routines that lead to market-
based objectives create path-dependent future
actions, whereas projects that do not materialize as
anticipated are quickly eliminated or temporarily
shelved (Ahuja and Lampert, 2001). The firm’s
ability to apply the lessons learned during these
types of experiences makes a difference in its speed
and integrative innovative abilities over another firm
facing a similar situation. Put differently, learning
Copyright © 2010 Strategic Management Society
111
builds upon existing knowledge domains and other
familiar assets, and these influences affect the direc-
tion of future entrepreneurial behaviors of firms
(e.g., Nerkar, 2003; Teece, 1988).
Extensive past experience with a particular knowl-
edge or innovation practice may also result in greater
inertia for change or learning. In the case of strong
in-house R&D facilities, particularly among large
firms (e.g., Fischer, 2001), established mental
models geared toward incremental solutions tend to
be the norm. The strategy literature on learning sug-
gests that while innovation is fostered by diversity
in experience, repeated spirals of competition and
cooperation within a familiar setting can also lead to
blindness to new opportunities and threats that tran-
scend specific settings (Levinthal and March, 1993).
Within this context, newer innovation models may
be viewed as less attractive, and such projects may
have lower rates of acceptance by entrepreneurial
decision makers.
We argue that due to the increasing global pres-
sures being placed on firms, those investing in
searches for opportunities that leverage past knowl-
edge increase their risk of being locked out of acquir-
ing and investing in newer breakthrough technologies
(Narula, 2002). Fewer accumulated breakthrough
knowledge experiences, and even breakthrough fail-
ures, may further create a familiarity trap and
prevent firms from sensing opportunities beyond
their typical knowledge domain (e.g., Ahuja and
Lampert, 2001; Hayward, 2002). It is our contention
that increased access to successfully accumulated
breakthrough endeavors will expand the firm’s pool
of valuable, complementary resources from which it
can search for solutions to new technological pro-
cesses (Levitt and March, 1988).
Reduced association with past incremental inno-
vation successes may also create a tolerance for
ambiguity and an appreciation of knowledge com-
plexity. This tolerance is particularly important for
the development of breakthrough capabilities where
the firm’s core technologies are already pushed into
new, uncharted territories. Ahuja and Lampert
(2001) found that firms that were able to experiment
with novel, emerging, and pioneering technologies
were more successfully able to overcome familiarity
traps that typically inhibited breakthrough practices
in the past. According to Nerkar (2003), researchers
should measure the impact of successfully commer-
cialized products to better understand current
research on explorative and exploitive activities.
To this end, we argue that prior experience with
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
112 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
successfully commercialized breakthroughs is an
internal resource that is valuable, unique and
difficult to imitate and, thus enhances the firm’s
competitive advantage (Barney, 1991). Hence, we
hypothesize:
Hypothesis 1: The greater the firm’s pre-existing
stock of breakthrough innovations, the greater the
likelihood that the current innovation will be
a breakthrough rather than an incremental
innovation.
Hypothesis 2: The greater the firm’s pre-existing
stock of incremental innovations, the lower the
likelihood that the current innovation will be
a breakthrough rather than an incremental
innovation.
The role of foreign subsidiaries in
corporate entrepreneurship
A firm’s production of knowledge constitutes
a resource underpinning sustainable competitive
advantage (Barney, 1991). It has been suggested that
knowledge from a particular national location
embodies more than its national culture, but rather
its entire national system of innovation (NSI)
(Lundvall, 2007; Mudambi, 2008). Researchers
have found that NSIs, although seemingly resistant
to changes, do change, but in a manner that is slower
than what firms need in terms of new technological
requirements. For instance, Narula (2002) found that
larger, more traditional, resource-intensive sectors
that were highly embedded within Norway’s NSI
benefited most from the close-knit relationships
developed nearby their home location of compe-
tence. By virtue of having research facilities in
certain geographic regions, foreign subsidiaries can
be used to tap into unique local technological
capabilities.
The literature on geographic knowledge sourcing
of innovation in the context of the parent-foreign
subsidiary dyad reveals two different models—
centralization and decentralization. A centralization
strategy is more of a learning-by-doing process that
demands greater specialization (Cohen and Klepper,
1996). Scale economies in R&D are achieved in
single geographic locations, preferably within a
regionally or nationally concentrated knowledge
cluster. Empirical evidence supports the notion that
knowledge spillovers tend to occur in geographi-
cally localized clusters (e.g., Jaffe, Trajtenberg, and
Copyright © 2010 Strategic Management Society
Henderson, 1993). There are both demand-driven
(i.e., close interaction with customers) and supply-
side (i.e., spillover advantages) reasons why geo-
graphical proximity at a single location is preferred.
Sakakibara and Porter (2001) found that intense
domestic rivalry among a sample of Japanese indus-
tries was positively associated with international
trade performance. The extent to which the R&D
efforts by MNEs are undertaken in their home
country depends on the nature of the industry in
which it operates, the incentives of the host country,
the absolute size of the home country, the nature of
knowledge itself, and the extent of foreign competi-
tion (Cantwell and Mudambi, 2005).
The counter argument to this philosophy is that
centralization leads to risks of lock-in into techno-
logical and institutional systems of innovations that
are self-reinforcing and not always efficient (Narula,
2002; Redding, 2002). The decentralization concept
of managing, developing, and exploiting global
knowledge from strategically advantageous centers
of excellence is not new. Other related concepts
include Bartlett and Ghoshal’s (1989) transnational
organization, Prahalad and Doz’s (1987) multifocal
organization, Hedlund’s (1986) heterarchy, Cantwell
and Mudambi’s (2005) competence-creating subsid-
iaries, and Perlmutter’s (1969) geocentric organiza-
tion. According to De Meyer (1992), a tightly
coordinated, yet decentralized strategy allows firms:
(1) the ability to reap the benefits of cost differentials
in different countries (i.e., traditional neoclassical
economic theory); (2) the capacity to become more
responsive to markets (i.e., international product life
cycle theory) (Vernon, 1966); and (3) the potential
to reduce difficulties with technological transfer of
know-how through networks (i.e., transaction cost
approach) (Teece, 1981).
Frost, Birkinshaw, and Ensign (2002) found that
MNEs need to reassess the innovative role played
by foreign subsidiaries. A survey of the literature
reveals that there are a limited number of frame-
works that examine how the innovation function of
foreign subsidiaries affects firm outcomes (Frost,
2001; Frost et al., 2002). Most frameworks have
characterized the knowledge and financial linkages
between the MNE parent, its subsidiaries (Asakawa,
2001), and most recently its subsidiary host sites
(Cantwell and Mudambi, 2005; Mudambi, 2008).
Current research suggests that firms are gradually
allowing their foreign R&D subsidiaries to take
on more specialized roles based on the compara-
tive advantages of the subsidiary’s location. This
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
perspective asserts that when subsidiaries take on
important roles such as centers of excellence,
intrafirm knowledge flows are greater (e.g., Frost,
2001; Frost et al., 2002).
Notwithstanding, decentralization strategies are
complicated, costly, and require a great deal of
expertise. Further, foreign subsidiaries that under-
take knowledge-intensive activities have been found
to be difficult for headquarters to control (Mudambi
and Navarra, 2004). This may be why some studies
report an overall insignificant flow of cross-border
knowledge sharing. For instance, by using patent
statistics as a proxy for measuring the world’s largest
firms in different countries and sectors, researchers
have determined that the technological production
activities of firms are far from globalized (e.g., Jaffe
et al., 1993; Patel and Pavitt, 1991). Most recently,
Kelley, Peters, and Colarelli O’Connor (2009)
reported that a firm’s knowledge base is theoreti-
cally likely to be enhanced when organizational
members share similar, but uniquely different, and
diverse knowledge structures. In reality, however, it
is difficult to recruit global organizational members
that want to participate in high-risk projects filled
with uncertainty, unfamiliarity, and potential failure.
Put simply, the authors found that organizational
members were more cautious to be seen as failing
on a global rather than on a local platform and, thus,
were less willing to collaborate with others, particu-
larly on breakthrough innovation projects involving
organizational legitimacy concerns.
Building on this logic, we argue that foreign sub-
sidiaries have their own individual characteristics
and unique ties with their parent organizations. We
contend that even though the decentralization of
subsidiary roles is increasing and firms are tapping
into NSIs for knowledge spillovers, subsidiaries are
still required to undertake traditional support func-
tions with central oversight of their R&D budgets.
According to Aldrich and Kim (2007), people within
these environments are likely to be constrained by
small world networks. Thus, it can be difficult for
subsidiaries to secure the kind of long-term support
and financial resources necessary for the exploration
of technological breakthroughs.
The path to developing novel products within the
global pharmaceutical industry is a long and expen-
sive process due to strict governmental regulations
which forbid firms from marketing new product
developments without approval from the relevant
government agency (i.e., the Food and Drug Admin-
istration in the U.S.). For instance, it can cost
Copyright © 2010 Strategic Management Society
113
upwards of $800 million to $1 billion and 10 to 15
years of clinical research to bring a new drug to the
market. Moreover, less than one percent of drugs
tested will end up being used by patients and of
those drugs that are approved, only 30 percent will
recover their R&D expenses (DiMasi, Hansen, and
Grabowski, 2003). With these seemingly insur-
mountable costs, we argue that foreign centers of
excellence from Greenfield operations will be more
reluctant to engage in and be associated with high-
risk projects aimed at breakthrough innovation. We
suggest that when firms increase their heterogeneity
of collaborative efforts outside of their home-base
system of innovation, they are more likely to open
up the doorway to exploit technology trajectories
that are predictable and slow down their existing
product life cycles or s-curves (Rogers, 1983). From
this rationale, we suggest the following:
Hypothesis 3: Innovations emanating from foreign
subsidiaries are less likely to be breakthrough
rather than incremental innovations.
Hypothesis 4: Innovations emanating from Green-
field operations are less likely to be breakthrough
rather than incremental innovations.
The role of joint ventures and strategic alliances
in corporate entrepreneurship
By engaging in acquisition and alliance strategies,
firms increase their scope of future learning possi-
bilities. The idea that creativity, synergies, and new
ideas come from the interaction and recombination
of these potentially conflicting knowledge sets is
well accepted in the knowledge literature (Simon,
1985). In fact, knowledge that is not regularly
renewed can lead to core rigidities in technological
innovation advancement (Leonard-Barton, 1992) so
that ‘firms are compelled to augment their R&D
capacity by collaborating and sourcing-in . . . dis-
coveries, inventions, and innovations from other
players and institutions’ (Markman, Siegel, and
Wright, 2008: 1401).
Due to the ongoing and growing importance of
partnerships, joint ventures, alliances, and other
forms of collaborative learning experiences, Lane
and Lubatkin (1998) refined Cohen and Levinthal’s
(1990) seminal absorptive capacity construct to
examine the interorganizational learning aspects
that occur in knowledge-sharing alliances. As
firms participate in these alliances, they face unique
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
114 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
challenges. Moreover, it is argued that alliances
have to benefit both partners if they are to ‘reduce
risk and facilitate knowledge transfer’ (Cumming
and Macintosh, 2000: 360). If alliance partners are
unwilling to devote energies into resource combina-
tion, then the empirical benefit from such knowledge
sharing is limited (Wiklund and Shepherd, 2009). In
many cases, some firms may naturally take on the
role of student firm, while other firms may take on
the role of teacher firm. Interorganizational and
intraorganizational learning processes have been
found to be positively enhanced when student-
teacher learning dyads shared similar basic knowl-
edge bases, compensation practices, and research
communities (Lane and Lubatkin, 1998; Yang,
Mudambi, and Meyer, 2008).
While it is important that alliance dyads share
basic experiences, it is equally important that their
knowledge experiences do not become so similar to
one another that there is little room for creativity,
collaboration, continuous learning, and most impor-
tantly, the exploration of breakthrough opportuni-
ties. In other words, even though alliance
characteristics (e.g., sizes and structure) vary greatly
across firms and industries, firms entering into them
should continually have something new to learn
from each other. The joining of diverse, yet comple-
mentary, skill sets is an essential and necessary com-
ponent of technological innovation (Cohen and
Levinthal, 1990; March 1991; Kotabe and Swan,
1995; Rosenkopf and Schilling, 2007). In many
cases, new and established firms benefit most from
alliance partnerships. Rothaermel and Boeker (2008)
found that biotechnology and pharmaceutical firms
entered into alliances when the biotechnology firm
was younger. Cultural factors also matter according
to Coombs, Mudambi, and Deeds (2006). They
found that foreign firms, unlike domestic firms,
entered into alliances to access local knowledge
embedded in dominant regional clusters. Their
results suggest that foreign firms value location-
specific knowledge because they are not only able
to have access to direct types of knowledge, but they
are also able to tap into indirect types of knowledge
(i.e., this involves specialized resources and infor-
mation available only through network member-
ship). Finally, the literature on open innovation
provides strong evidence of the value of recombin-
ing diverse knowledge resources (Laursen and
Salter, 2006; von Hippel, 1998). Given these posi-
tive and convincing arguments, we hypothesize
the following:
Copyright © 2010 Strategic Management Society
Hypothesis 5: Innovations emanating from joint
ventures and strategic alliances are more likely
to be breakthrough rather than incremental
innovations.
METHODOLOGY
Data
Our objective in this study was to examine empirical
data on corporate entrepreneurs that are attempting
to balance the competing demands associated with
developing breakthrough and incremental product
innovations. We gathered data from the global phar-
maceutical industry since it is an industry that is both
science- and technologically-intensive and has a
high propensity to create both types of new product
innovations. According to a recent IMS Health
Report (2007), worldwide sales in the industry were
approximately $643 billion dollars in 2006. Our
initial sample was drawn from a population of U.S.-
based multinationals and foreign multinationals
operating in the United States that received approval
from the U.S. Food and Drug Administration’s
(FDA) Center for Drug Evaluation and Research
(CDER) that their new products were safe and effec-
tive for marketing under U.S. law.
The FDA’s new drug application (NDA) is a
lengthy process. It involves extensive analysis of
clinical testing methods as well as reviews of label-
ing information and manufacturing methods. Appli-
cations are nearly 100,000 pages long and take about
two years to be processed; average testing periods
are more than eight years. Receiving FDA approval
for the marketing of a new drug often has a funda-
mental impact on the firm’s growth and profitability.
For instance, investors who had a strong financial
stake in the company Bristol-Myers Squibb received
a 15 percent decrease in their stock values when the
FDA refused to review the firm’s breakthrough
cancer treatment candidate, Erbitux, in which the
firm had invested more than $1 billion (Sellers,
2002).
We limited our sample to product innovations that
occurred over the time period of 1992 to 2002. This
sample comprised a total of 2,885 new drug applica-
tions (n = 2,590 incrementals, n = 295 breakthroughs)
from both private and public firms. For our final
analysis, private companies, companies that only
created abbreviated new drug applications (i.e.,
generics), and firms that did not create innovations
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
after 1992 were excluded. This left a total of 1,789
new drug applications in our dataset. Due to missing
financial and other construct measures, our final
usable samples (e.g., depending on model specifica-
tion) comprised from 1,496 to 1,699 new drug appli-
cations from 98 firms.
Companies in the pharmaceutical industry exploit
their new product innovations through new efficacy
supplements (NESs), which represent a labeling
modification to the original new drug application.
Up until the year 2008, we recorded 16,684 NESs
to our original sample of 2,885 new drugs. FDA
review of these applications takes approximately six
to 12 months. NESs are marketed as new innova-
tions proposing to add a new usage to an already
approved drug or product application. No matter
how many supplements are created, these innova-
tions still maintain their original new product appli-
cation number.
Constructing measures
Breakthrough and incremental innovations
The FDA has a detailed evaluation system to deter-
mine whether or not a new drug is new to the market
or a product improvement. Thus, we believe that our
measure is appropriate for the organizational ambi-
dexterity focus developed in our article. Product
innovations were coded as 1 for breakthrough and 0
for incremental. Drugs that the FDA defines as new
molecular entities (NMEs) are new active ingredi-
ents that have never been marketed before in any
other form in the United States. We recorded these
products as breakthrough innovations. We classified
incremental innovations as: (1) new drug applica-
tions (NDAs) chemical types that have been mar-
keted in the U.S. before in one form or another (i.e.,
this is a nongeneric new chemical dosage form)
according to FDA standards; and (2) all abbreviated
new drug applications (ANDAs) or generics that
are bioequivalent to previously approved NDAs.
Further, we gathered data on new efficacy supple-
ment applications (NESs) up until 2008 for all origi-
nal new drug applications from the years 1992 to
2002. We computed the number of NESs per year
since drugs developed in 1992 are more likely to
accumulate more NESs than drugs developed in
2002. We classified NESs as another exploitative
and incremental form of corporate entrepreneurship
(see Discussion section).
Copyright © 2010 Strategic Management Society
115
Prior accumulated knowledge
This construct was measured as the stock of break-
through (i.e., prior breakthrough knowledge) and
incremental (i.e., prior incremental knowledge)
product successes that the firm had in 1992, the year
before the period of analysis. For incremental inno-
vations, we further categorized between nongeneric
versus generic forms of innovation. While both
product categories involve some form of minor tech-
nological change to an existing knowledge platform,
theorists, scientists, and the FDA draw a distinction
between them. Nongenerics, unlike generics are
generally protected by patents. Once a patent has
expired, a generic alternative can be sold. Generic
innovations often radically increase price competi-
tion among firms and the number of choices avail-
able to consumers in a product class category.
Foreign subsidiary unit
An essential aspect of this study was to characterize
how foreign subsidiary knowledge impacts corpo-
rate entrepreneurship. To accomplish this objective,
we initially recorded the name of the firm for each
FDA approved drug application. We further deter-
mined whether or not the firm was a subsidiary or
the parent firm. Based on information from CHI
Research Inc.’s Tech-Line Products of Companies
and extensive additional public records (i.e., these
are 10-K reports, press releases, Mergent, Who
Owns Whom, etc.) we were able to properly unify
parent-subsidiary relationships. We recorded the
two-digit country code for both parent and subsid-
iary firm. Foreign subsidiaries were coded as 1 and
domestic subsidiaries were coded as 0.
Greenfield operations
This construct was specified by identifying, through
public records (i.e., these are 10-K reports, press
releases, Mergent, Who Owns Whom, etc.), whether
or not the subsidiary was a Greenfield operation
(coded as 1) or acquired (coded as 0) up until the
year 2002.
Joint venture and strategic alliances
The degree to which the parent firm relied on a joint
venture or other forms of strategic alliances was
measured by analyzing the self-report information
on the FDA new drug application. If more than
two companies were responsible for a NDA, then
this information was listed on the original FDA
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
116 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
application. Through public records (i.e., these are
10-K reports, press releases, Mergent, etc.), we
recorded whether or not specific drugs were the cre-
ation of joint ventures and/or other alliance agree-
ments. We coded 1 for JVs and alliances and 0
otherwise.
Firm-level controls
As discussed earlier, firm age and firm size—both
continuous variables in our study—are important
influences on corporate entrepreneurship. Thus, we
needed to control for these influences. We gathered
data at the parent firm level. First, we collected infor-
mation on the firm’s date of incorporation from
Mergent. We subtracted this date from the year in
which a new product was developed. Firms in our
study ranged from being in their first year of business
to being over 150 years old. Financial data for firm
net sales and firm R&D intensity were gathered from
10-K reports, Mergent, Edgar, and Osiris. Firm size
was measured as the firm’s net sales per year (i.e., in
thousands of U.S. dollars). A large body of empirical
work has examined the relationship between R&D
resources and firm performance. It is well established
that increasing R&D intensity positively affects firm
outcomes (e.g., DeSanctis, Glass, and Ensing, 2002).
This is particularly evident in the pharmaceutical
industry where R&D is vital to firm performance.
R&D intensity was measured as a ratio of R&D
expenditures by the firm to its total net sales.
National system of innovation (NSI)
Given the current research on NSIs, we expected
that a parent firm’s NSI (or its home-base headquar-
ters location) to have a significant influence on its
corporate entrepreneurial behavior. We defined a
firm’s NSI as the country in which the firm is head-
quartered. Firms were coded based on internation-
ally recognized two-digit codes. We generated three
primary NSI dummy variables for the firms in this
study: the United States (U.S.), the United Kingdom
(U.K.), and Switzerland (CH). Firms from less rep-
resented countries and territories, such as Belgium
(BE), Bermuda (BM), Canada (CA), Denmark (DK),
France (FR), Finland (FI), Germany (DE), Ireland
(IE), Israel (IL), Japan (JP), and the Netherlands
(NL), were coded as other or 0.
Industry controls
The global pharmaceutical industry was selected as
the primary focus of this study. For each parent firm,
Copyright © 2010 Strategic Management Society
we gathered information on their primary standard
industry classification code (SIC) from Mergent. We
classified firms according to their four-digit SIC code.
Mainstream pharmaceutical firms (coded as 1) had
the following SIC codes: 2833 (i.e., medicinal chemi-
cals), 2834 (i.e., pharmaceutical preparations), 2835
(i.e., in vitro diagnostic substances), and 5122 (i.e.,
drugs and drug proprietaries, and druggists’ sundries).
Nonpharmaceutical-related firms (coded as 0)
were from the following SIC code classifications:
2836 (i.e., biological products), 2841 (i.e., soap),
2844 (i.e., toiletries, perfumes), 2869 (i.e., organic
chemicals), 2874 (i.e., fertilizer), 2891 (i.e., adhe-
sives), 3563 (i.e., gas compressor), 3841 (i.e., surgi-
cal), 3851 (i.e., ophthalmic), 8733 (i.e., noncommercial
research), etc.
Product controls
We controlled for whether or not the new product
was a prescription (i.e., RX was coded as 1) or an
over-the-counter (i.e., OTC was coded as 0) drug.
Further, in this research, we controlled for patent
knowledge (Jaffe et al., 1993) since patents encom-
pass know-how into the inner workings of product
and process technologies (Kogut and Zander, 1993).
The appropriateness of utilizing patent data as a
measure of innovation is well recognized in the lit-
erature (Cantwell, 1989). The pharmaceutical indus-
try relies heavily on patent protection and has a high
propensity to patent. Data on 1,273 product patents
for new drug approvals were initially obtained from
CDER’s Electronic Orange Book. Excluding private
firms, products created before 1992, and products
with missing values, our final sample included a
total of 1,208 patent counts used to protect 443 of
the 1,496 products. Firms in our study used as many
as 14 patents and as few as zero patents to protect
their new drug or product innovations.
ESTIMATION AND RESULTS
We adopt a discrete probabilistic approach to esti-
mating the likelihood associated with breakthrough
versus incremental innovation. Each product inno-
vation is categorized in a bivariate manner as either
a breakthrough or not. This generates a discrete
limited dependent variable that we estimate using
unconditional maximum likelihood logit. We chose
this estimation method since it is less sensitive to
assumptions regarding the error distribution and,
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation 117

Table 1. Summary statistics

VARIABLES Mean S.D. N

Dependent variable
Breakthrough innovation 0.134 0.3409 1,789
Control variables
Firm age (years) 48.589 36.7611 1,789
Firm size (sales, $’000) 6,780.984 9,454.0147 1,789
R&D intensity 24.779 6.1641 1,789
HQ location: U.S. 0.579 0.4936 1,789
HQ location: U.K. 0.065 0.2457 1,789
HQ location: CH 0.089 0.2846 1,789
Industry dummy(1) 0.908 0.2895 1,789
Product type(2) 0.869 0.1743 1,786
Patent count 0.701 1.6459 1,723
Prior innovations-total(3) 4.785 6.3109 1,789
Hypothesized variables
Prior breakthrough(3) 0.484 1.0091 1,789
Prior nongeneric(3) 1.367 2.5511 1,789
Prior generic(3) 2.935 3.5635 1,789
Foreign unit 0.337 0.4729 1,788
Greenfield operation 0.463 0.4988 1,788
JV, alliance, etc. 0.052 0.0499 1,788

Notes: (1) If pharmaceuticals industry (SIC = 2833, etc.) = 1; else = 0;

(2) If prescription drug, product type = 1; else = 0;
(3) Prior innovations = stock of past firm innovative activities.

therefore, likely to produce more robust results
(Aldrich and Nelson, 1984). Since our sample
includes data on 98 firms over the period 1992 to
2002, we estimated the models using firm-fixed
effects with time (i.e., year) dummies.
The fixed effects logit model is consistent in the
number of units (i.e., the bias tends to zero as
the number of units is arbitrarily increased)
(Chamberlain, 1980). Further, it has been shown that
the bias associated with the unconditional fixed effects
logit model tends to become insignificant when the
number of units exceeds 16 (Katz, 2001) or 20
(Greene, 2004). Thus, the unconditional maximum
likelihood fixed effects logit model is appropriate
given the large number of firms in our sample. None-
theless, we estimate the model using conditional
maximum likelihood logit as a robustness check.
None of our results are qualitatively affected.
We estimate the likelihood that a particular inno-
vation will be a breakthrough innovation based on
firm, industry, and location level variables. Con-
versely, the estimation provides the likelihood that
an innovation will not be a breakthrough and, there-
fore, will be an incremental innovation. The summary
Copyright © 2010 Strategic Management Society
statistics related to all the variables used in the anal-
ysis are presented in Table 1. The correlations of all
these variables appear in Table 2.
We first estimate a set of models (Models 1, 2,
and 3) using only our basic control variables. These
results are presented in Table 3. We find reassuring
parameter stability across the three models, a finding
that is supportive of robustness. The likelihood
ratios indicate that the models are all highly signifi-
cant. Furthermore, all the specifications have in-
sample predictive accuracies of greater than 85
percent, a very high rate of success (Graubard and
Korn, 1999).
We find that firm size is consistently a significant
covariate of breakthrough innovation and, therefore,
is an important control. R&D intensity is also a
significant covariate in Models 1 and 2. Comparing
Models 1 and 2 with Model 3, we find that when we
introduce the patent count associated with the
product innovation, R&D intensity is no longer sig-
nificant. This can be interpreted as evidence in
support of the process view of innovation (Basberg,
1987), wherein R&D is an input into the patenting
process which, in turn, is an input in the innovation
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 118

Table 2. Correlation matrix and descriptive statistics

VARIABLES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Copyright © 2010 Strategic Management Society

1. Breakthroughs 1
2. Firm age (years) 0.076 1
3. Size (sales, $’000) 0.168 0.563 1
4. R&D intensity −0.00 −0.06 −0.03 1
5. HQ location: U.S. −0.06 −0.21 −0.17 0.029 1
6. HQ location: U.K. 0.136 −0.11 0.138 −0.01 −0.31 1
7. HQ location: CH −0.01 0.406 0.519 −0.01 −0.41 −0.10 1
8. Industry dummy(1) 0.040 −0.16 0.110 0.011 −0.16 0.062 0.084 1
9. Product type(2) 0.062 −0.14 −0.17 0.169 −0.05 0.011 −0.02 −0.01 1
10. Patent count 0.158 0.137 0.212 0.247 0.01 0.234 −0.02 0.055 −0.06 1
11. Prior total innov.(3) 0.200 0.416 0.467 0.226 0.046 0.413 0.013 0.086 −0.028 0.326 1
12. Prior breakthrough(3) 0.215 0.290 0.443 0.193 −0.13 0.617 0.054 0.141 −0.04 0.310 0.810 1
13. Prior nongeneric(3) 0.173 0.423 0.427 0.215 0.112 0.287 −0.00 0.055 −0.02 0.299 0.972 0.648 1
14. Prior generic(3) −0.22 0.106 −0.00 −0.04 −0.06 −0.23 0.422 0.251 0.096 −0.19 −0.12 −0.11 −0.11 1
D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi

15. Foreign unit −0.02 0.099 −0.07 −0.02 −0.33 −0.05 0.105 0.096 0.047 −0.04 −0.23 −0.23 −0.21 −0.01 1
16. Greenfield operation 0.033 0.039 0.027 0.039 0.409 −0.16 −0.12 −0.17 −0.12 0.009 0.036 −0.10 0.087 −0.18 −0.31 1
17. JV, alliance, etc. 0.054 −0.06 −0.03 −0.04 −0.17 0.163 0.130 0.159 0.111 −0.05 −0.05 0.093 −0.10 0.189 0.280 −0.97 1

Notes: (1) If pharmaceuticals industry (SIC = 2833, 2834, 2835, 5122) = 1; else = 0;
(2) If prescription drug, product type = 1; else = 0;
(3) Prior innovations = stock of past firm innovative activities.

DOI: 10.1002/sej
Strat. Entrepreneurship J., 4: 106–127 (2010)
 Breakthrough versus Incremental Innovation 119

Table 3. Estimation of the likelihood of breakthrough innovations: Base-line control models (1993–2002) and maximum
likelihood logit results with firm fixed effects(1)

Regressand: Breakthrough innovation = 1 if FDA new drug approval is a new molecular entity; otherwise = 0
(incremental innovation).

REGRESSOR Coefficient (‘t’ stat(4))

MODEL 1 MODEL 2 MODEL 3

CONSTANT −2.27 (4.89)*** −4.14 (4.08)*** −5.51 (4.23)***

Controls
Firm age (years) 1.4 × 10−3 (0.74) 2.4 × 10−3 (1.23) 2.1 × 10−3 (0.91)
Firm size (sales) 4.5 × 10−8 (5.61)*** 5.0 × 10−8 (5.09)*** 4.8 × 10−8 (4.91)***
R&D intensity 0.03 (2.71)*** 0.05 (2.54)** 8.5 × 10−3 (0.61)
HQ location: U.S. −0.20 (1.17) −0.16 (1.34) −0.28 (1.81)*
HQ location: U.K. 0.68 (2.49)** 0.72 (2.51)** 0.32 (1.26)
HQ location: CH −0.93 (3.31)*** −1.12 (3.02)*** −1.23 (3.22)***
Industry dummy(2) −0.25 (0.81) −0.32 (0.79) 0.38 (0.94)
Product type(3) — 2.81 (2.82)*** 2.95 (2.83)***
Patent count — — 0.27 (4.26)***
Diagnostics
N 1,699 1,696 1,498
Log-likelihood −590.9231 −582.7721 −476.3685
Restricted log-likelihood; −675.4641 −675.0252 −573.5764
d.f. 115 116 117
Likelihood ratio χ2(d.f.); (‘p’ value) 169.082 (0.001) 184.5062 (0.000) 194.4158 (0.000)
Prediction ratio 0.8915 0.9014 0.9128

Notes: (1) Time dummies included;

(2) If pharmaceuticals industry (SIC = 2833, 2834, 2835, 5122) = 1; else = 0;
(3) If prescription drug, product type = 1; else = 0;
(4) ‘t’ statistics computed using White’s heteroskedasticity-consistent variance-covariance matrix.
***‘t’ statistics significant at the 0.001 level;
**‘t’ statistics significant at the 0.05 level;
*‘t’ statistics significant at the 0.10 level.

process (Ernst, 2001; Macdonald, 2004). Products
aimed at the prescription drug market are signifi-
cantly more likely to be breakthrough innovations.
We obtain some significant results related to U.K.
and Swiss headquarters location, but these stem
from the firms’ prior innovation profiles and dis-
appear in the full models.
Estimates of our hypothesized models are pre-
sented in Table 4. We first present results for an
exploratory specification (Model 4) in which we
include total prior innovations as a regressor. This
regressor emerges as highly significant, indicating
preliminary support for our thesis of path depen-
dency in innovation performance. Next, we estimate
a specification where we unpack prior innovations
into breakthrough and incremental categories. Prior
incremental innovations are further disaggregated
Copyright © 2010 Strategic Management Society
into nongeneric and generic innovations in these
specifications (Model 5).
Prior breakthrough innovations are highly signifi-
cant and raise the likelihood that the current innova-
tion is a breakthrough. This result provides evidence
in support of Hypothesis 1. Prior nongeneric incre-
mental innovations are not significant. However, a
larger stock of prior generic innovations signifi-
cantly reduces the likelihood that the current innova-
tion will be a breakthrough and correspondingly
increases the likelihood that it is incremental. There-
fore, the support for Hypothesis 2 is mixed; the
nature of prior incremental innovation seems to
matter. In our final and most detailed specification
(Model 6), we were left with 1,496 product innova-
tions. In this model, we include the location
and nature of the innovating unit. We find that
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
120 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi

Table 4. Estimation of the likelihood of breakthroughs: hypothesized models, full controls (1993–2002)—Maximum
likelihood logit results with firm fixed effects(1)

Regressand: Breakthrough innovation = 1 if FDA new drug approval is a new molecular entity; otherwise = 0
(incremental innovation).

REGRESSOR Coefficient (‘t’ stat(5))

MODEL 4 MODEL 5 MODEL 6

CONSTANT −4.34 (3.78)*** −3.72 (4.36)*** −5.09 (4.31)***

Controls
Firm age (years) −1.1 × 10−3 (0.52) 1.9 × 10−3 (0.83) 1.7 × 10−3 (0.54)
Firm size (sales) 4.9 × 10−8 (4.42)*** 3.1 × 10−8 (2.70)*** 2.1 × 10−8 (2.22)**
R&D intensity 4.6 × 10−3 (0.61) 7.4 × 10−3 (0.34) 1.9 × 10−3 (0.14)
HQ location: U.S. −0.41 (2.41)** −0.14 (0.62) −0.54 (1.43)
HQ location: U.K. −0.33 (0.73) −0.38 (1.12) −0.72 (1.61)
HQ location: CH −0.92 (2.91)*** 0.49 (0.58) 0.38 (0.92)
Industry dummy(2) 0.12 (0.17) 1.14 (1.96)* 1.17 (2.17)**
Product type(3) 2.48 (2.53)** 3.13 (2.90)*** 3.35 (2.88)***
Patent count 0.27 (3.61)*** 0.24 (1.16) 1.09 (1.81)*
Prior innovations-total(4) 0.09 (2.86)*** — —
Hypotheses
Prior breakthrough(4) — 0.31 (2.89)*** 0.21 (2.27)**
Prior nongeneric(4) — 0.14 (1.22) 0.09 (1.42)
Prior generic(4) — −0.33 (7.24)*** −0.31 (8.06)***
Foreign unit — — −0.77 (1.31)
Greenfield operation — — 0.43 (1.07)
JV, alliance, etc. — — 1.83 (2.92)***
Diagnostics
N 1,498 1,498 1,496
Log-likelihood −465.1746 −443.1226 −430.1763
Restricted log-likelihood −573.5764 −573.5764 −573.3019
d.f. 118 120 123
Likelihood ratio χ2(d.f.); (‘p’ value) 216.8036 (0.000) 260.9076 (0.000) 286.2512 (0.000)
Prediction ratio 0.8908 0.9011 0.9126

Notes: (1) Time dummies included;

(2) If pharmaceuticals industry (SIC = 2833, 2834, 2835, 5122) = 1; else = 0;
(3) If prescription drug, product type = 1; else = 0;
(4) Prior innovations = stock of past firm innovative activities;
(5) ‘t’ statistics computed using White’s heteroskedasticity-consistent variance-covariance matrix.
***‘t’ statistics significant at the 0.001 level;
**‘t’ statistics significant at the 0.05 level;
*‘t’ statistics significant at the 0.10 level.

innovations emerging from foreign subsidiaries do
not have a significantly lower likelihood of being
breakthrough innovations than those developed in
home country units. Thus, the data do not appear to
support Hypothesis 3 that posited a home country
bias in breakthrough innovations. Focusing on the
nature of the innovating unit, we find that innova-
tions emanating from Greenfield operations are not
less likely to be breakthrough innovations, contrary
to Hypothesis 4. However, innovations originating
Copyright © 2010 Strategic Management Society
from joint ventures and strategic alliances are
significantly more likely to be breakthroughs. This
provides evidence in support of Hypothesis 5.
DISCUSSION AND IMPLICATIONS
Contributions and future research
According to Baron (2007: 167), ‘entrepreneurship
happens because entrepreneurs conceive of new
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
products or services and then develop them through
the launch and operation of new ventures. In this
sense, entrepreneurs truly are the active element in
new venture creation . . . but which aspects of their
behavior and cognition are most relevant and hence
most deserving of careful attention?’ Our model
adds insight into what conditions cause some entre-
preneurs to be more successful than others (Baron,
2004). A primary theme in our article has been to
uncover and identify those competencies that allow
firms to breakthrough traditional firm routines and
create unprecedented new innovations that have the
potential to increase the quality of our lives. Through-
out our article, we have noted that successful long-
term growth hinges on the firm’s ability to punctuate
these more explorative types of innovations along-
side exploitative types of incremental innovations
(Raisch and Birkinshaw, 2008). Our corporate entre-
preneurship model emphasizes the collective impor-
tance of developing both mountain climbing (i.e., the
ex ante discovery process is replete of uncertainty)
and mountain building (i.e., the ex post value of
discovery process is uncertain) types of innovations
(Alvarez and Barney, 2007).
From an organizational learning perspective, a
key contribution of our work is the insight that
breakthrough and incremental innovations follow
different logics. We find supporting evidence that
innovation is an outcome of expertise based on accu-
mulated assets (e.g., Dierickx and Cool, 1989),
whereby prior experience in breakthrough innova-
tion increases the likelihood that a current innova-
tion will be breakthrough. In contrast, the effects of
prior experience in incremental innovations are
more complex, depending on whether the experi-
ence is in the nongeneric or generic realm. Prior
experience in generics significantly reduces the like-
lihood that the current innovation will be a break-
through and increases the likelihood that it will be
incremental. However, prior experience in nonge-
nerics does not significantly impact the likelihood of
breakthrough innovation.
It is possible that nongeneric incremental innova-
tion depends on many of the same creative processes
that underpin breakthrough innovation, so that the
difference between the two is more one of scale
(Nelson and Winter, 1982). Prior experience in non-
generic innovation does not appear to create a famil-
iarity trap (Ahuja and Lampert, 2001) that negatively
affects the firm’s absorptive capacity to assimilate
and develop new breakthroughs. Conversely, it may
be the case that an increase in cumulative generic
Copyright © 2010 Strategic Management Society
121
knowledge significantly reduces the knowledge
complexity and diversity within the firm, which
dampens the need for effective communication and
coordination, which is necessary for new learning
and technological change. Thus, we find that prior
innovation experience in generics handicaps (or
results in greater inertia for) breakthrough
innovation.
This finding has important implications for orga-
nizational ambidexterity (Raisch and Birkinshaw,
2008; Tushman and O’Reilly, 1996). Our results
imply that in the global pharmaceutical industry, an
exploitation strategy focused on incremental innova-
tion in generics is incompatible with organizational
ambidexterity. However, an exploitation strategy
focused on incremental innovation in nongenerics is
compatible with the successful pursuit of break-
through innovation and hence with organizational
ambidexterity. Consequently, the ambidexterity lit-
erature needs to recognize that exploitation strate-
gies are likely to be heterogeneous and that only
some of these strategies are compatible with organi-
zational ambidexterity.
The benefits and costs associated with sharing
knowledge across national borders has been a topic
of much discussion. With regard to the organization
of innovation, we find that innovation experience
dominates country of origin location effects. When
we ran our full model with all hypothesized vari-
ables, industry, product type, and patent count
emerge as significant predictors of breakthrough
innovation, while headquarters locations became
insignificant. It may be the case that through the use
of today’s ever-advancing electronic technologies,
traditional geographic proximity concerns have
decreased over the years. As a result, it has become
easier for firms to increase their frequency of inter-
national knowledge sharing activities. Moreover,
current research recognizes the important role that
foreign subsidiaries play in the context of compe-
tence creation (e.g., Cantwell and Mudambi, 2005;
Frost et al., 2002).
A goal of this study was to better understand the
parent-subsidiary dyad and go beyond previous
studies that have made arguments both for and
against the decentralization of knowledge creation
in MNEs. Interestingly, in our analysis, we do not
find support for a home country bias in breakthrough
innovation. At a broader level, additional research is
needed to understand whether or not innovation net-
works are becoming truly globalized. For instance,
Patel and Pavitt (1991) found that nearly 90 percent
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
122 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
of U.S. patenting activities from 1985 to 1990 were
concentrated in the firm’s home country. Further
analysis is needed to understand how the world
has changed since these seminal studies were
conducted.
We found intriguing results with regard to the
effect of firm organization and architecture on inno-
vation. Greenfield operations and acquired units
do not differ significantly in terms of the likelihood
of breakthrough versus incremental innovation.
However, innovations emanating from joint ven-
tures and alliances are significantly more likely to
be breakthroughs. This finding is very much in line
with the recent literature on open innovation (Laursen
and Salter, 2006) and suggests a link between firm
organization and R&D outcomes. Greenfield opera-
tions and acquired units are generally assessed on
the basis of bottom line performance as profit centers
(e.g., Delios and Beamish, 2001). Hence, they have
an incentive to focus on what has been called loss
prevention (Chandler, 1991). Firms with this men-
tality tend to be relatively myopic in their decision
making, with a tendency toward using financial tools
(e.g., IRR, NPV) that support shorter-term, less risky
projects (Mote et al., 2001). Correspondingly, they
have a strong disincentive to undertake breakthrough
innovations, which by their very nature are highly
risky and unlikely to yield bottom-line results in the
short run.
By contrast, strategic alliances and nonequity
joint ventures tend to be assessed as cost centers
(Hamel and Prahalad, 1990), since it is always dif-
ficult and often impossible to assess them on the
basis of their bottom-line contributions. As such,
they are more likely to be treated by their multiple
parent locations as a type of strategic option
(Bowman and Hurry, 1993). Further, the lack of
single headquarter locus of control allows for the
retention of a diversity of skill sets, which is a very
important precursor to breakthrough innovation
(Powell, Koput, and Smith-Doerr, 1996). Put simply,
we find that diversity is less likely to be stamped out
in strategic alliances. Firms are allowed to bring
their own individual perspectives to the partnership
filled with a unique repertoire of skills, knowledge,
and strategic assets. The dynamic interaction
between partners increases organizational learning,
resulting in greater innovative performance. Thus,
our findings are consistent with breakthrough inno-
vations occurring and being ascribed to joint ven-
tures and strategic alliances (Cohen and Levinthal,
1990; Kotabe and Swan, 1995; Laursen and Salter,
Copyright © 2010 Strategic Management Society
2006; March, 1991; Rosenkopf and Schilling, 2007).
It is likely that following a breakthrough, the joint
venture or strategic alliance will be acquired by one
or the other party; subsequent exploitative incremen-
tal innovations would then be ascribed to the acquired
unit. This storyline is consistent with our results.
Our results also indicate some notable conclu-
sions about firm age and firm size on corporate
entrepreneurship activities. In general, empirical
research supports the notion that smaller and younger
firms tend to be more innovative than larger firms
because they have less crowded local knowledge
networks, which reduces inertia and increases the
exploration of new opportunities (e.g., Almeida and
Kogut, 1997; Hannan and Freeman, 1984). Larger
and older firms tend to have more rigid organiza-
tional routines, which limit the rewards associated
with corporate intrapreneurship. Sorensen and
Stuart (2000) reported contradictory findings with
respect to age and innovation. They found that even
though older firms were more likely to have higher
rates of innovation than younger firms, they had
greater difficulties than their younger counterparts in
creating innovations that matched the current tech-
nological needs of their buyers. The data in our
research suggest that an important factor affecting
the creation of breakthroughs was firm size, not firm
age. The influence of firm age on the firm’s propen-
sity to innovate was not conclusive. Future research
is necessary to explore this preliminary finding in
more detail.
We also present data that raise some intriguing
research questions. Firms exploit both breakthrough
and incremental innovations as platforms to generate
further types of incremental innovation. In our data-
base, these types of innovations appear as new effi-
cacy supplements (NESs). These drugs represent a
modification to an existing new drug application.
While they are not assigned a new drug application
number, they are considered as new product innova-
tions and must meet conditions specified by the FDA
before they can be legally marketed. We tracked
these efficacy supplements for each of the innova-
tions in our study starting from the years 1992 to
2002, for both public and private firms, up until the
year 2008. The results are presented in Table 5. For
the 2,885 original new drugs in our database, we
recorded a total of 16,684 efficacy supplements. We
found that some drugs over their lifetime were
the basis for more than 75 efficacy supplements,
while some led to no efficacy supplements at all.
Overall, we find that breakthrough innovations and
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
Table 5. Innovation and exploitation: new efficacy supplements associated with original new drug approvals (NDAs)
through 2008
Innovation type Firm type
Innovations 1992–2002
Breakthrough innovations Public
Private
All firms
Incremental innovations Public
Private
All firms
Notes: (1) NES = New efficacy supplement.
incremental innovations appear to generate about the
same number of NESs per year (0.465 versus 0.458).
It is possible, however, that breakthrough and incre-
mental innovations are exploited in different ways.
We suggest that tracking and analyzing the future
exploitation of commercialized innovations is a
fruitful path for further research.
Limitations and future research
This study has limitations, which proffer opportuni-
ties for future research. There exists considerable
research about the importance of industry-related
pressures and how deviation from the norm may
dramatically affect the firm’s performance. This
study recognizes that responding to industry-level
demands may have different repercussions for dif-
ferent firms. A generalizability concern of this study
is that it focused on a single industry, albeit an
important one. The path to developing new products
within the global pharmaceutical industry is a long
and expensive one. A consideration for the future is
to examine corporate entrepreneurship variations
across other industries.
Another limitation is that our study was based on
secondary data analysis from legally constituted
organizations. Like all research of this kind, we were
unable to measure the firm’s attitudes, opinions, and
perceptions about the effectiveness/ineffectiveness
of product-level innovation strategies. We believe
that the collection of primary data can enhance our
empirical understanding of the link between entre-
preneurship and corporate innovation (Snow, 2007).
In this study, we measured entrepreneurship by
examining only breakthrough and incremental com-
mercialized innovations. Due to the type of data that
we collected, there were relatively small numbers of
Copyright © 2010 Strategic Management Society
123
Number of NDA Number of NES(1) per NDA
innovation per year* up to 2008
263 0.416
32 0.513
295 0.465
2,145 0.522
445 0.394
2,590 0.458
breakthroughs. While much of the current literature
is based on examining this dichotomy, it is well
recognized that the innovation continuum between
breakthrough and incremental contains many shades
of gray. Future research should develop more sophis-
ticated innovation models that specifically recognize
the innovation continuum while examining the link
between commercialized innovations of this kind
and market performance.
CONCLUSION
In this study, we examined the corporate entrepre-
neurship activities of firms in the global pharmaceu-
tical industry (see Figure 1). We examined a total of
about 1,500 commercialized innovations from the
years 1993 to 2002, accounting for the effect of past
organizational learning in 1992. From a managerial
and theoretical perspective, we have shown that
developing breakthrough and incremental innova-
tions are complex processes and appear to be based
on different logics. Our results support the view that
breakthrough innovations are based on accumulated
expertise, as represented by a successful prior track
record in managing this process.
In contrast, the effect of accumulated expertise in
incremental innovation varies depending on whether
the experience is in nongenerics or generics. Prior
experience in innovation in generics reduces the
likelihood of breakthrough innovation, while prior
experience in nongeneric innovation does not. Thus,
while breakthrough innovation seems to be highly
path dependent, incremental innovation seems to be
a more heterogeneous phenomenon. Some forms of
incremental innovation (i.e., this involves nongener-
ics) are compatible with breakthrough innovation
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
124 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
and, hence, with organizational ambidexterity. Other
forms of incremental innovation (i.e., this involves
generics) significantly reduce the likelihood of
breakthrough innovation and are, therefore, incom-
patible with organizational ambidexterity.
It is important to note that acquiring foreign sub-
sidiary knowledge did not negatively impact our
sample firms’ breakthrough capabilities. While
foreign subsidiaries have long been considered bases
for innovatory activities, many studies have sug-
gested that truly breakthrough innovations will be
concentrated in the firm’s home base. In a relatively
large and comprehensive database, we find that
innovations emanating from foreign subsidiaries are
just as likely to be breakthroughs as those emerging
from the firm’s home base. Finally, we find that
innovations stemming from alliances and joint ven-
tures are more likely to be breakthroughs. As empha-
sized in the literature on alliances and open
innovation, it is likely that the diversity implicit in
such organizational forms is supportive of break-
through innovation.
Corporate entrepreneurship is increasingly viewed
as a valuable instrument for rejuvenating and revital-
izing existing companies. Our results indicate the
nature of innovation that policy makers may expect
from existing populations of firms. We provide evi-
dence that intangible resources in the form of accu-
mulated expertise are critical signals presaging
future success in breakthrough innovation. Our find-
ings also suggest that incremental innovation pro-
cesses are heterogeneous and that some are less
conducive to breakthrough innovation than others.
ACKNOWLEDGEMENTS
An earlier version of this article was presented at the
2009 SEJ Special Issue Conference at York University,
Toronto. We appreciate the supportive comments that
we received from the conference participants and our
discussant, Mike Wright. We are also grateful to the
special issue editors (Doug Cumming, Don Siegel and
Mike Wright) and the two anonymous reviewers, whose
extensive comments helped improve the quality of this
manuscript.
REFERENCES
Abernathy WJ, Utterback JM. 1978. Patterns of industrial
innovation. Technology Review 80(7): 40–47.
Copyright © 2010 Strategic Management Society
Ahuja G, Lampert CM. 2001. Entrepreneurship: the large
corporation: a longitudinal study of how established
firms create breakthrough inventions. Strategic Manage-
ment Journal 22(6–7): 521–543.
Aldrich HE, Kim PH. 2007. Small worlds, infinite possibili-
ties? How social networks affect entrepreneurial team
formation and search. Strategic Entrepreneurship
Journal 1(1–2): 147–165.
Aldrich JH, Nelson FD. 1984. Linear Probability, Logit, and
Probit Models. Sage Publications: Beverly Hills, CA.
Ali A. 1994. Pioneering versus incremental innovation:
review and research propositions. Journal of Product
Innovation Management 11(1): 46–61.
Almeida P, Kogut B. 1997. The exploration of technologi-
cal diversity and the geographic localization of innova-
tion. Small Business Economics 9(1): 21–31.
Alvarez SA, Barney JB. 2007. Discovery and creation:
alternative theories of entrepreneurial action. Strategic
Entrepreneurship Journal 1(1–2): 11–26.
Anderson P, Tushman ML. 1991. Managing though cycles
of technological change. Research Technology Manage-
ment 34(3): 26–31.
Asakawa K. 2001. Evolving headquarters-subsidiary
dynamics in international R&D: the case of Japanese
multinationals. R&D Management 31(1): 1–14.
Audretsch DB. 1995. Innovation and Industry Evolution.
The MIT Press: Cambridge, MA.
Barney JB. 1991. Looking inside for competitive advan-
tage. Academy of Management Executive 9(4): 49–61.
Baron RA. 2004. The cognitive perspective: a valuable tool
for answering entrepreneurship’s basic ‘why’ questions.
Journal of Business Venturing 19: 221–239.
Baron RA. 2007. Behavioral and cognitive factors in entre-
preneurship: entrepreneurs as the active element in new
venture creation. Strategic Entrepreneurship Journal
1(1–2): 167–182.
Bartlett CA, Ghoshal S. 1989. Managing Across Borders.
Harvard Business School Press: Boston, MA.
Basberg BL. 1987. Patents and the measurement of techno-
logical change: a survey of the literature. Research Policy
16(2–4): 131–141.
Binks MR, Vale PA. 1990. Entrepreneurship and Economic
Change. McGraw-Hill: New York.
Birkinshaw J. 1997. Entrepreneurship in multinational
corporations: the role of subsidiary initiative. Strategic
Management Journal 18(3): 207–229.
Bowman EH, Hurry D. 1993. Strategy through the option
lens: an integrated view of resource investments and the
incremental-choice process. Academy of Management
Review 18(4): 760–782.
Burgleman RA. 1983. Corporate entrepreneurship and
strategic management: insights from a process study.
Management Science 29(12): 1349–1363.
Cantwell J. 1989. Technological Innovation and Multina-
tional Corporations. Blackwell Publishers: Cambridge,
MA.
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
Cantwell J, Mudambi R. 2005. MNE competence-creating
subsidiary mandates. Strategic Management Journal
26(12): 1109–1128.
Chamberlain G. 1980. Analysis of covariance with qualita-
tive data. Review of Economic Studies 47(1): 225–38.
Chandler AD. 1991. The functions of the HQ unit in the
multi-business firm. Strategic Management Journal
12(8): 31–50.
Christensen CM. 1997. The Innovator’s Dilemma. Harvard
Business Press: Boston, MA.
Cohen WM, Klepper S. 1996. Firm size and the nature of
innovation within industries: the case of process and product
R&D. Review of Economic Studies 78(2): 232–244.
Cohen W, Levinthal D. 1990. Absorptive capacity: a new
perspective on learning and innovation. Administrative
Science Quarterly 35(1): 128–152.
Coombs JR, Mudambi R, Deeds DL. 2006. An examination
of the investment in U.S. biotechnology firms by foreign
and domestic corporate partners. Journal of Business
Venturing 21(4): 405–428.
Cooper AC. 2007. Behavioral characteristics of entrepre-
neurial activity. Strategic Entrepreneurship Journal
1(1–2): 145–146.
Cumming D, Macintosh J. 2000. The determinants of R&D
expenditures: a study of the Canadian biotechnology
industry. Review of Industrial Organization 17(4):
357–370.
Cyert R, March J. 1963. A Behavioral Theory of the Firm.
Prentice-Hall: Englewood Cliffs, NJ.
Delios A, Beamish PW. 2001. Survival and profitability: the
roles of experience and intangible assets in foreign sub-
sidiary performance. Academy of Management Journal
44(5): 1028–1038.
De Meyer A. 1992. Management of international R&D
operations. In Technology Management and Interna-
tional Business: Internationalization of R&D and Tech-
nology, Grandstand O, Hakanson L, Sjolander S (eds).
John Wiley and Sons: New York; 163–179.
DeSanctis G, Glass JT, Ensing IM. 2002. Organizational
designs for R&D. Academy of Management Executive
16(3): 55–66.
Dierickx I, Cool K. 1989. Asset stock accumulation and
sustainability of competitive advantage. Management
Science 35(12): 1504–1511.
DiMasi JA. 2000. New drug innovation and pharmaceutical
industry structure: trends in the output of pharmaceutical
firms. Drug Information Journal 34(4): 1169–1194.
DiMasi JA, Hansen RW, Grabowski HC. 2003. The price
of innovation: new estimates of drug development costs.
Journal of Health Economics 22(2): 151–185.
Drucker P. 1985. Innovation and Entrepreneurship. Harper
& Row: New York.
Ernst H. 2001. Patent applications and subsequent changes
of performance: evidence from time-series cross-section
analyses on the firm level. Research Policy 30(1):
143–157.
Copyright © 2010 Strategic Management Society
125
Fischer M. 2001. Innovation, knowledge creation, and
systems of innovation. The Annals of Regional Science
35(2): 199–216.
Frost TS. 2001. The geographic sources of foreign subsid-
iaries’ innovations. Strategic Management Journal 22(2):
101–123.
Frost TS, Birkinshaw JM, Ensign PC. 2002. Centers of
excellence in multinational corporations. Strategic Man-
agement Journal 23(11): 997–1018.
Graubard BI, Korn EL. 1999. Predictive margins with
survey data. Biometrics 55(2): 652–659.
Greene W. 2004. The behavior of the maximum likelihood
estimator of limited dependent variable models in the
presence of fixed effects. Econometrics Journal 7(1):
98–119.
Guth WD, Ginsberg A. 1990. Guest editor’s introduction:
corporate entrepreneurship. Strategic Management
Journal 11(Summer Special Issue): 5–16.
Hamel G, Prahalad CK. 1990. On the core competence of
the corporation. Harvard Business Review 68(May-
June): 79–91.
Hannan M, Freeman J. 1984. Structural inertia and organi-
zational change. American Sociological Review 49(2):
149–164.
Hayward M. 2002. When do firms learn from their acquisi-
tion experience? Evidence from 1990–1995. Strategic
Management Journal 23(1): 21–39.
Hedlund G. 1986. The hypermodern MNC: a heterarchy?
Human Resource Management 25(1): 9–35.
IMS Health. 2007. IMS Health reports global pharmaceuti-
cal market grew 7.0 percent in 2006, to $643 Billion. IMS
Press Release, 20 March. Available at: http://ir.imshealth.
com/phoenix.zhtml?c=67124&p=irol-news&nyo=3
[accessed 7 August 2009].
Ireland RD, Covin JG, Kuratko DF. 2009. Conceptualizing
corporate entrepreneurship strategy. Entrepreneurship
Theory and Practice 33(1): 19–46.
Ireland RD, Hitt MA, Simon DG. 2003. A model of strate-
gic entrepreneurship: the construct and its dimensions.
Journal of Management 29(6): 963–989.
Jaffe AB, Trajtenberg M, Henderson R. 1993. Geographic
localization of knowledge spillovers as evidenced by
patent citations. Quarterly Journal of Economics 108(3):
577–598.
Katz E. 2001. Bias in conditional and unconditional fixed
effects logit estimation. Political Analysis 9(4): 379–384.
Kelley DJ, Peters L, Colarelli O’Connor G. 2009. Intra-
organizational networking for innovation-based corpo-
rate entrepreneurship. Journal of Business Venturing
24(3): 221–235.
Ketchen DJ Jr., Ireland RD, Snow CC. 2007. Strategic
entrepreneurship, collaboration, innovation, and wealth
creation. Strategic Entrepreneurship Journal 1(3–4):
371–385.
Kogut B, Zander U. 1993. Knowledge of the firm and the
evolutionary theory of the multinational corporation.
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
126 D. Dunlap-Hinkler, M. Kotabe, and R. Mudambi
Journal of International Business Studies 24(4):
625–645.
Kotabe M, Swan KS. 1995. The role of strategic alliances
in high technology new product development. Strategic
Management Journal 16(8): 621–636.
Kuratko DF, Audretsch DB. 2009. Strategic entrepreneur-
ship: exploring different perspectives of an emerging
concept. Entrepreneurship Theory and Practice 33(1):
1–17.
Kuratko DF, Welsch HP. 2001. Strategic Entrepreneurial
Growth. Harcourt College Publishers: Fort Worth,
TX.
Lawson B, Samson D. 2001. Developing innovation capa-
bility: a dynamic capabilities approach. International
Journal of Innovation Management 5(3): 377–400.
Lane PJ, Lubatkin M. 1998. Relative absorptive capacity
and interorganizational learning. Strategic Management
Journal 19(5): 461–477.
Laursen K, Salter A. 2006. Open for innovation: the role of
openness in explaining innovation performance among
U.K. manufacturing firms. Strategic Management
Journal 27(2): 131–150.
Leiblein MJ. 2007. Environment, organization, and innova-
tion: how entrepreneurial decisions affect innovative
success. Strategic Entrepreneurship Journal 1(1–2):
141–144.
Leonard-Barton D. 1992. Core capabilities and core rigidi-
ties: a paradox in managing new product development.
Strategic Management Journal 13(Summer Special
Issue): 11–125.
Levinthal DA, March JG. 1993. The myopia of learning.
Strategic Management Journal 14(Winter Special Issue):
95–112.
Levitt B, March JG. 1988. The architecture of simplicity.
Academy of Management Review 18(1): 116–138.
Lundvall B. 2007. National innovation systems: analytical
concept and development tool. Industry and Innovation
14(1): 95–119.
Macdonald S. 2004. When means become ends: consider-
ing the impact of patent strategy on innovation. Informa-
tion Economics and Policy 16(10): 135–158.
Mandel M. 2009. Innovation interrupted. Businessweek
15 June: 34–39.
March GJ. 1991. Exploration and exploitation in organiza-
tional learning. Organization Science 2(1): 71–87.
Markman GD, Siegel DS, Wright M. 2008. Research and
technology commercialization. Journal of Management
Studies 45(8): 1410–1423.
Masalin L. 2003. Nokia leads change through continuous
learning. Academy of Management Learning and Educa-
tion 2(1): 68–72.
Miller D. 1983. The correlates of entrepreneurship in three
types of firms. Management Science 29(7): 770–791.
Mote J, Boylan R, Rice M. 2001. A modeling view of
breakthrough innovation. Engineering Management
Journal 13(3): 35–40.
Copyright © 2010 Strategic Management Society
Mudambi R. 2008. Location, control, and innovation in
knowledge intensive industries. Journal of Economic
Geography 8(5): 699–725.
Mudambi R, Navarra, P. 2004. Is knowledge power?
Knowledge flows, subsidiary power, and rent seeking
within MNCs. Journal of International Business Studies
35(5): 385–406.
Narayanan VK, Yang Y, Zahra SA. 2009. Corporate ventur-
ing and value creation: a review and proposed frame-
work. Research Policy 38(1): 58–76.
Narula R. 2002. Innovation systems of ‘inertia’ in R&D
location: Norwegian firms and the role of systemic lock-
in. Research Policy 31(5): 795–816.
Nelson R, Winter S. 1982. An Evolutionary Theory of Eco-
nomic Change. The Belknap Press of Harvard University
Press: Cambridge, MA.
Nerkar A. 2003. Old is gold? The value of temporal explo-
ration in the creation of new knowledge. Management
Science 49(2): 211–229.
Palmer R, Brookes R. 2002. Incremental innovation: a case
study analysis. Journal of Database Marketing 10(1):
71–83.
Patel P, Pavitt K. 1991. Large firms in the production
of the world’s technology: an important case of ‘non-
globalization.’ Journal of International Business Studies
22(1): 1–21.
Perlmutter HV. 1969. The tortuous evolution of the
multinational corporation. Columbia Journal of World
Business 4(1): 9–19.
Peterson R, Berger D. 1972. Entrepreneurship in organiza-
tions. Administrative Science Quarterly 16(1): 97–
106.
Phan PH, Wright M, Ucbasaran D, Wee-Liang T. 2009.
Corporate entrepreneurship: current research and future
directions. Journal of Business Venturing 24(3):
197–205.
Pinchott G. 1985. Intrapreneuring. Harper & Row: New
York.
Powell WW, Koput KW, Smith-Doerr L. 1996. Interorga-
nizational collaboration and the locus of innovation:
networks of learning in biotechnology. Administrative
Science Quarterly 41(1): 116–145.
Prahalad CK, Doz Y. 1987. The Multinational Mission:
Balancing Local Demands and Global Vision. The Free
Press, Macmillian: New York.
Raisch S, Birkinshaw J. 2008. Organizational ambidexter-
ity: antecedents, outcomes, and moderators. Journal of
Management 34(3): 375–409.
Redding S. 2002. Path dependence, endogenous innovation,
and growth. International Economic Review 43(4):
1215–1249.
Rogers E. 1983. The Diffusion of Innovations. Free Press:
New York.
Romanelli E, Tushman ML. 1994. Organizational transfor-
mation as punctuated equilibrium: an empirical test.
Academy of Management Journal 37(5): 1141–1166.
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej
 Breakthrough versus Incremental Innovation
Rosenkopf L, Schilling M. 2007. Comparing alliance
network structure across industries: observations and
explanations. Strategic Entrepreneurship Journal
1(3–4): 191–209.
Rothaermel FT, Boeker W. 2008. Old technology meets
new technology: complementarities, similarities, and alli-
ance formation. Strategic Management Journal 29(1):
47–77.
Rothaermel FT, Deeds DL. 2004. Exploration and exploita-
tion alliance in biotechnology: a system of new product
development. Strategic Management Journal 25(3):
201–221.
Sakakibara M, Porter ME. 2001. Competing at home to win
abroad: evidence from Japanese industry. Review of Eco-
nomics & Statistics 83(2): 311–324.
Schumpeter JA. 1934. The Theory of Economic
Development. Harvard University Press: Cambridge,
MA.
Schumpeter JA. 1939. Business Cycles: A Theoretical, His-
torical and Statistical Analysis of The Capitalist Process.
McGraw-Hill: New York.
Schumpeter JA. 1942. Capitalism, Socialism, and Democ-
racy. Harper & Brothers: New York.
Sellers LJ. 2002. Bristol-Myers Squibb: wounded and
vulnerable. Pharmaceutical Executive 1 May: 18.
Simon HA. 1985. What we know about the creative process.
In Frontiers in Creative and Innovative Management,
Kuhn RL (ed). Ballinger: Cambridge, MA; 3–
22.
Smith A. 2007. Pfizer to cut 10,000 jobs, shut 5 plants.
CNNMoney.com, 22 July. Available at: http://money.
cnn.com/2007/01/22/news/companies/pfizer/index.htm
[accessed 20 December 2008].
Snow CC. 2007. Innovation. Strategic Entrepreneurship
Journal 1(1–2): 101–102.
Sorensen JB, Stuart TE. 2000. Aging, Obsolescence, and
Organizational Innovation. Administrative Science Quar-
terly 45(1): 81–112.
Teece DJ. 1981. The multinational enterprise: market
failure and market power considerations. Sloan Manage-
ment Review 22(3): 3–17.
Teece DJ. 1988. Technological change and the nature of the
firm. In Technical Change and Economic Theory, Dosi
Copyright © 2010 Strategic Management Society
127
G, Freeman C, Nelson R, Soete L (eds). Pinter: London:
256–281.
Tidd J. 2001. Innovation management in context: environ-
ment, organization, and performance. International
Journal of Management Reviews 3(3): 169–183.
Tushman ML, Anderson P. 1986. Technological disconti-
nuities and organizational environments. Administrative
Science Quarterly 31(3): 439–465.
Tushman ML, Nadler D. 1986. Organizing for innovation.
California Management Review 28(3): 74–92.
Tushman ML, O’Reilly CA. 1996. Ambidextrous organiza-
tions: managing evolutionary and revolutionary change.
California Management Review 38(4): 8–30.
Utterback JM, Abernathy WJ. 1975. A dynamic model of
process and product innovation. Omega, International
Journal of Management Science 3(6): 639–656.
Vale PA, Addison M. 2002. Promoting entrepreneurship
and innovation in a large company: creating a virtual
portfolio. Journal of Change Management 2(4):
334–343.
Vernon R. 1966. International investment and international
trade in the product cycle. Quarterly Journal of Econom-
ics 41: 255–267.
von Hippel E. 1998. Economics of product development by
users: the impact of ‘sticky’ local information. Manage-
ment Science 44(5): 629–644.
von Hippel E, Thomke S, Sonnack M. 1999. Creating
breakthroughs at 3M. Harvard Business Review 77(5):
47–56.
Wiklund J, Shepherd DA. 2009. The effectiveness of alli-
ances and acquisitions: the role of resource combination
activities. Entrepreneurship Theory and Practice 33(1):
193–212.
Yang Q, Mudambi R, Meyer K. 2008. Conventional and
reverse knowledge flows in multinational corporations.
Journal of Management 34(5): 882–902.
Zahra SA. 1991. Predictors and financial outcomes of
corporate entrepreneurship: an extension. Journal of
Business Venturing 6(4): 259–285.
Zahra SA, Coven JG. 1995. Contextual influences on the
corporate entrepreneurship-performance relationship:
a longitudinal analysis. Journal of Business Venturing
10(1): 43–58.
Strat. Entrepreneurship J., 4: 106–127 (2010)
DOI: 10.1002/sej