Strategic Management Journal

Strat. Mgmt. J., 19: 1063–1077 (1998)

EVOLUTIONARY DIFFUSION: INTERNAL AND

EXTERNAL METHODS USED TO ACQUIRE
ENCOMPASSING, COMPLEMENTARY, AND
INCREMENTAL TECHNOLOGICAL CHANGES IN THE
LITHOTRIPSY INDUSTRY
ANURADHA NAGARAJAN AND WILL MITCHELL*
University of Michigan Business School, Ann Arbor, Michigan, U.S.A.

This study links theories concerning methods that firms use to acquire technology with theories
concerning types of technological change. We place particular emphasis on interorganizational
relationships. We predict that firms will often acquire know-how needed for encompassing
technological change through equity-based arrangements with other organizations, complemen-
tary technological changes through nonequity interorganizational arrangements, and incremental
changes through internal R&D. Our theory draws on perspectives that emphasize the need to
develop new competencies within a business organization and to protect the value of existing
competencies. Our empirical analysis examines methods of technology acquisition that firms
have used in the commercialization of medical lithotripters, which are devices that fragment
stones in the kidney and gall bladder. The analysis contributes to a better understanding of
how technology acquisition methods vary with the manner in which technological change relates
to a firm’s existing capabilities. The study also helps develop our understanding of the
evolutionary processes by which capabilities diffuse through an industry.  1998 John Wiley &
Sons, Ltd.

Firms often must acquire new know-how as tech- zational methods of acquiring know-how vary
nology changes in an industry. Past research iden- with the type of technological change. We define
tifies different ways that firms acquire new know- technological change as either encompassing,
how when technological change affects their busi- complementary, or incremental, where the cate-
nesses, including internal development and acqui- gories differ in the degree and the focus of a
sition from other firms (e.g., Teece, 1986; Mitch- technological change. We argue that the different
ell and Singh, 1992; Pisano, 1990; Williamson, types of change will tend to have different effects
1991). In parallel, other research describes differ- on firms’ existing capabilities. We then predict
ent types of technological change (e.g., Abernathy that firms will often acquire know-how needed
and Utterback, 1978; Abernathy and Clark, 1985; for encompassing technological change through
Tushman and Anderson, 1986; Henderson and equity-based interorganizational relationships,
Clark, 1990; Mitchell, 1991). Although the complementary technological changes through
research streams relate to each other, analysts nonequity interorganizational relationships, and
often do not develop the link between them. incremental changes through internal R&D. A
This paper shows that internal and interorgani- study of 44 cases of technological change that
have affected medical lithotripters between 1989
and 1991 supports our predictions. The analysis
Key words: technological change; internal and exter- contributes to a better understanding of how inter-
nal know-how acquisition; firm capabilities nal development and interorganizational tech-
*
Correspondence to: Will Mitchell, University of Michigan
Business School, 701 Tappan Street, Ann Arbor, MI 48109- nology acquisition methods vary with the manner
1234, U.S.A. in which technological change affects a firm’s

CCC 0143–2095/98/111063–15 $17.50 Received 6 May 1996

 1998 John Wiley & Sons, Ltd. Final revision received 22 December 1997
1064 A. Nagarajan and W. Mitchell

existing capabilities. In turn, the study helps Arrow, 1974; Nelson and Winter, 1977; Winter,
develop our understanding of the evolutionary 1987). Such codes often intertwine in the form
processes by which capabilities diffuse through of cumulative tacit organizational routines that
an industry. span external organizational boundaries only with
difficulty (Cyert and March, 1963; Nelson and
Winter, 1982). By contrast, people working
BACKGROUND AND PREDICTIONS within a single business unit often can exchange
and develop tacit information owing to their com-
Methods of acquiring know-how when
mon understanding of the business’s routines
technology changes
(Cohen and Levinthal, 1990), while people work-
Technology is the ability to create products, proc- ing within a multiproduct firm may be able to
esses, and services (Friar and Horwitch, 1985). build on a common technology and develop appli-
This definition embraces physical assets and non- cations in different product areas (Nelson, 1959;
physical knowledge, and views technology as sys- Argyres, 1996). Thus, internal R&D offers the
tems of activities as well as individual products potential to protect existing competencies and to
and processes. Changing an organization’s tech- develop new competencies within a business
nology involves adjusting the assets, knowledge, organization.
and techniques that the organization uses to con- The strength of internal R&D is also a limi-
vert inputs to outputs (Rosenberg, 1972; Tushman tation. It is often costly and difficult for busi-
and Anderson, 1986). First-mover advantages in nesses to develop capabilities that are new to a
capturing key resources and learning curve bene- firm’s existing internal R&D repertoire. Cost is-
fits create incentives for investing in technological sues arise because the irreversible commitments
adjustment (Schumpeter, 1934; Lieberman and required of the firm under conditions of signifi-
Montgomery, 1988). Firms can carry out such cant uncertainty may impede the investment in
technological adjustments through internal R&D internal R&D that requires new assets (Mueller
or by forming relationships with other organi- and Tilton, 1969; Ghemawat, 1991). In addition
zations. External relationships include equity- to cost issues, internal development encounters
based associations, such as joint ventures and difficulties that arise owing to forces both outside
direct investments, and nonequity associations, and within a focal business. Externally, insti-
such as technology licensing, technology tutional forces may limit a firm’s investment paths
exchanges, testing agreements, technology sharing by constraining its behavior within accepted
agreements, and research contracts. activities and arenas (Scott, 1987). Within the
An extensive literature discusses many sources business organization, organizational inertia may
of advantages for internal R&D over external limit the expansion of the firm’s existing bundle
arrangements for adjusting technology. Two basic of routines and inhibit the firm’s ability to
types of the advantages of internal R&D are at develop technological competency in areas
the core of the discussions: mitigating risks of beyond the firm’s local search area (Cyert and
opportunistic behavior and building on tacit March, 1963; Nelson and Winter, 1982). For
organizational routines. In part, internal R&D reasons stemming from both cost and difficulty,
protects a firm from opportunistic behavior by a therefore, internal R&D is likely to be most
partner in an external research relationship, effective when the new capabilities being
including both reducing inadvertent leakage of developed relate to the organization’s existing set
proprietary information to a partner and guarding of capabilities.
against active opportunism by a partner Firms often use external know-how acquisition
(Williamson, 1975; Teece, 1986, 1988; Pisano, methods when faced with limits and constraints
1990). Moreover, and often more importantly on internal research, where external methods
than issues stemming from opportunism, tech- sometimes involve equity holdings and sometimes
nology development typically involves the com- taking nonequity approaches. Using external
munication of tacit knowledge. The members of sources allows a firm to acquire technology that
a firm often have a common code of communi- is outside its current capabilities (Singh and
cation for discussing tacit knowledge and Mitchell, 1996). Equity-based relationships allow
developing new capabilities (Penrose, 1958; greater control of technology acquisition than do
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1065

nonequity approaches, but also require longer- or reduce the cost of an existing product
term financial and managerial commitment to the (Utterback and Abernathy, 1975; Abernathy and
other party. Clark, 1985). Incremental innovations tend to
We expect that different types of technological build on existing capabilities of firms and exploit
change will tend to have different relationships the potential of established designs (Argyres,
with firms’ existing capabilities and, therefore, 1996).
influence the choice between internal development Like the distinction between radical and
and external acquisition of new know-how. The incremental changes, the distinction between core
following section describes different types of and complementary activities exists within many
technological change and develops our arguments strands in the technology literature. Authors have
concerning how they will affect the methods firms used several terms to describe the general con-
use to acquire new know-how. cepts. Thompson (1967) distinguished between
core activities at the focus of a firm’s attention
and peripheral activities that extend beyond the
Encompassing, complementary, and
boundaries of the firm but affect the firm’s activi-
incremental technological change
ties. In parallel, Richardson (1972) distinguished
We distinguish among three types of change between similar and complementary activities.
within a technological system: encompassing, Richardson defined similar activities as the bundle
complementary, and incremental changes. The of routines that a firm is most familiar with and
types of change vary on two dimensions: first, monitors most closely, and defined complemen-
the degree of change and, second, whether the tary activities as those activities essential to the
changes tend to focus on core or complementary overall commercialization process that do not fall
activities of the firm. Encompassing changes within the specialized set of a firm’s routines.
involve radical alteration of core activities, Cleland and Bursic (1992) distinguished between
regardless of the degree to which the changes also core technology that firms require to make a
affect complementary activities. Complementary product or service and complementary technology
changes involve radical alteration of complemen- that supports the product or the service. Similarly,
tary activities that do not also involve radical Henderson and Clark (1990) distinguished
alteration of core activities. Incremental changes between core product technologies and comple-
involve incremental adjustments to core or com- mentary technologies, while also identifying link-
plementary activities. The two dimensions, of rad- ages among components as key foci of change.
ical vs. incremental innovations and of core vs. Argyres (1996), meanwhile, distinguishes
complementary activities, are common to many between changes that deepen a firm’s existing
perspectives in the technology literature. knowledge and changes that broaden a firm’s
Many analyses view technological change as a existing capabilities.
process in which incremental changes follow rad- Drawing on the prior research, we use the
ical advances in technology (Sahal, 1985). Rad- concepts of core and complementary activities for
ical innovations are changes that draw on a differ- our study. Core activities are the set of routines
ent set of engineering and scientific principles involving physical assets, knowledge, and com-
and significantly alter the set of competencies petencies that are intrinsic to the engineering
required for success and survival (Jewkes, Saw- and manufacturing of a product. Complementary
ers, and Stillerman, 1958; Tilton, 1971; Aberna- activities are the set of routines involving physical
thy and Utterback, 1978). These innovations alter assets, knowledge and competencies that contrib-
the competitive dynamics of the industry signifi- ute to the production of or enhance the commer-
cantly. Radical innovations often destroy capabili- cial utility of a product. In some cases, there will
ties, both in the ability to create goods and be ambiguity between core and complementary
services and in the ability to sell them in a activities. Indeed, the existence of indistinct gray
market (Mitchell and Singh, 1993; Christensen areas helps explain why similar firms respond
and Bower, 1996). Incremental innovations are differently in similar environments, as managerial
improvements, refinements, minor modifications, perceptions may differ concerning what is core
and extensions of core or complementary product and what is complementary, or seemingly minor
and process technology that enhance the utility firm-level differences may underlie substantial
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
1066 A. Nagarajan and W. Mitchell

differences in what actually are core and comple- basically unchanged but the complementary
mentary activities at different firms. Nonetheless, activities undergo radical change. Such cases may
the concepts of core and complementary activities involve the processes used by the buyers in their
are sufficiently distinct that they have both commercialization of the product and by the firms
empirical and conceptual value. that provide components and capital equipment
The intersection of the two dimensions of tech- for the production process. Tire technology, for
nological change, that is, of radical or incremental instance, complements the core technology of a
type of change with core or complementary locus motor vehicle. The move to radial tires from
of change, produces our basic typology of bias ply did not directly affect core automobile
encompassing, complementary, and incremental technology but had significant impacts on auto-
change. As we noted above, encompassing motive manufacturers by affecting the design and
changes involve radical alteration of core activi- manufacture of suspension, brake, and steering
ties, regardless of the degree to which the changes systems. Incremental technological change occurs
also affect complementary activities. Complemen- when incremental innovations affect core or
tary changes involve radical alteration of comple- component activities. Incremental changes tend to
mentary activities that do not also involve radical improve the product or process without radically
alteration of core activities. Incremental changes altering their paradigms. For instance, Sony
involve incremental adjustments to core or com- offered 8 mm Camcorder versions TR31, TR71,
plementary activities. Our three-part typology is TR81 and TR101 during the early 1990s. All
similar to Henderson and Clark’s (1990) four- camcorders have the same broad functionalities
part framework of technological changes, which but each version has some aspect that differs
distinguishes between incremental, modular, from the previous model.
architectural, and radical innovation. Radical and
modular innovation in the Henderson and Clark
Hypotheses
framework both involve changes to core tech-
nology and, therefore, are equivalent to Encompassing, complementary, and incremental
encompassing change in our typology. Architec- changes often have substantially different effects
tural and incremental innovation in the Henderson on a firm’s existing knowledge and, in turn, on
and Clark framework may be either complemen- the methods that firms use to acquire new knowl-
tary or incremental innovation in our typology, edge. The following hypotheses address the
depending on whether complementary technology methods that firms will tend to use to acquire
changes. The primary advantage of our three-part know-how needed for different types of change.
approach is that it incorporates complementary We begin with encompassing changes, next dis-
technology and is empirically tractable in situ- cuss complementary changes, and then turn to
ations involving multiple firms and innovations. incremental changes.
Some examples may help clarify the concepts Encompassing technological changes substan-
of encompassing, complementary, and incremen- tially alter the principles by which the product
tal change. As an encompassing change, consider achieves its functionality (Dewar and Dutton,
Henderson and Clark’s (1990) example of the 1986). New technology that is radically different
central air conditioner as the replacement for the from the previous technology renders many
electric fan. Both products target similar cus- resources and competencies of existing businesses
tomers to fulfill similar needs but the air- obsolete, so that the businesses must acquire new
conditioning unit uses different inputs, different capabilities if they wish to continue to operate
scientific principles, and different distribution (Cooper and Schendel, 1976; Abernathy and
channels when compared to the electric fan. In Utterback, 1978; Ettlie, Bridges, and O’Keefe,
an earlier period, electrical fans embodied 1984; Abernathy and Clark, 1985; Tushman and
encompassing changes with respect to fans pow- Anderson, 1986). Encompassing changes directly
ered by manual or other sources. In our terms, affect the core technology that a firm employs,
in both cases the core activities of firms produc- so that the impact of such obsolescence is often
ing the fans have undergone encompassing extreme and may threaten business survival.
change. Complementary technological change We expect that firms will often form inter-
occurs when the core activities required remain organizational relationships in order to acquire the
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1067

know-how needed for encompassing technological and process know-how, but lack industry-specific
change. Many authors have argued that firms knowledge and financial resources required for
have limited repertoires of key routines (e.g., independent entry and often are willing to become
Cyert and March, 1963; Rumelt, 1974; Nelson partners of incumbent firms (Mitchell and Singh,
and Winter, 1982; Prahalad and Bettis, 1987; 1992). The growing recognition of these benefits
Dosi, 1988; Ramanujam and Varadarajan, 1989). has given rise to the idea of interfirm virtual
Firms faced with encompassing change often corporations as a means for businesses to adapt in
require new routines that are outside the realm changing environments (Byrne, 1993). Although
of their internal R&D capabilities. The firms must there are substantial limits to interorganizational
acquire knowledge of new scientific principles, adaptability (Miles and Snow, 1992; Singh and
manufacturing processes, and marketing Mitchell, 1996), such interfirm relationships play
approaches. Technology development within the key roles in modern markets.
firm is often both slow and expensive under these We expect that firms will tend to use equity-
conditions, because the expertise developed by based interorganizational relationships for
the firm will often not be relevant to the change. encompassing technological changes. Because the
Acquiring firms with the requisite technology is encompassing change will often diminish the
an alternative means of acquiring resources, but value of a firm’s capabilities, the firm must focus
several factors limit acquisitions as alternatives on acquiring new technology with a view to
to interorganizational relationships. First, the rebuilding its core activities. Given the tacit na-
acquiring firm must possess the financial and ture of much of the knowledge that the firm
management resources needed for acquisition. needs to acquire and the high stakes that the firm
Second, potential target firms may have an assort- must attach to the outcome of this decision,
ment of resources, including many capabilities equity-based relationships provide a framework
that the acquiring firm does not need. Third, for both learning about new technology and for
technological change may occur more rapidly protecting technology acquisition (Nueno and
than firms can identify acquisition targets and Oosterveld, 1988; Klein, 1988; Pisano and Teece,
negotiate acquisitions. Fourth, institutional 1989; Mitchell, 1991). On the learning dimension,
reasons such as external regulatory barriers may the organization must learn about engineering and
preclude acquisitions. Interorganizational relation- technological principles that are new to the firm,
ships provide the particular external capability which often requires close communication and
that the firm requires, without encountering the interaction with another firm. On the protection
financial, managerial, temporal, and institutional dimension, equity-based relationships usually
barriers to acquisition. involve administrative hierarchies for monitoring
Many analysts argue that interorganizational and control and thereby offer a guard against
relationships often provide viable alternatives to opportunistic behavior by a partner (Williamson,
internal development and firm acquisition as a 1985; Pisano, 1990; Jorde and Teece, 1990).
means of acquiring new skills, particularly in
dynamic environments in which firms lack time Hypothesis 1: Firms will tend to use equity-
needed to develop internal skills or identify and based interorganizational relationships to
negotiate with potential acquisition targets (e.g., acquire know-how when technological change
Richardson, 1972; Mariti and Smiley, 1983; Jorde is encompassing.
and Teece, 1990; Williamson, 1991; Martin, Swa-
minathan, and Mitchell, 1998). From the point Complementary changes have different influences
of view of potential interorganizational partners, on a firm’s capabilities than do encompassing
industry incumbents will sometimes be willing to changes. Complementary technological changes
ally together to address the uncertainties and sometimes trigger adjustments in a firm’s core
expense of encompassing changes. Moreover, routines. Taking the tire example used earlier, the
encompassing technological change often pro- car manufacturer must accommodate the new tire
vides the foundation for entry to an industry by specifications into the automotive design and
start-up firms and by established companies that adjust to the new weight and friction character-
undertake diversifying entry from outside the istics of the tires. The changes have less drastic
industry. Entrants often possess valuable product effects on a firm’s resource base than do
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
1068 A. Nagarajan and W. Mitchell

encompassing changes, however, and in some competitive advantage arising out of strategic or
cases the core activities will remain unchanged efficiency enhancements in complementary activi-
even when a radical change takes place in the ties. Nonequity participation serves as reassurance
complementary technology, so that most of the and commitment by one party to the other.
firm’s capabilities will remain intact and relevant.
Nonetheless, although the changes may appear to Hypothesis 2: Firms will tend to use non-
be minor, a firm must manage the change in the equity interorganizational relationships to
complementary technology carefully, for at least acquire know-how when technological change
two reasons. First, the firm must adjust to the is complementary.
new technology in order to maintain competitive
positioning. Second, changes in complementary Finally, we expect firms to use internal R&D for
technologies may provide a window of oppor- most incremental changes. Although adapting to
tunity for the firm to improve the core technology incremental changes is critically important to firm
of the product. Therefore, firms often benefit by survival (Banbury and Mitchell, 1995), incremen-
encouraging, moderating, and exploiting comple- tal changes usually do not drastically alter the
mentary technological change. competencies required by firms in the core and
We expect nonequity relationships to be more support industries. Instead, the changes tend to
common than internal R&D and equity-based reinforce the competitive positions of firms,
relationships in cases of complementary techno- because they build on their core competencies
logical change. Internal R&D is often ill equipped (Abernathy and Clark, 1985) or they are com-
to develop complementary technologies because petence enhancing (Tushman and Anderson,
the firm’s capabilities often focus on the core 1986). Because these innovations build on exist-
technologies and the activities defined by the core ing architectural and component knowledge, firms
technological system. Equity-based associations, compete among themselves to innovate incremen-
meanwhile, may require substantial financial and tally so as to obtain temporary competitive advan-
managerial commitments that firms can not tage. Firms use incremental innovations to capture
reverse easily. Complementary technologies, transient market or cost advantages, as well as
while enhancing the core product functionality, to build on their existing knowledge and skill
often draw upon different principles and capabili- base to attain competitive advantages (Cohen and
ties. The extent to which the complementary Levinthal, 1990). Firms tend to use their existing
technologies apply technologies that are unrelated competence base to lead or respond to incremen-
to the core product’s technology determines the tal change, and often are able to rely on internal
potential for economies of scope (Teece, 1980). R&D to develop specific innovations that enhance
Unless there are substantial transactions cost a firm’s particular capabilities.
advantages to internalizing the development of
complementary technologies and potential for Hypothesis 3: Firms will tend to use internal
achieving economies of scope, firms would tend R&D to acquire know-how when technological
to use nonequity relationships when faced with change is incremental.
complementary technological change. Nonequity
relationships allow the firms to monitor techno- In summary, we expect that firms will often
logical advance that could enhance the product’s acquire know-how needed for encompassing tech-
effective utilization while retaining the option of nological change through equity-based inter-
soliciting other sources of technology. From the organizational relationships, complementary tech-
point of view of a provider of complementary nological changes through nonequity
technology, a nonequity relationship can add relationships, and incremental changes through
credibility to the innovation especially if the firm internal R&D. We believe that the predictions
is relatively unknown (Shan, 1990; Mitchell, concerning technology acquisition via interfirm
1991). Further, given the inherent risk and cost relationships will hold in many settings. Relevant
of new technology, the close involvement of their settings include industries in which there are
technology suppliers is desirable. Such arrange- many small firms that lack the resources to under-
ments provide the flexibility required in high-risk take frequent acquisitions, cases in which tech-
situations while allowing firms to exploit any nology changes more rapidly than firms can nego-
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1069

tiate business acquisition agreements, and in cases technologies as alternatives to the ultrasound tech-
when the capabilities required for encompassing nology used as the core scientific basis in the
and complementary changes exist within busi- original Dornier lithotripter. Examples include the
nesses and other organizations that have many Coumadin green laser fiber lithotripter, pulse-
capabilities beyond those that are the target of dyed laser lithotripter, and Pulsolith. Laser-based
the particular innovation. These conditions exist approaches such as pulse-dyed laser lithotripsy
within the lithotripsy industry, which is the proved to be an effective technique in patients
empirical setting for this study, and in many other with difficult bile duct stones according to
industrial sectors. researchers in Germany (Brambs et al., 1996).
These procedures are less invasive, less painful,
and less expensive than the initial product based
NEW TECHNOLOGY IN THE on ultrasound technology. Other encompassing
MEDICAL LITHOTRIPSY INDUSTRY technological changes include the electromechani-
cal impactor and the echo locator. The develop-
The lithotripter is a medical device that uses ment of the electrohydraulic lithotripsy probes
electrohydraulically generated shock waves to dis- was a complementary technological innovation
integrate kidney stones into tiny fragments that that affected the component linkages of the prod-
easily pass in urine. Technological innovation in uct but left the core concepts of the lithotripter
the medical industry has had dramatic impact on intact. According to a study conducted at the
surgical therapies used for kidney stones. Prior Washington University School of Medicine, the
to the introduction of the lithotripter, removal of development of the 1.9F electrohydraulic probes
kidney stones involved painful surgery and sev- provides the urologist with a safe, highly effec-
eral days in hospital. The development of extra- tive, and inexpensive method for performing
corporeal shock wave lithotripsy (ESWL) has lithotripsy and is the only form of lithotripsy
successfully changed kidney stone therapy. sufficiently malleable to allow routine access for
According to published reports, ESWL is now the treatment of lower pole renal calculi (Elashry
the preferred treatment for approximately 80–85 et al., 1996). Other complementary innovations
percent of kidney stones (Lingeman, 1996). The in the lithotripsy industry include the mobile
entire procedure lasts less than an hour and most lithotripter and the catheter acoustic transponder.
patients feel little discomfort. A collaborative Incremental changes include modifications to the
effort of Dornier Medical Systems, Inc. and the lithotripter that allow biliary procedures as well
Urology Department of Ludwig Maximillians as renal applications. Our goal is to determine
University of Munich required 6 years of research whether firms used systematically different
and investigation before the commercialization of methods to acquire know-how needed to produce
the first lithotripter in 1983 (Coleman, Saunders, lithotripters that incorporated encompassing, com-
and Palfrey, 1987). The Dornier lithotripter plementary, and incremental technological
received FDA approval for sale in the United changes.
States in December 1984. The rapid clinical We drew the sample for this study from an
acceptance of ESWL sparked other equipment extensive archival study of published sources and
manufacturers to design and test renal lithotripters a mailed survey to lithotripter manufacturers. We
that differ from the original Dornier system reviewed all references to ‘lithotripsy’ or ‘litho-
(Gallivan, 1986). Entrants to the worldwide litho- tripters’ in the imaging industry newsletter,
tripsy industry include a wide range of businesses SCAN, plus the Nexis-Lexis, ABI Inform, and
from the United States, several European coun- InfoTrac electronic data bases. We searched these
tries, Japan, and Israel, including entrepreneurial sources for references relating to product inno-
start-up firms, established medical device manu- vations and for information concerning how firms
facturers, and diversifying entrants from outside acquired know-how required for the innovations.
the medical sector. We classified the acquisition methods as equity-
Many encompassing, complementary, and based relationships, nonequity relationships, and
incremental technological changes occurred internal R&D based on the information in the
within the lithotripsy industry between 1989 and published sources. The initial data set for the
1991. Some encompassing innovations use laser analysis included 25 product innovations between
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
1070 A. Nagarajan and W. Mitchell

1989 and late 1991, totaling 63 cases of know- provide a clear picture of the cases. The logit
how acquisition. The cases involved 24 com- analysis provides confidence that the descriptive
panies, including nine lithotripter manufacturers results are robust.
and 15 suppliers. Table 1 compares the types of technological
We mailed a questionnaire listing the inno- change with the methods the firms used to acquire
vations to the senior managing executives of the know-how. We use log-likelihood ratio ␹2 sta-
nine lithotripter manufacturers, requesting their tistics and Fisher’s exact test for univariate tests
opinion on whether the innovation affected the of the hypotheses, based on the contingency table
core concepts or component systems of the litho- in Table 1. These tests demonstrate the relation-
tripter. We received responses from three of the ship between the type of technological change
nine manufacturers, two of which we could use and the methods that firms used to acquire tech-
for the study. (The third response simply noted nology. Log-likelihood ratio ␹2 tests of the hom-
an interest in our study.) The two-firm response ogeneity of a contingency table are sometimes
somewhat conditions the interpretations of the misleading when the sample size is small. There-
study’s results, as one must be cautious concern- fore, we also calculated Fisher’s exact test based
ing the reliability of the classifications of techno- on the hypergeometric distribution.
logical changes. Nonetheless, there was substan- The distribution of observations in Table 1 is
tial agreement about how the changes affected consistent with the proposition that the degree
core and component lithotripsy systems, both in and focus of technological change influence the
the responses, as we discuss below, and in written methods firms use to acquire technology. Twenty-
reports of the changes. We believe that the com- nine of the 44 cases (66%) conform to the pre-
bination of archival and survey research provides dicted relationships. Both the ␹2 test for hom-
a solid base for the study. ogeneity and Fisher’s exact test reject the null
The responses from the firms allowed us to hypothesis of no association among the variables.
classify innovations as encompassing, comple- As sensitivity analysis, we excluded the 14 cases
mentary, or incremental. We classed the inno- of incremental innovations in order to eliminate
vations as encompassing when the respondents the possibility that the incremental innovations,
stated that an innovation changed the core con- which realized the greatest conformity with the
cepts. We classified the innovation as comple- predictions, might cause the lack of independence
mentary when the respondents stated that an inno- in the results. Therefore, we tested for homogen-
vation changed the activities in the component eity using a 2 × 3 cell contingency table contain-
technological system and did not change the core ing only the observations relating to the nine
concepts in the lithotripter. We classified the encompassing and 21 complementary innovations.
innovation as incremental when the respondents The log-likelihood ratio ␹2 statistic (␹2 = 13.7, 2
stated that neither the core nor component con- d.f.; probability = 0.001) and the Fisher’s exact
cepts changed. The classification based on the test (probability = 0.003) for the restricted sample
two responses was the same for 19 (6 confirm the conclusion that there is a relationship
encompassing, 10 complementary change, 3 between the method firms use to acquire tech-
incremental) of the 25 innovations (p ⬍ 0.05, nology and the nature of technological change.
binomial test for independence). We used the 19
innovations with consistent classification for the
Univariate analysis: General support
analysis, which involved 44 cases of know-how
acquisition (9 encompassing, 21 complementary, The results in Table 1 support each hypothesis.
and 14 incremental). The Appendix lists the inno- Hypothesis 1 predicted that firms will tend to use
vations. equity-based interorganizational relationships to
acquire know-how required for encompassing
technological change. Of the nine instances of
ANALYSIS encompassing technological change in the indus-
try, firms used equity participation for six cases
We first test the hypotheses with univariate sta- (67%) and in-house R&D for three cases. It is
tistics based on contingency tables and then esti- noteworthy that there were no nonequity relation-
mate binary logit models. The contingency tables ships for the encompassing changes, so that the
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1071
Table 1. Case-wise comparison: Types of technological change and methods used to acquire know-how (The
cases on the diagonal, reported in italic type face, are consistent with the predictions.)

Methods used to acquire technology

Technological Equity-based Nonequity Internal Total

change relationships relationships R&D cases As predicted

Encompassing 6 0 3 9 67%
Complementary 3 12 6 21 57%
Incremental 1 2 11 14 79%
Total cases 10 14 20 44 66%

Log-likelihood ratio ␹2 statistic = 23.3 (4 d.f., probability ⬍ 0.001)

Fisher’s exact test: probability ⬍ 0.001

entire set of observations in the encompassing incremental technological change (79%) came
technology category involved equity participation, through in-house R&D.
either through internal R&D or relationships with
other firms. This pattern suggests that
Univariate analysis: Nonconforming cases
encompassing technological changes redefine the
rules of competition, such that the stakes are high We examined the individual cases in our study in
when a firm responds to the change. Therefore, order to attempt to understand the variance in the
firms are likely to use technology acquisition outcomes. Examining the cases that do not adhere
modes that involve partial or complete ownership to the predictions provides further insight concern-
in order to achieve effective monitoring and con- ing factors that may influence firm behavior. Two
trol when they acquire technology needed for factors, industry entry and firm size, appear to be
encompassing changes. particularly relevant. We first consider the unex-
Table 1 provides moderate support for Hypoth- pected outcomes for encompassing change, next
esis 2, which predicted that nonequity interorgani- address complementary change, and then turn to
zational relationships will be common in cases incremental change.
involving complementary technologies. Over half Three of nine cases in Table 1 do not fit
the responses to such changes involved nonequity the prediction of equity-based relationships for
relationships (12 of 21 cases; 57%). The widest encompassing change, all involving internal R&D.
variation in responses occurred along this seldom- When we examined the three cases, we found that
researched area of technological change, as three all involved diversifying entrants that used internal
firms acquired the required know-how to adapt expertise gained from activities in other medical
to such change through equity-based relationships equipment industries to enter the lithotripter indus-
and six firms used their in-house R&D facilities. try. The implication of this observation is that
We discuss implications of this variance later in internal experience in other industrial sectors pro-
this section of the paper. At this point, though, vided the entry point to the lithotripter industry,
we will note that the results show that comple- so that the firms did not need relationships with
mentary technological changes are common other businesses. Instead, businesses can some-
components of the competitive landscape. The times obtain new capabilities from other parts of
analysis demonstrates the importance of consider- their firms. This outcome follows the interdi-
ing the broader context of the product when visional expertise logic of Nelson (1959) and
studying the effect of technological change on Argyres (1996).
firms. Table 1 reports the greatest divergence in choice
Table 1 provides strong support for Hypothesis of method of technology acquisition when techno-
3, which predicted that firms will tend to use logical change is complementary, with nine of 21
internal development to acquire know-how when cases having nonpredicted outcomes. There were
technological change is incremental. Consistent three instances of equity-based relationships. All
with the prediction, 11 of the 14 instances of three cases involved small lithotripter manufac-
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
1072 A. Nagarajan and W. Mitchell

turers that received equity investments as well as is appropriate for assessing the effects of categ-
complementary technology from larger firms. The orical variables on a dichotomous response vari-
likely explanation for the result is that the small able, providing a categorical response analog to
firms needed the financial resources of equity regression models for continuous response vari-
relationships as well as technical expertise from ables. We estimated binary logit models of the
their partners. The partners, in turn, likely viewed effect of technological change on each method
the equity investment as a form of option on the of technology adoption. At the firm level, the 44
results of the smaller firms’ efforts. There were cases of technology adoption involved 59 inci-
six cases in which firms used internal R&D to dents, because several cases involved relation-
acquire complementary know-how. Five of the six ships between two lithotripter manufacturers. We
cases involved existing lithotripter manufacturers omitted the case that involved the nonequity
that possessed sufficient expertise to adapt to the research consortium because such consortia are
particular complementary changes. As in the beyond the scope of the study. Thus, the firm-
encompassing technology exceptions that we level data set for the binary logit models consisted
described in the previous paragraph, the existence of 51 observations, including 14 instances of
of needed skills inside the companies obviated the equity-based relationships, 17 cases of nonequity
need for external relationships for these firms. It relationships, and 20 cases of internal R&D.
is notable that the firms used all three methods of Table 2 shows the firm-level cross-tabulation of
technology acquisition to acquire complementary the method of technology acquisition with the
know-how. Future studies could fruitfully examine nature of technological change. As in the case-
interorganizational relationships in the context of wise comparison of Table 1, the firm-level results
complementary technological change in finer in Table 2 are consistent with the three hypoth-
detail. Nonetheless, despite the variation, a eses.
majority of the cases were consistent with the We analyzed three binary logit models to verify
predicted relationship between complementary the interpretation of the contingency tables. Each
technological change and nonequity methods of logit model had one method of technology acqui-
technology acquisition. Thus, although there will sition as the dependent variable (Equity-based
be substantial variation owing to firm-level differ- relationships, Nonequity relationships, and
ences and other factors, the general tendencies Internal R&D). The explanatory variables
appear to hold. included the three types of technological change
Three of 14 cases in Table 1 did not conform (Encompassing change, Complementary change,
to the expected use of internal R&D when techno- and Incremental change), whether the firm was
logical change was incremental. One case entering the lithotripsy industry when it acquired
involved an equity-based relationship and two the know-how (Industry entrant), and the log of
cases involved nonequity relationships. Small corporate sales (Firm size). As we noted above,
firms created two of the three relationships, sug- our examination of the data suggested that indus-
gesting that resource scarcity may have been a try entry and firm size might affect acquisition
motivation for the external relationships. The third methods. Previous research suggests that size
case involved a nonequity research consortium of sometimes affects firm behavior (Scherer, 1980),
eight firms. The use of the consortium for although the direction of the effect is ambiguous.
incremental technological change might indicate Larger firms may have greater resources at their
that the firms were learning about new capabilities command and therefore are better able to under-
before making commitments to more substantial take internal R&D, but large firms often have
projects. substantial organizational inertia that may inhibit
internal R&D. We used corporate sales in the
year prior to the year in which the firms
Multivariate analysis
announced the technology acquisition. We used
In addition to the contingency table analysis, we the value of corporate sales in its log form so
also estimated binary logit models of firm-level as to minimize distortion in the results due to
technology adoption methods in order to control particularly large firms. We found similar results
for the potentially confounding effect of business for the predicted variables in sensitivity analysis
size and industry entry. The binary logit method that used untransformed sales, but the models
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1073
Table 2. Firm-level comparison: Types of technological change and methods used to acquire know-how (The
cases on the diagonal, reported in italic type face, are consistent with the predictions.)

Methods used to acquire technology

Technological Equity-based Nonequity Internal Total

change relationships relationships R&D cases As predicted

Encompassing 9 0 3 12 75%
Complementary 4 16 6 26 62%
Incremental 1 1* 11 13 85%
Total 14 17 20 51 71%

Log-likelihood ratio ␹2 statistic = 35.5 (4 d.f., probability ⬍ 0.001)

Fisher’s exact test: probability ⬍ 0.001.
*We omitted one Incremental/Nonequity case in the firm-level data because the case was a research consortium involving
eight firms.

using log sales achieved somewhat greater overall for complementary technological change is posi-
fit. Table 3 reports descriptive statistics and corre- tive and significant.
lations. To test Hypothesis 3, the third model used
Table 4 shows the results of the binary logit internal R&D as the dependent variable. As pre-
models of firms’ responses. All models are sta- dicted, the parameter estimate for incremental
tistically significant, based on the log-likelihood technological change is positive and significant.
␹2 statistics reported at the foot of the table. The The other influences reported in Table 4 also
logit results support each of the three hypotheses. are interesting. Larger firms are less likely to
To test Hypothesis 1, the first model in Table undertake either equity-based or nonequity inter-
3 uses equity-based relationships as the dependent organizational relationships and marginally more
variable. The parameter estimate for likely to undertake internal R&D, although the
encompassing technological change is positive internal R&D coefficient is not statistically sig-
and significant, as predicted. nificant. In addition, industry entrants tend to
To test Hypothesis 2, the second model used avoid equity-based relationships. In this study,
nonequity relationships as the dependent variable. the latter result likely stems from the high inci-
There were no cases where firms used nonequity dence of diversifying entry by established firms,
relationships to address encompassing technologi- which have less need than start-up companies of
cal change. Therefore, there were four explanatory the financial resources offered by equity relation-
variables in the model, including complementary ships.
change, incremental change, firm size, and indus- We carried out several sensitivity analyses of
try entrant. As predicted, the parameter estimate the data. We obtained similar results for the

Table 3. Summary statistics and product moment correlations

Variable Mean S.D. Range 1 2 3 4 5 6 7 8

1. Equity-based relationships 0.27 0.45 0–1 1 −0.43 −0.49 0.59 −0.28 −0.26 −0.18 0.08
2. Nonequity relationships 0.33 0.48 0–1 −0.43 1 −0.57 −0.39 0.61 −0.32 0.10 0.19
3. Internal R&D 0.39 0.49 0–1 −0.49 −0.57 1 −0.16 −0.34 0.54 0.07 −0.26
4. Encompassing change 0.24 0.43 0–1 0.59 −0.39 −0.16 1 −0.57 −0.32 0.05 0.36
5. Complementary change 0.51 0.50 0–1 −0.28 0.61 −0.34 −0.57 1 −0.60 0.00 −0.02
6. Incremental change 0.25 0.44 0–1 −0.26 −0.32 0.54 −0.32 −0.60 1 −0.05 −0.33
7. Firm size 4.10 2.45 0–10.5 −0.18 0.10 0.07 0.05 0.00 −0.05 1 −0.10
8. Industry entrant 0.51 0.50 0–1 0.08 0.19 −0.26 0.36 −0.02 −0.33 −0.10 1

 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
1074 A. Nagarajan and W. Mitchell
Table 4. Results of binary logit models of firm-level responses (51 cases) (The cases on the diagonal, reported
in italic type face, are the predicted relationships.)

Independent variables Dependent variable: Method used to acquire technology

1. Equity 2. Nonequity 3. Internal

based relationships R&D
relationships

Encompassing change 2.720** −1.581**

(1.218) (0.930)
Complementary change −0.518 1.257** −1.663**
(0.808) (0.619) (0.771)
Incremental change −1.328 −1.810* 1.302*
(1.202) (1.105) (0.917)
Firm size (log corporate sales) −0.340** −0.188** 0.108
(0.193) (0.109) (0.138)
Model log-likelihood ␹2 31.52*** 15.87*** 18.56***
(4 d.f.) (3 d.f.) (4 d.f.)

*p ⬍ 0.10; **p ⬍ 0.05; ***p ⬍ 0.01 (one-tailed tests; standard errors in parentheses).

predicted effects when we omitted the firm size nology acquisition from external sources. This
and industry entrant variables. We also tested for research shows that the intrinsic composition of
influences of acquisition frequency and acqui- technological change in terms of its degree and
sition date in sensitivity analyses, finding no sys- focus also affects technology acquisition patterns.
tematic patterns of acquisition methods by acqui- The results are consistent with Henderson and
sition year or among firms that participated in Clark’s (1990) conclusion that it is necessary to
more than two cases of know-how acquisition. look beyond the core set of activities defined by a
Therefore, we believe that the reported results product’s technology and include complementary
are robust. activities to better understand the competitive
effects of innovation in an industry. The results
also are consistent with Mitchell and Singh’s
CONCLUSION (1992) finding that supporting assets play a key
role in alliance formation. The frequent formation
This research examines the association between of interfirm alliances for encompassing and com-
types of technological change and the methods plementary technological changes suggests that
that firms use to acquire new know-how. We firms often use supporting assets in relationships
demonstrate that the firms in our sample of the with other firms when confronted by significant
medical lithotripsy industry tended to use equity- technological change. Thus, rather than view a
based interorganizational relationships to acquire single product-development organization as the
know-how for encompassing changes, nonequity central player in the product-development process,
interorganizational relationships for complemen- our analysis emphasizes that innovation takes
tary changes, and internal R&D for incremental place as a series of complementary activities
changes. The results provide insight into tech- among firms.
nology acquisition strategies used to manage tech- The results also highlight new aspects of the
nological change. process by which interorganizational relationships
The results extend the existing literature con- provide an evolutionary conduit for the diffusion
cerning types of technological change and of capabilities through an industry. Drawing on
methods of technology acquisition. Pisano (1990) ideas developed by Nelson and Dosi (1993) and
found that transaction cost factors such as small- Nelson (1995), Singh and Mitchell (1996) argued
numbers bargaining hazards and firm-level factors that interfirm relationships are part of an evolu-
such as R&D experience may motivate tech- tionary process of industrial change. The main
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1075

elements of such an evolutionary process include ner in which firms acquire technology is a mosaic
sources of new ideas, means by which the ideas of organizational and interorganizational arrange-
diffuse through portions of an industry, barriers ments.
to universal diffusion of the ideas, and means by The study highlights the importance of inter-
which firms that do not adopt successful new organizational relationships in the process of tech-
ideas will ultimately fail. This study directly nological change. Firms in most modern indus-
addresses the diffusion element of this evolu- tries must contend with the frequent technological
tionary process, showing that equity and nonequ- change. The sources of innovation and the
ity interorganizational relationships provide com- capacity to commercialize effectively are the cor-
mon means for firms to obtain new ideas nerstones of corporate and economic progress.
concerning encompassing and complementary When there is technological change, firms must
technologies. Implications for imperfect diffusion adapt to the new environment in a manner that
are also implicit in this study, because almost all limits costs while permitting rapid response to
interorganizational relationships involve only two the changing environment. Without some measure
firms or, at most, a small proportion of the firms of interfirm cooperation and planning, firms will
in an industry. Thus, new ideas that arise in not be able to make many supplementary and
a single organization or in a small number of complementary investments needed to develop
organizations will diffuse through only a portion and commercialize new technologies.
of an industry in a given period, so long as firms
require interorganizational contact to monitor and
learn about the use of the ideas. Further study ACKNOWLEDGEMENTS
could usefully develop our understanding of such
evolutionary processes. We greatly appreciate comments and suggestions
Several additional avenues for further research from two Strategic Management Journal
seem fruitful, concerning environmental and firm- reviewers. We also appreciate suggestions from
level factors. Investigating how firms respond to reviewers for the 1994 meetings of the Academy
exogenous environmental variations within com- of Management TIM division, at which we
plementary technologies would add valuable presented an earlier version of this paper.
insight to the technological challenge confronting
firms. Other environmental factors that might
influence the nature of the technology association REFERENCES
include value added by the complementary
activity, the pace of technological change, and Abernathy, W. J. and K. B. Clark (1985). ‘Innovation:
technology life cycles. In addition to exogenous Mapping the winds of creative destruction’,
Research Policy, 14, pp. 3–22.
environmental factors, firm-level heterogeneity is Abernathy, W. J. and J. G. Utterback (1978). ‘Patterns
likely to affect technology acquisition methods. of industrial innovation’, Technology Review, 80,
Differences among firms include at least two pp. 97–107.
components, including their technological knowl- Argyres, N. (1996). ‘Capabilities, technological diversi-
edge and their network positions. A firm’s prior fication and divisionalization’, Strategic Manage-
ment Journal, 17(5), pp. 395–410.
technological knowledge will constrain the type Arrow, K. (1974). The Limits of Organization. W. W.
and manner of capabilities it seeks to develop Norton, New York.
(Nelson and Winter, 1982; Teece and Pisano, Banbury, C. M. and W. Mitchell (1995). ‘The effect
1994). Similarly, a firm’s experience in dealing of introducing important incremental innovations on
with external partners is an integral component market share and business survival’, Strategic Man-
agement Journal, Summer Special Issue, 16, pp.
of its stock of skills (Harianto and Pennings, 161–182.
1990). Firms learn how to deal with alliance Brambs, H. J., S. H. Duda, A. Rieber, M. Scheulen
partners, what to expect from interorganizational and C. D. Claussen (1996). ‘Treatment of bile duct
interaction, and how best to utilize the alliance stones: Value of laser lithotripsy delivered via percu-
to their advantage. Future research could examine taneous endoscopy’, European Radiology, 6, pp.
734–740.
how a firm’s historical technological and net- Byrne, J. A. (8 February 1993). ‘The virtual corpo-
working experience affects its methods of acquir- ration’, Business Week, No. 3304, pp. 98–103.
ing technology. It is clear, though, that the man- Christensen, C. M. and J. L. Bower (1996). ‘Customer
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
1076 A. Nagarajan and W. Mitchell
power, strategic investment, and the failure of lead- relationship revisited’, Journal of Law, Economics
ing firms’, Strategic Management Journal, 17(3), and Organization, 4, pp. 199–212.
pp. 197–218. Lieberman, M. B. and D. B. Montgomery (1988).
Cleland, D. I. and K. M. Bursic (1992), Strategic ‘First-mover advantages’, Strategic Management
Technology Management: Systems for Products and Journal, Summer Special Issue, 9, pp. 41–58.
Processes. AMACOM, New York. Lingeman, J. E. (1996). ‘Lithotripsy and surgery’, Sem-
Cohen, W. M. and D. Levinthal (1990). ‘Absorptive inar on Nephrology, 16, pp. 487–498.
capacity: A new learning perspective on learning Mariti, P. and R. H. Smiley (1983). ‘Co-operative
and innovation’, Administrative Science Quarterly, agreements and the organization of industry’, Jour-
35, pp. 128–152. nal of Industrial Economics, 31(4), pp. 437–451.
Coleman, A. J., J. E. Saunders and E. L. H. Palfrey Martin, X., A. Swaminathan and W. Mitchell (1998).
(January/February 1987). ‘The destruction of renal ‘Organizational evolution in an interorganizational
calculi by external shock waves: Practical operation environment: Vertical, competitor, and community-
and initial results with the Dornier lithotripter’, Jour- level influences on supplier expansion to foreign
nal of Medical Engineering and Technology, 11, pp. markets’, Administrative Science Quarterly, forth-
4–10. coming.
Cooper, A. C. and D. Schendel (1976). ‘Strategic Miles, R. E. and C. C. Snow (1992). ‘Causes of failure
response to technology threats’, Business Horizons, in network organizations’, California Management
19, pp. 61–69. Review, 34(4), pp. 53–72.
Cyert, R. and J. March (1963). A Behavioral Theory Mitchell, W. (1991). ‘Probability and timing of expan-
of the Firm. Prentice-Hall, Englewood Cliffs, NJ. sion by industry incumbents following evolutionary
Dewar, R. D. and J. E. Dutton (1986). ‘The adoption major product innovation’. In P. Shrivastava, A.
of radical and incremental innovations: An empirical Huff and J. Dutton (eds.), Advances in Strategic
analysis’, Management Science, 32, pp. 1422–1433. Management, Vol. 7. JAI Press, Greenwich, CT, pp.
Dosi, G. (1988). ‘Procedures and microeconomic 43–60.
effects on innovation’, Journal of Economic Litera- Mitchell, W. and K. Singh (1992). ‘Incumbents’ use
ture, 26, pp. 1120–1230. of preentry alliances before expansion into new
Elashry, O. M., R. B. DiMeglio, S. Y. Nakada, E. M. technical subfields of an industry’, Journal of Eco-
McDougal and R. V. Clayman (1996). ‘Intracorpo- nomics, Behavior and Organization, 18(3), pp.
real electrohydraulic lithotripsy of ureteral and renal 347–372.
calculi using small caliber (1.9F) electrohydraulic Mitchell, W. and K. Singh (1993). ‘Death of the
lithotripsy probes’, Journal of Urology, 156, pp. lethargic: Effects of expansion into new technical
1581–1585. subfields of an industry on performance in a firm’s
Ettlie, J. E., W. P. Bridges and R. D. O’Keefe (1984). base business’, Organization Science, 4(2), pp.
‘Organizational strategy and structural differences 152–180.
for radical vs incremental innovation’, Management Mueller, D. C. and J. E. Tilton (1969). ‘R&D cost as
Science, 30, pp. 692–695. a barrier to entry’, Canadian Journal of Economics,
Friar, J. and M. Horwitch (1985). ‘The emergence of 2, pp. 570–579.
technology strategy: A new dimension of strategic Nelson, R. R. (1959). ‘The simple economics of basic
management’, Technology in Society, 7(2/3), pp. scientific research’, Quarterly Journal of Economics,
143–178. 74, pp. 297–306.
Gallivan, M. (20 July 1986). ‘The race for competing Nelson, R. R. (1995). ‘Recent evolutionary theorizing
lithotripters heats up’, Hospitals, 60, pp. 70–71. about economic change’, Journal of Economic
Ghemawat, P. (1991). Commitment. The Dynamic of Literature, 33, pp. 48–90.
Strategy. Free Press, New York. Nelson, R. R. and G. Dosi (1993). ‘Evolutionary
Harianto, F. and J. M. Pennings (1990). ‘Technological theories in economics: Assessment and prospects’,
innovation through interfirm linkages’. In M. W. working paper WP-93-064, International Institute for
Lawless and L. R. Gomez-Meijia (eds.), Strategic Applied Systems Analysis, A-2361, Laxenburg, Aus-
Management in High Technology Firms. JAI Press, tria.
Greenwich, CT, pp. 15–42. Nelson, R. R. and S. G. Winter (1977). ‘In search of
Henderson, R. M. and K. B. Clark (1990). ‘Architec- useful theory of innovation’, Research Policy, 6, pp.
tural innovation: The reconfiguration of existing 36–76.
product technologies and the failure of established Nelson, R. R. and S. G. Winter (1982). An Evolutionary
firms’, Administrative Science Quarterly, 35, pp. Theory of Economic Change. Harvard University
9–30. Press, Cambridge, MA.
Jewkes, J., D. Sawers and R. Stillerman (1958). The Nueno, P. and J. Oosterveld (1988). ‘Managing techno-
Sources of Invention. Macmillan, London. logical alliances’, Long Range Planning, 21(3), pp.
Jorde, T. M. and D. J. Teece (1990). ‘Innovation 11–17.
and cooperation: Implications for competition and Penrose, E. T. (1958). Theory of the Growth of the
antitrust’, Journal of Economic Perspectives, 4, pp. Firm. Wiley, New York.
75–96. Pisano, G. P. (1990). ‘The R&D boundaries of the
Klein, B. (1988). ‘Vertical integration as organizational firm: An empirical analysis’, Administrative Science
ownership: The Fisher Body–General Motors Quarterly, 35, pp. 153–176.
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)
 Internal and External Technology Acquisition 1077
Pisano, G. P. and D. J. Teece (1989). ‘Collaborative Williamson, O. E. (1975). Markets and Hierarchies.
arrangements and global technology strategy: Some Free Press, New York.
evidence from the telecommunications equipment Williamson, O. E. (1985). The Economic Institutions
industry’, Research on Technological Innovation, of Capitalism. Free Press, New York.
Management and Policy, 4, pp. 227–256. Williamson, O. E. (1991). ‘Comparative economic
Prahalad, C. K. and R. P. Bettis (1987). ‘The dominant organization: The analysis of discrete structural alter-
logic: A new linkage between diversity and perform- natives’, Administrative Science Quarterly, 36, pp.
ance’, Strategic Management Journal, 7(6), pp. 269–296.
485–501. Winter, S. G. (1987). ‘Knowledge and competence as
Ramanujam, V. and P. Varadarajan (1989). ‘Research strategic assets’. In D. Teece (ed.), The Competitive
on corporate diversification: A synthesis’, Strategic Challenge. Harper & Row, New York, pp. 159–184.
Management Journal, 10(6), pp. 523–551.
Richardson, G. B. (1972). ‘The organization of indus-
try’, Economic Journal, 82, pp. 833–896.
Rosenberg, N. (1972). Technology and American Eco- APPENDIX: CATEGORIES OF
nomic Growth. Harper & Row, New York. TECHNOLOGICAL CHANGE AND
Rumelt, R. P. (1974). Strategy, Structure, and Eco- LITHOTRIPTER INNOVATIONS
nomic Performance. Harvard Business School Press,
Boston, MA. A. Encompassing innovations
Sahal, D. (1985). ‘Technological guideposts and inno-
vation avenues’, Research Policy, 14(2), pp. 61–82. 1. Coumadin green laser fiber
Scherer, F. M. (1980). Industrial Market Structure and 2. Electromechanical impactor
Economic Performance (2nd ed.). Rand McNally, 3. Pulse-dyed laser
Chicago, IL. 4. Pulsolith
Schumpeter, J. A. (1934). The Theory of Economic 5. Ultrasound imager (Echo locator)
Development. Harvard University Press, Cam-
bridge, MA. 6. Ursodeoxycholic acid
Scott, W. R. (1987). Organizations: Rational, Natural
and Open Systems. Prentice-Hall, Englewood B. Complementary innovations
Cliffs, NJ. 1. Catheter acoustic transponder
Shan, W. (1990). ‘An empirical analysis of organi- 2. Electrohydraulic lithoptripsy probe (1.9F)
zational strategies by entrepreneurial high technology
firms’, Strategic Management Journal, 11(2), pp. 3. Flashlamp dye laser
129–139. 4. Lithox system (contact dissolution system)
Singh, K. and W. Mitchell (1996). Precarious collabora- 5. Mobile ESW
tion: Business survival after partners shut down 6. Mobile lithotripter
or form new partnerships’, Strategic Management 7. Modular renal lithotripter
Journal, Summer Special Issue, 17, pp. 95–115.
Teece, D. J. (1980). ‘Economies of scope and the 8. ORACLE PTCA catheter
scope of the enterprise’, Journal of Economic 9. Second-generation sonographic piezoe-
Behavior and Organization, 1, pp. 223–247. lectric lithotripters (Piezolith 2300)
Teece, D. J. (1986). ‘Profiting from technological inno- 10. Triper XI
vation: Implications for integration, collaboration,
licensing, and public policy’, Research Policy, 15,
pp. 285–305. C. Incremental innovations
Teece, D. J. (1988). ‘Technological change and the 1. Biliary lithotripter
nature of the firm’. In G. Dosi, C. Freeman, R. 2. Hydrophones
Nelson, G. Silverberg and C. Soete (eds.), Technical 3. Therasonic system
Change and Economic Theory. Pinter, London, pp.
256–281.
Teece, D. J. and G. P. Pisano (1994). ‘The dynamic We omitted six innovations that had conflicting
capabilities of firms: An introduction’, Industrial survey responses concerning categories of
and Corporate Change, 3(3), pp. 537–556. technological change:
Thompson, J. (1967). Organizations in Action. 1. Diagnostic and therapeutic urological
McGraw-Hill, New York. pharmaceuticals
Tilton, J. E. (1971). International Diffusion of Tech-
nology: The Case of Semiconductors. Brookings 2. Human monoclonal antibodies
Institution, Washington, DC. 3. Laparoscopic cholecystectomy
Tushman, M. L. and P. Anderson (1986). ‘Technologi- 4. Organic chemical contact dissolution sys-
cal discontinuities and organizational environments’, tem
Administrative Science Quarterly, 31, pp. 439–465. 5. Percutaneous nephrostolithotomy
Utterback, J. M. and W. J. Abernathy (1975). ‘A
dynamic model of product and process innovation’, 6. Urological diagnosis/therapy table
Omega, 3(6), pp. 639–656.
 1998 John Wiley & Sons, Ltd. Strat. Mgmt J., 19: 1063–1077 (1998)