COMPLEMENTARY TECHNOLOGIES, KNOWLEDGE RELATEDNESS AND INNOVATION OUTCOMES IN HIGH TECHNOLOGY MERGERS AND ACQUISITIONS

Transcription

1 WPS-MGT COMPLEMENTARY TECHNOLOGIES, KNOWLEDGE RELATEDNESS AND INNOVATION OUTCOMES IN HIGH TECHNOLOGY MERGERS AND ACQUISITIONS Marianna Makri University of Miami School of Business Administration Jenkins Building, 414 L 5250 University Drive Coral Gables, FL Voice: (305) Fax: (305) [email protected] Peter J. Lane University of New Hampshire Whittemore School of Business & Economics Durham, NH Voice: (603) Fax: (603) [email protected] Michael A. Hitt Mays Business School Texas A&M University College Station, TX Voice: (979) Fax: (979) [email protected] 2007 by Makri, Lane and Hitt No part of this publication can be reproduced or transmitted in any form or by any means electronic, mechanical, photocopying or otherwise (including the Internet) without the written permission of the authors

2 COMPLEMENTARY TECHNOLOGIES, KNOWLEDGE RELATEDNESS AND INNOVATION OUTCOMES IN HIGH TECHNOLOGY MERGERS AND ACQUISITIONS ABSTRACT Building on research in organizational learning and technology management, this study develops and tests a knowledge-based model of acquirer-target relatedness and innovation performance in high technology mergers and acquisitions. Knowledge relatedness emphasizes both complementarity (supporting or facilitating knowledge bases) and similarity (overlapping knowledge bases). An examination of high-technology M&As provided support for theoretical arguments linking science and technology relatedness with post-merger innovation quantity and quality. Also, the interactions of technology complementarity with the product/market similarity between the acquiring and acquired firms and with acquirer diversification were found to affect innovation outcomes. The findings provide a richer understanding of the role of knowledge relatedness in post M&A innovation performance.

3 The speed of innovation and need for novel solutions in high technology industries often motivates firms to extend their resources and capabilities through mergers and acquisitions (M&As) (Hagedoorn, 2002). In the latter half of the 20 th century, M&As became a prominent strategy for many companies, large and smaller. This is exemplified by the fact that global mergers and acquisitions were valued at $3.6 trillion in 2006 (Cimilluca, 2007). Furthermore, the strategic use of acquisitions to acquire new knowledge and capabilities has become a well institutionalized corporate phenomenon (Uhlenbruck, Hitt & Semadeni, 2006; Vermeulen & Barkema, 2001). In a study of over 9,000 transactions between 1990 and 2000, Villalonga and McGahan (2005) found that the likelihood a firm would choose acquisition over other forms of collaboration increased with the technological resources of the potential target/partner. Technologically rich acquisition targets create opportunities for organizational learning by exposing the acquirer to new and diverse ideas (Ghoshal, 1987; Hitt et al., 1996). Research on high technology M&As has identified the relatedness of the buyer s and the target s technological knowledge as an important predictor of post-merger innovation performance (Cloodt, Hagedoorn & Kranenburg, 2006; Cassiman et al., 2005; Hagedoorn & Duysters, 2002). The positive effect on innovation is, in part, based on absorptive capacity; the more similar the two firms technological knowledge, the more quickly the acquired firm s knowledge can be assimilated and commercially exploited (Cohen & Levinthal, 1990; Lane & Lubatkin, 1998). However, too much similarity reduces the acquirer s opportunities for learning (Ghoshal, 1987; Hitt et al., 1996). Cloodt et al. (2006), for example, found an inverted U-shaped relationship between technological relatedness and post-merger innovation performance in a study of several high technology industries. Innovation performance (innovation quantity) was lowest for M&As where the firms were in highly similar or largely unrelated technology areas and highest when there was a moderate degree of overlap. Not only do M&As that integrate 1

4 highly similar technologies narrow the range of potential learning, they also reduce the incentives to explore divergent research opportunities available due to the M&A. Cassiman et al. s (2005) in-depth study of 31 high technology M&As found that firms are much more likely to reduce R&D effort, shorten the time horizon of projects and emphasize development over research when they acquire targets in similar technology areas than when they acquire targets in complementary areas. In short, the M&As that most improve innovation performance are those where the technological knowledge is similar enough to facilitate learning but different enough to create both new opportunities and the incentives to explore them. This paper extends these important findings by developing and testing a more fully developed model of relatedness and innovation performance of high technology M&As. Prior research suggests that the independent effects of similarity and complementarity of the acquirer s and the target s technological knowledge are important predictors of post-merger innovation performance (innovation quantity) (Ahuja & Katila, 2001; Cloodt, Hagedoorn & Kranenburg, 2006; Cassiman et al., 2005; Hagedoorn & Duysters, 2002). Further, research suggests that science and technology have independent as well as interactive effects on firm performance (Henard & McFayden, 2005). Science influences not only technology development but also the techniques available for that exploration (Nightingale, 1998). Thus, drawing on recent findings regarding science, technology and innovation (Makri, Lane & Gomez-Mejia, 2006; Henard & McFadyen, 2005) we propose that the addition of complementary science from an acquisition influences post-merger innovation performance directly and indirectly through its interaction with technology. More specifically, we suggest that complementary scientific knowledge and complementary technological knowledge both contribute to post-merger innovation performance. These two types of complementarities and their interactions fuel innovation by enlarging 2

5 the ideas and tools considered. Yet, too much breadth can lead to divergent R&D initiatives with little or no relationship to the firm s strategic objectives. Building on prior research, we propose that two types of product/market relatedness can help to focus the breadth provided by complementary science and technology. If the target and the acquirer are in the general industries, it is more likely that the ideas developed from the complementarities will be relevant to the acquirer s strategic objectives. Conversely, when the acquirer operates in multiple productmarkets, it can more easily find ways to utilize acquired complementarities than when it is in a single line of business. Simply put, the more diversified the acquirer, the greater the influence of complementary science and technology on post-merger innovation. [Insert Figure 1 Here] In summary, prior research suggests that combining technological complementarities through M&As may improve the innovation outcomes of high technology firms. We believe that the relationship is more complex, that technological complementarity must be combined with other forms of relatedness to enhance innovation. Building on absorptive capacity research, we propose that high technology acquisitions will lead to higher innovation performance when the acquirer and target have related areas of science as well as complementary technologies because this facilitates communication and the recognition of promising ideas. Also, we suggest that post M&A innovative performance is enhanced when the acquirer and target operate in related products/markets as well as having complementary technologies. This creates a shared context and similar dominant logics that in turn facilitates faster knowledge assimilation and innovation processes. Finally, we argue that high technology acquisitions will lead to higher innovation performance when the acquirer is diversified as well as having technologies complementary to those of the target. This provides a wide range of opportunities in which to apply the most valuable assimilated knowledge. 3

6 Our study makes several contributions to the M&A literature. First, the study shows that acquisitions can lead to greater internal innovation if the appropriate knowledge is acquired. Second, the research emphasizes the importance of knowledge complementarity to the success of high-technology acquisitions. Third, it introduces an objective measure of knowledge relatedness that distinguishes between science and technology relatedness. This measure makes possible a fine-grained assessment of firms knowledge relatedness and enables a more comprehensive assessment of the value creation potential between two merging firms. Fourth, this research emphasizes the value of the target-picking phase of acquisitions, because the type and degree of knowledge relatedness are significant determinants of post M&A success. We first address the two types of knowledge considered, science and technology, and discuss their role in innovation to provide the foundation for the definition of knowledge relatedness. SIMILARITY, COMPLEMENTARITY AND THE ROLE OF SCIENCE AND TECHNOLOGY IN INNOVATION Innovation has been defined in different ways in engineering, marketing, economics and management. The 1991 OECD provides one of the most comprehensive definitions of innovation as: an iterative process initiated by the perception of a new market and/or service opportunity for a technology-based invention which leads to development, production and marketing tasks striving for the commercial success of the invention (Garcia & Calantone, 2002; p.112). This definition emphasizes two important issues for understanding innovation. First, it suggests that the process of innovation includes the technological development of an invention along with the commercialization of that invention to end-users. Stated differently, at the origin of the innovation process are inventions and while innovations and inventions are related, they are not identical. The distinction is that innovation is a process that begins with an invention, proceeds with the development of the invention and results in the introduction of a new product, process, 4

7 or service to the marketplace (Edwards & Gordon, 1984; p. 1) (Acs & Audretsch, 2005). While innovation refers to the commercialization of an invention, invention refers to the act of creating something new (Ahuja & Lampert, 2001). Here we do not use the term innovation in the Schumpeterian sense as the recognition of economically viable inventions and taking actions to bring them to market (Baumol, 2002) but rather, we adopt an innovation production function framework and model patenting quantity and quality as the output of this production function. Second, the definition suggests that the innovation process is iterative in nature as products develop over time with an initial emphasis on product performance, followed by an emphasis on product variety and later emphasis on product standardization and cost reduction (Utterback & Abernathy, 1975). The iterative nature of innovation implies that it involves a process of recombination (Fleming & Sorenson, 2004), in which existing technological components are reconfigured, or new ones are discovered. The recombination process is facilitated by two types of knowledge. The first is knowledge about the core design concepts and the way in which they are implemented in a particular component (scientific knowledge). The second is knowledge about the ways in which the components are integrated and linked together into a coherent whole (technological knowledge). From these two types of knowledge different innovation types can emerge, ranging from incremental innovations for products at the advanced stages of the product life cycle to radical innovations for products at the early stages of diffusion and adoption. Incremental innovations are not based on dramatically new science (Henderson & Clark, 1990). They are improvements, refinements, modifications and extensions of core or complementary products that enhance the utility or reduce the cost of an existing product (Abernathy & Clark, 1985). Radical innovations, on the other hand, are based on a different set of engineering and scientific principles, often open wholly new markets and provide new potential applications (Dess & 5

8 Beard, 1984). Therefore, incremental innovations reinforce existing capabilities while radical innovations require asking a new set of questions, as well as drawing from new technical and scientific knowledge (Tushman & Anderson, 1986). Understanding and differentiating scientific knowledge and technological knowledge are important because science and technology behave uniquely and affect the process of discovery in different ways. For instance, technology facilitates an incremental recombination of components (Fleming & Sorenson, 2004) and it is more competence-enhancing than competence-destroying (Gatignon, Tushman, Smith &Anderson, 2002). Thus, technology-based innovations could take the form of incremental or architectural 1 innovations (Henderson & Clark, 1990), both of which work within the existing core design concepts for a product. Alternatively, science-based innovations could make obsolete the core design concepts that underlie existing technology and thereby lead to modular 2 or radical innovations (Henderson & Clark, 1990). The differences between science and technology are discussed next. The Role of Science and Technology in Innovation Much definitional diversity exists with the concept of technology. Traditional definitions focused on the physical characteristics of technology, as a system consisting of components and linkages among the components (e.g. Hughes 1983). More recently, this definition has been expanded from a physical artifact to include the knowledge embodied in such artifact (e.g. Thirtle & Ruttan 1987). For instance, Das and Van de Ven (2000) conceptualized technology as both an artifact and knowledge. Burgelman, Madique, and Wheelwright, (1996) provide a more comprehensive definition of technology as the theoretical and practical knowledge, skills and 1 Architectural innovations involve changes in a component to develop new interactions and new linkages with other components. 2 Modular innovations involve changes in the core design concepts but not the linkages between the components. 6

9 artifacts that can be used to develop products and services as well as their production and delivery systems (Burgelman et al, 1996: 2). The Burgelman et al (1996) conceptualization of technology maps onto the National Science Foundation s (NSF) definition of applied research. According to the NSF, applied research involves research projects with specific commercial objectives regarding either products or processes. Alternatively, basic research is the set of research projects which represent original investigation for the advancement of scientific knowledge and which do not have specific commercial objectives, although they may be in fields of present or potential interest to the reporting company (Link, 1996, p. 397). We integrate the NSF conceptualization of applied and basic research with Burgelman et al s (1996) work to define a firm s technology domain as the applied research and development techniques the firm employs to develop its product portfolio. A firm s science domain is defined as the set of scientific knowledge bases from which a firm draws to conduct research projects which do not have specific commercial objectives. Despite the tendency of strategic management researchers to treat technology and science interchangeably, technology is not science engineers are not scientists (Allen, 1984: p.307). Technology is about exploitation, adapting and combining what is known to achieve what is desired (Rip, 1992). It is driven by pressures from markets for products and services (Clark, 1987; Balmer & Sharp, 1993) and begins with an idea of what is needed to respond to those pressures. When a solution to a technological problem is not obvious, the firm works backwards from its preconceived ends and evaluates potential starting points (solutions) until an optimal one is found (Nightingale, 1998). Science, on the other hand, is exploratory in nature and tends to value unexpected findings even those that contradict previous experience or results (Barnes & Edge, 1982; Mulkay, 1977). While it has a known starting point, it searches for unknown outcomes (Nightingale, 1998). It is usually developed by scientific communities; 7

10 groups of scientists in different organizations that are bound by a common interest in a broadly defined problem (Kuhn, 1970; Crane, 1972; Latour, 1987; Tushman & Rosenkopf, 1992). A key distinction between science and technology is that while technological knowledge can be used to create products and services, scientific knowledge cannot do so directly. Rather, scientific knowledge influences products and services indirectly through its role in technology development (Nightingale, 1998). More specifically, science plays three roles in the innovation process. First, it produces an understanding of how technologies work. Second, it can help screen alternatives before technologists test them, thereby placing limits on the number and types of tests needed. Third, it can help extrapolate patterns or theories applicable to similar problems (Nightingale, 1998). Thus, science provides an understanding of how technological components interact and how technologies can be recombined (Fleming & Sorenson, 2004; Evenson & Kislev, 1976). The highlighted differences between science and technology suggest that science can play an enabling role in innovation by stimulating a better understanding of the problem at hand. Further, it could change or invalidate core design concepts underlying existing technology and lead to modular or radical innovations (Henderson & Clark, 1990). Because technology involves exploiting, adapting and combining what is known, it more likely leads to incremental or architectural innovations (Henderson & Clark, 1990). Thus, while technology facilitates an exploitative recombination process most likely leading to incremental innovations, science enables an exploratory process more likely leading to radical innovations. Knowledge Relatedness Redefined Defining and measuring relatedness are central issues in strategic management research; prior research has shown an equivocal effect of relatedness on firm performance, partly due to definitional concerns. While some diversification studies suggest that greater relatedness among 8

11 business units leads to higher profitability (Mahajan & Wind, 1988) due to greater synergies, others suggest that it can result in inflexibilities and poor responsiveness to market changes that adversely affect performance (Hill & Hoskisson, 1987). Similarly, some M&A studies show that related acquisitions outperform unrelated ones (Singh & Montgomery, 1987), while others find the exact opposite (Elgers & Clarke, 1980). Yet, a recent meta-analysis of M&A research suggests that relatedness has no observable effect on the performance of an acquisition (King et al, 2004). Still many scholars in the field believe that relatedness can produce value in acquisitions (Hitt, Harrison & Ireland, 2001). The conflicting results regarding the outcomes of relatedness arise from two interconnected problems. First, relatedness along product and market domains does not imply relatedness in knowledge domains and vice versa. Because knowledge is the primary resource on which competitive advantage is created, especially for high technology firms, it is important to examine relatedness in terms of firms knowledge domains. Secondly, relatedness has been defined rather broadly often using similarity and complementarity interchangeably (i.e., Davis, et al, 1992; Farjoun, 1998). For example, Davis, Robinson, Pearce and Park (1992) defined relatedness as: the degree to which business units support or complement each other's business activities. Even when complementarity is not used as a synonym for similarity, it has been conceptualized as the lack of similarity or overlap between core businesses or capabilities (e.g. Harrison et al, 1991). Only a few studies have considered knowledge as a basis for relatedness. For example, Singh and Montgomery (1987) classified acquisitions as related if the acquirer and target had at least one of the following: similar production technologies, similar science-based research, and similar products or markets. Defining relatedness more narrowly, Mowery, Oxley and Silverman (1998) used patents to measure technological similarity defining it as the extent to which two 9

12 firms technological resources draw from the same pools of technological knowledge (1998: p. 512). Following Mowery et al s (1998) lead, Ahuja and Katila (2001) used patent data to capture the acquiring and acquired firms knowledge bases. As they indicate, patent citation data corresponds closely to the conceptual abstraction of a firm s knowledge base as a set of knowledge elements. Since each patent number uniquely identifies a distinct component of knowledge, the higher the number of patents that are common across two knowledge bases, the higher the relatedness between those knowledge bases (Ahuja & Katila, 2001; 202). Also, Lane and Lubatkin (1998) considered the knowledge relatedness of alliance partners by examining the similarity of their basic and applied scientific knowledge. While these studies have been influential in providing a better understanding of the concept of knowledge relatedness, they often considered only one type of knowledge (Lane & Lubatkin, 1998; Ahuja & Katila, 2001) or treated science and technology as interchangeable (Singh & Montgomery, 1987). Such actions lack precision and are problematic because as discussed earlier, science and technology represent two distinct types of knowledge that affect innovation in different ways. Also, the prior studies do not adequately address an important type of relatedness: complementarity. They either have incomplete or tautological definitions of complementarity (Mowery, et al., 1998), or they ignore complementarity (Lane & Lubatkin, 1998; Ahuja & Katila, 2001). A handful of studies contributed to the development of a more comprehensive view of relatedness by differentiating complementarity from similarity. For example, Hitt, Harrison, Ireland and Best distinguished between relatedness and complementarity by indicating that relatedness of products/markets does not necessarily imply that the two firms have complementary resources resources do not have to be similar to be complementary (1998: p. 107). In the same vein, Kale, et al (2000), conceptualized complementarity in an alliance as the 10

13 valuable but different capabilities each partner brings to the relationship. Larsson and Finkelstein (1999) offered the most detailed study of complementarity in the strategy literature to date addressing many of the shortcomings of earlier studies. Considering marketing and production relatedness in M&As, they defined similarity as economies of sameness and complementarity as the combination potential between two firms (Larsson & Finkelstein, 1999: p. 6). Thus, they captured the super-additive value of complementarities (Tanriverdi & Venkatraman, 2005), albeit in a marketing and production context. Larsson and Finkelstein s (1999) work provides the basis for a more fully specified definition of knowledge relatedness. Integrating their concepts of economies of sameness and combination potential with the definitions of science and technology domain presented earlier suggests the following: Science similarity between firms is the degree to which their scientific research focuses on the same narrowly-defined areas of knowledge. Science complementarity between firms is the degree to which their scientific research focuses on different narrowly-defined areas of knowledge within a broadly defined area of knowledge that they share. Technology similarity between firms is the degree to which their technological problem-solving focuses on the same narrowly-defined areas of knowledge. Technology complementarity between firms is the degree to which their technological problem-solving focuses on different narrowly-defined areas of knowledge within a broadly defined area of knowledge that they share. [Insert Figures 2 & 3 here] The definitions of science and technology complementarity refer to knowledge complementarities within a value chain activity (R&D) as opposed to asset complementarities across different value chain activities. While some prior research has examined 11

14 complementarities in assets along the value chain (e.g. specialized manufacturing capability, access to distribution channels, service networks, and exclusive supplier relationships) (e.g., Teece, 1986; Tripsas, 1997; Rothaermel & Hill, 2005; Helfat, 1997), this research is focused on the source of complementarity grounded in firms knowledge domains. While synergies can arise from knowledge relatedness at multiple points of the value chain (Tanriverdi & Venkatraman, 2005), because innovation is critical for the success of high technology firms, it is important to assess knowledge relatedness in R&D. Therefore, focusing on knowledge relatedness in one aspect of the value chain, R&D, allows more complete specification of this construct and allows us to directly measure knowledge relatedness at the firm level. KNOWLEDGE RELATEDNESS IN HIGH TECHNOLOGY M&A For firms that rely on continuous innovations as a source of competitive advantage, knowledge synergies have become increasingly critical. Ahuja and Katila (2001) and Cloodt et al (2006) found that relatedness of the acquiring and target firms knowledge bases has a curvilinear effect on innovation output, while Cassiman et al (2005) found that when the merged entities are technologically complementary, their R&D productivity increases. Yet, more research is needed to address the differential effects of science and technology and the effects of knowledge complementarities on innovation outcomes. Knowledge Relatedness and Innovation in High Technology M&As Exploration facilitates the discovery of new opportunities and, simultaneously provides the potential for exploitation. Exploration is 'the pursuit of knowledge, of things that might come to be known,' and exploitation is 'the use and development of things already known' (Levinthal & March, 1993: 105). Science and technology map onto the exploration/exploitation framework. More specifically, technology similarities lead to local searches with firms using knowledge that is closely related to their preexisting knowledge bases (e.g. Martin & Mitchell, 1998; Stuart & 12

15 Podolny, 1996). Alternatively, science complementarities lead to distant search behaviors using knowledge that lies outside of existing knowledge bases (Miner, Bassoff, & Moorman, 2001). When two firms similar in both science and technology domains merge and combine their knowledge bases, they are expected to be dancing to the same music, using similar steps because they are familiar with the types of technological problems likely to surface and are relying on similar sets of scientific theories to understand and resolve them. They experience similar know-whats (the semantics and challenges of a particular information domain) and similar know-hows (an understanding of how semantics and challenges are causally linked) (Lubatkin et al, 2001). Further, using our understanding of absorptive capacity (Cohen & Levinthal, 1990; Lane & Lubatkin, 1998), the more similar the new knowledge is to the firm s existing knowledge, the more easily the new knowledge can be understood, assimilated and applied operationally. Also, experience with similar knowledge domains is likely to make the search process more predictable and more efficient. These arguments suggest that knowledge similarities (technology similarity in particular) facilitate the exchange and combination of existing knowledge (Nonaka, Takeuchi, & Umemoto, 1996) and encourage exploitation of what is already known. Cloodt et al. (2006), for example, found an inverted U-shaped relationship between technological relatedness and post-merger innovation in a study of several high technology industries. Innovation was lowest for M&As where the firms were in highly similar or largely unrelated technology areas and highest when there was a moderate degree of overlap. While knowledge similarity between the acquiring and target firm enhances exploitation and thereby innovation productivity, knowledge complementarities (science complementarities in particular), facilitate a process of exploration by experimenting with new competencies and technologies (March, 1991). Thus, acquiring complementary knowledge may help extend the 13

16 scope of innovation search and thereby contribute to richer innovations; however, it may also increase knowledge integration costs (Katila & Ahuja, 2002). Integrating complementary technology or complementary science can require effort and is more complex than integrating similar knowledge domains (Grant, 1996: 377). Yet, when the acquiring and acquired firms have knowledge complementarities, they have common knowledge stocks (in broadly defined areas), thus facilitating communication and coordination. Also, the common knowledge in broad areas helps each party understand the value of the unique but complementary sets of knowledge. These conditions help the two firms to achieve integration of their two complementary knowledge sets thereby increasing innovation productivity. Rothaermel, Hitt and Jobe (2006) found that firms able to integrate complementary knowledge from internal and external sources (through strategic alliances) increased the number of related new products introduced to the market. When knowledge complementarities between the acquiring and acquired firms are high, the merged firm s ability to respond correctly to new information increases. In this way, the common general knowledge stocks enhance the probability of success in innovation search processes (Cyert & March, 1963). Thus, knowledge complementarities contribute positively to more and richer innovations. Stated differently, attempts to integrate knowledge complementarity in both technology and science simultaneously post acquisition, escalate the potential benefits nonlinearly. If the acquirer has to assimilate and integrate a complementary science knowledge base in addition to a complementary technology knowledge base, such a combination is likely to have an increasingly positive effect on innovation productivity. Knowledge complementarities not only affect innovation productivity, but also affect the novelty of a firm s innovations (Hall, Jaffe, & Trajtenberg, 2001, DeCarolis & Deeds, 1999; Deng, Lev & Narin, 2001; Rosenkopf & Nerkar, 2001). The theory of recombinant innovation 14

17 suggests that innovations are the result of integrating existing or new knowledge elements to create new combinations (Fleming, 2001; Tushman & Rosenkopf, 1992). The merger of two firms can potentially lead to the creation of high quality innovations when they have similarities in their knowledge bases but also when some fraction of their knowledge is fairly diverse to permit effective creative utilization of the new knowledge (Cohen & Levinthal, 1990; p. 136). The similar activities in scientific research conducted by the acquiring and acquired firms and their work on technologies that share broadly defined areas of knowledge, allow them to communicate, coordinate and cooperate in effective ways. Yet, their foci of research in different specific knowledge areas of science and technology allow merging firms to complement the knowledge and research of each other and increase their exploration search processes. Exploring diverse realms of knowledge introduces a firm to more heterogeneous problem sets and can increase the probability that a radically different approach to solving a problem will emerge (Fiol & Lyles, 1985; Levinthal & March, 1993). Exposure to new sets of routines, new modes of reasoning, and challenges to existing understandings of cause-effect relationships helps a firm discover novel solutions to its problems. In Ahuja and Lampert s words the irritant of new, imperfectly understood streams of knowledge can foster the pearls of insight (2001, p. 527). The integration of these complementary knowledge stocks can lead to unique combinations and thus, novel innovations. Moreover, integrating both complementary technologies and complementary science provides the potential for a much greater portfolio of new and unique knowledge combinations. Science knowledge provides the base for advances in technological knowledge but technological knowledge advances also facilitate the further development and application of science knowledge. The great strides in mapping the human genome, for example, have been made possible by advances in instrumentation and computerized analytical tools technological knowledge embodied in products and processes. The result is an on-going series of 15

18 interactions in which science and technology are dancing partners that evolve around each other yet follow different steps to do so (Rip, 1992). As such, the interaction between technological and science complementarities of acquiring and acquired firms has a positive effect on the quality of innovations produced by the merged firm 3. These arguments lead to the following hypothesis: Hypothesis 1: The interaction of complementarities in the acquiring and acquired firms science and technology domains is positively related to innovation outcomes post M&A. Earlier it was suggested that acquiring and target firms with high similarity in their knowledge domains are not likely to develop radical innovations (all else being equal) after merging. While similarity in technology and science provides substantial absorptive capacity for the acquiring firm to learn from an acquired firm, it also creates significant path dependence. As such, there is a limit to the number of new recombinations that can be created using the same set of knowledge elements. On the other hand, firms with similar knowledge bases need less time and effort to integrate their R&D activities thereby facilitating innovation productivity. Science similarity suggests that the two firms scientists have similar understandings of how technologies work and thereby search for new solutions in the neighborhood of old or existing solutions. Such knowledge redundancy diminishes the opportunities for creating radically new knowledge. but when integrated with complementary technology it can increase potential recombinations leading to a greater number of innovations. Further, the interaction of technology complementarities with science similarities could lead to influential innovations since technology 3 The learning distance (Hoskisson & Busenitz, 2002) that exists with complementarity can increase the risk of moral hazard. This risk exists because an acquiring firm may not be able to adequately evaluate the target when the target s knowledge is reasonably distant from the knowledge held by the acquirer. Thus, knowledge-based information asymmetries are likely greater when two firms knowledge stocks are complementary rather than when they are similar. Due to the increased information asymmetries the premiums paid for complementary acquisitions are likely to be higher than those for more similar targets. 16

19 complementarities increase the scope of innovation search and thereby contribute to richer innovations. Formally stated: Hypothesis 2: The interaction of technology complementarities and science similarities in the acquiring and acquired firms knowledge domains is positively related to innovation outcomes M&A. As Casssiman et al (2005) noted, technological relatedness and product/market relatedness distinctly affect the relationship between M&A and innovation and should both be considered. Product and market relatedness can help to focus the breadth provided by complementary science and technology. If the target and the acquirer operate in the same general industries, (overlap in products/markets) it is more likely that the ideas developed from the complementarities will be relevant to the acquirer s strategic objectives. Conversely, when the acquirer is in multiple product-markets i.e. highly diversified, it is more likely to find ways to utilize acquired complementarities than if it was operating in a single line of business. Simply put, the more diversified the acquirer, the greater the influence of complementary technology on post-merger innovation. Formally stated: Hypothesis 3: The interaction of technology complementarities and product/market similarities between acquiring and acquired firms is positively related to innovation outcomes post M&A. Hypothesis 4: The more diversified the acquirer is, the greater the influence of technology complementarities on innovation outcomes post M&A. METHODS We tested these hypotheses using a 1996 sample of 95 high technology M&As (the science domain data were only available for 1996). Our sample was constructed in five steps. First, we used the Securities Data Corporation database (SDC) to select all M&As completed in 1996 in the drugs, chemicals and electronics industries because they are knowledge-intensive 17

20 industries that have been identified to use science in the development of innovation (Deng et al, 2001; Von Hippel, 1986). SDC is a financial transactions database which covers capital market, M&A, and private equity activities in the U.S. and internationally. We excluded M&As motivated to obtain access to distribution, gain entry into new markets, or obtain financial synergies or market power because such acquisitions do not generate science and technology inputs, and thus are unlikely to affect innovative output. In Ahuja and Katila s words, nontechnological acquisitions add less to the knowledge base of the acquirer and are less likely to lead to innovation output enhancing effects such as inventive recombinations (2001: 199). Second, following Finkelstein and Haleblian (2002) Harrison et al. (1991), and Chatterjee and Lubatkin (1990), we excluded deals smaller than $10 million and greater than $500 million because firms tend to adopt a hands-off approach with small acquisitions as their effects are likely to be negligible and acquisitions greater than $500 million tend to be driven by motivations other than knowledge exchange (market power, economies of scale, etc ). Third, data restrictions necessitated excluding all acquirers that were not publicly traded in the U.S. Applications of these criteria for selecting the sample resulted in 278 deals, 31 in drugs, 37 in chemicals and 210 in electronics industries. Fourth, we identified all name variants (i.e., names of divisions, subsidiaries, etc.) for acquirers and targets using the Directory of Corporate Affiliations, which contains business profiles and corporate linkages for 114,000 companies worldwide. We matched these to names in the National Bureau of Economic Research (NBER) and Center for Research Planning (CRP) databases because it was essential to have data on both acquirers and targets. These necessary reduced the sample to 100 deals for which complete data for acquirers and targets were available from the NBER and CRP database. The acquirer or the target in each of the remaining 178 deals did not patent or did not publish in scientific journals; in a majority of those deals (150 out of 178), either the acquirer or the target 18

21 did not publish in scientific journals thus the reduction in size was driven by lack of science domain data. Further, out of those 150 deals, 133 of them were in electronics while the other 17 were in drugs or chemicals. These differences in numbers reflect the common wisdom that drugs and chemicals are traditionally science-based industries while electronics is an emerging user of science (Lane & Makri, 2006). Comparative t-tests revealed that the 178 deals eliminated from the study in the first four steps were not statistically different on sales, R&D intensity and ROA from the final sample. Lastly, to ensure that the acquired firm had the potential to have a significant effect on the acquiring firm, all deals were eliminated from the sample where the target firm size was less than 5 percent of the acquiring firm s total annual sales. The final sample of 95 deals consisted of 24 deals in drugs, 27 deals in chemicals and 44 in electronics. Dependent Variables Innovation Quantity and Quality. We use two measures to capture innovation activity. The first, patent counts, is a direct measure of innovation quantity (e.g. Griliches, 1990, 1995, 1998) used extensively to indicate innovation productivity (Hitt, Hoskisson, Ireland & Harrison, 1991; Mowery et al, 1998; Malerba & Orsenigo, 1999; Sorensen & Stuart, 2000; DeCarolis & Deeds, 1999; Hall et al, 2001). The second, patent citations, is the extent to which a firm s patents are cited in subsequent patents. It measures innovation quality because it reflects the ability of a set of patents to support future innovation by creating a ripple effect to stimulate subsequent patents. We obtained annual data on the number of patents granted (quantity) and patents citations (quality) for each firm from from CHI research, which provides patent analyses for the National Science Foundation s bi-annual science and engineering 19

22 reports 4. Patent citations reflect the total number of forward citations in a given year to the company s patents issued in the most recent five years, divided by the average number of citations to all U.S. patents in the CHI database in the corresponding years. Independent Variables The technology similarity and complementarity measures were created using patent data from the NBER Patent Citations Data File constructed by Hall, Jaffe and Trajtenberg (2001). The U.S. Patent and Trademark Office (USPTO) classifies technologies into 417 main (3-digit) patent classes (Hall et al, 2001). Hall et al., (2001) aggregated these 417 classes into 36 two-digit technological subcategories, which in turn are further aggregated into 6 main categories. We constructed the similarity and complementarity measures for firms science and technology domains using data from CRP s Science Model TM (CRP). Each year CRP uses the raw data from ISI s Science and Social Science Citation Indexes (approximately 500,000 papers annually) and analyzes the patterns of co-citations for each of them (approximately 8 million citations annually). Co-citation analysis uses frequently occurring pairs of citations in peerreviewed articles to identify the social and intellectual structure of science and provides insights into how this structure changes over time (Garfield, Sherer & Torpie, 1964; Garfield, 1979). This technique is widely used by science historians (for a review see Kuhn, 1970, p ), and has recently been applied to analyses of strategic issues (Lane & Lubatkin, 1998; Deeds, 2001). 5 CRP s unit of analysis is the research community (RC) which is composed of two sets of papers: current papers (published in the year being examined) and base papers (earlier papers highly co-cited in that year). A firm s participation in research communities reflects the research 4 After our study was completed, CHI Research was acquired by Thomson Scientific s Delphion patent service. 5 The validity of CRP s method of modeling the structure of science was tested in depth by the United Kingdom Advisory Board for the Research Councils (ABRC) in 1983 and 1984 (Healy, Rothman & Hock, 1986), and more recently by Klavans and Boyack (2006). 20

23 streams in which the firm s researchers are involved. CRP also considers a firm s participation in science disciplines where disciplines are defined as fundamental fields of academic instruction. Research communities and science disciplines represent a classification of firms scientific knowledge that is analogous to the classification system for technologies (patent classes and subcategories respectively). By examining the patterns of co-citation, CRP develops an annual model of the social structure of science and can assess the performance and trends of each area of research. A visual representation of the structure of the CRP and NBER data is presented in Figure 4. [Insert Figure 4 about here] Technology Relatedness Technology Similarity Technology similarity between two firms was operationalized with the degree of overlap in patents in the same 3-digit patent class. This measure captures the extent to which two firms develop technology in the same patent classes, hence using similar technological knowledge. It is calculated as degree of patent overlap between the target and acquirer, weighted by the importance of each patent class for the acquirer. Technology Complementarity While the measures of technological overlap are effective proxies for the similarity of technological assets, they do not capture possible technological complementarities. Technological complementarity is operationalized here as the overlap in patents in the same subcategory but in a different patent class. This measure mirrors the theoretical conceptualization of complementarity because it captures the integrative potential between two firms. For example, if two firms patents build on subcategory Biotechnology in the category Drugs & Chemical, the extent to which they build on different technological classes (e.g. Patent Class 435 Vs Patent Class 800) captures their integrative potential (Figure 1). 21

24 Science Relatedness Science relatedness was measured using the two types of data from CRP described earlier: a) indicator of a firm s participation (membership) in research communities, and b) indicator of a firm s participation in science disciplines. The first CRP measure used indicates a firm s membership in a research community (RC), a self-organized (emergent) group of current researchers whose current papers address similar research topics and have a common base bibliography (i.e., similar theories and methods). The second measure indicates a firm s membership in a science discipline (Disc). If two firms scientists publish in the same research communities, the firms are familiar with the same methods and theories, and are examining similar research questions. If two firms have scientists who publish in the same science disciplines but in different research communities, they will have some shared knowledge from general research questions, theories and methods, but also have different areas of specialized expertise. These two types of data permit the calculation of a science similarity and a science complementarity measure. Science Similarity The logic underlying this construct is that the more similar the research publications of two firm s scientists, the more similar is the firms scientific knowledge. Thus, the science similarity variable was operationalized as the overlap in the research communities in which the acquirer and target publish. Science complementarity Science complementarity is defined as the extent to which the target and acquirer build on the same science disciplines but not on the same research communities. In other words, it captures the extent to which two firms draw from related but different areas of science and it is based on the number of unique (non-overlapping) research communities (RCs) in which each firm participates within the science disciplines that they share. All measures of science and technology similarity and complementarity are described and 22

25 summarized in Table 1. [Insert Table 1 here] To better illustrate our definitions of science and technology relatedness, consider the following hypothetical example of a merger between two firms (A and B) with an equal number of total patents (20) in a small range of patent categories and sub-categories. These relationships are illustrated in Figure 5. Firms A and B each have 12 patents in the sub-category Biotechnology, which is under the main category Drugs and Medical. The 60% overlap in this narrowly defined domain (Biotechnology) between these two firms indicates a high degree of technological similarity. Both firms also have 2 other patents under the same main category, but in different sub-categories: Molecular Biology (Firm A) and Microbiology (Firm B). The 10% overlap in areas of knowledge content focused on different narrowly defined topics (Molecular Biology, Microbiology) but similar broad areas of knowledge (Drugs and Medical) reflects a low degree of technological complementarity. The same type of analyses could be done for the science relatedness of firms A and B. [Insert Figure 5 here] Because the measures of complementarity are new and highly important to the efficacy of this study and in response to recent calls to demonstrate the construct validity of important measures in strategic management research (Boyd, Gove & Hitt, 2005), we took further steps to test the construct validity of our measures of complementarity. To do so, we collected data from trade magazines in the industries represented in our sample. The data provided information on the acquiring and acquired firms research content and foci for 24 deals in the electronics industry and for 24 deals in the drugs industry (n = 48). We asked two highly qualified professionals to read carefully through all of the information provided for each firm and to rate the knowledge complementarity present between the two firms in each deal. One of the raters has 23

26 a Ph.D. in computer science and the other rater has a Ph.D. in immunology. They rated the knowledge complementarity between the two firms on a five-point Likert type scale. We then calculated the correlations between their ratings on knowledge complementarity and our objective measures of technology complementarity and science complementarity. There was a positive and statistically significant correlation between our objective measure of technology complementarity and the raters evaluation of knowledge complementarity (r =.69, p <.01). There was also a positive and statistically significant correlation between our objective measure of science complementarity and the raters evaluation of knowledge complementarity (r =.41, p <.01). These results provide support for the construct validity of our measures of complementarity. Acquirer Product Diversification Acquirer product diversification pre M&A was captured using the entropy measure (Hitt, Hoskisson & Kim, 1997) for Product/Market Similarity Product/market similarity was operationalized by capturing whether the acquirer and target fall within the same 4 digit primary SIC. Control Variables Several variables are used to help control for alternative explanations of the findings. Acquirer characteristics used as controls include industry (e.g. Hoskisson & Hitt, 1990), R&D intensity (e.g. Morck, Shleifer & Vishny, 1988; Morck & Yeung, 1991), industry weighted average pre-merger ROA, and relative size of target/acquirer in terms of assets (e.g. Haleblian & Finkelstein, 1999). All of these variables have been suggested or shown to affect one or more of the outcomes examined in this study. Also, we accounted for cost reductions that might be associated with the acquisition measured as the change in Selling, General and Administrative 24

27 expenses one year after the deal was completed because of the potential effect on the short-term returns (e.g., ROA). Further, we controlled for prior acquisition experience (Simonin, 1997) as well as the number of post M&A acquisitions each firm completed following 1996 that could affect innovation outcomes. Statistical Analysis We tested our hypotheses using panel data econometric methodology, as we have innovation quantity and quality data for years Specifically, we use a random-effects model (GLS) because we have time-invariant independent variables and also because estimates computed using fixed-effects models can be biased for panels over short periods (Greene, 1993; Judge, Hill, Griffiths & Lee, 1985; Hsiao, 1986). This bias is not a problem with random-effects models. Further, even though a fixed-effects model has the advantage over a random-effects model of allowing the possibility that the omitted variables may be correlated with other independent variables in the model (Baltagi, 1995), the Hausman s chi square statistic indicated that a random-effects test was the most appropriate with our data. Because innovation outcomes are likely to lag M&A activity, we incorporated lagged measures. One year after the deal is completed i.e. in 1997, is the earliest the merged companies could file for patents resulting from joined R&D efforts. Because post M&A performance has been examined three-five years after the deal completion year (Bettis & Mahajan, 1985; Robins & Wiersema, 1995; Harrison et al, 1991), we consider innovation quantity, innovation quality, and financial performance, for three-five years after the completion of the acquisition ( ). The data on all independent variables (including control variables) were collected for Thus, there is a lag structure of three-five years. DATA ANALYSIS AND RESULTS Table 2 presents descriptive statistics for the variables included in the analyses in the 25

28 years 1996 and In 1996, acquiring firms spent on average $198.8 million on R&D, (11.3 percent of their sales) and were granted on average 59 patents. In 2000, these same firms spent 37 percent of their sales on R&D and were granted on average 76 patents. [Insert Table 2 here] Acquirers and targets technological knowledge similarity was 7 percent on average, (7 percent of the acquirers and targets patents in 1996 were classified by the U.S. Patent Office in the same patent classes), indicating that these firms were building on similar applied knowledge. The range of technological similarity was 0-90 percent, which is in line with Mowery et al. s (1998) research on technological overlap in alliances where technological similarities were 5-50 percent. Acquirers and targets technological knowledge complementarity was 17 percent suggesting that 17 percent of acquirers and targets patents in 1996 were classified by the U.S. Patent Office as being in the same subcategory but not in the same class, signifying that these firms were building on related technological knowledge. The range of value in technological complementarity was 0-80 percent. While there are no prior benchmarks for technological complementarity these levels appear to be reasonable. The levels of scientific similarity are lower than those of technological similarity. On average, about 1 percent of the acquirers and targets scientific knowledge was similar with the range 0-11 percent. This suggests that before the merger, few target and acquirer firms addressed similar sets of problems using similar theories and methods. Scientific complementarity is also lower than technological complementarity as on average, about 9 percent of the acquirers and targets scientific knowledge was complementary (similar sets of problems but addressed using different theories and methods) with a range of 0-78 percent. Science and Technology Relatedness and Innovation Outcomes The first hypothesis examined the effects of science and technology complementarity on 26

29 innovation quantity and quality. Table 3 summarizes the regression analysis for testing these effects. [Insert Table 3 about here] The results shown in Table 3 depict a positive interaction effect between science and technology complementarity on innovation quantity (p <0.05) and innovation quality (p<0.10). To examine these interaction effects further we plotted the results using the same method shown in Hitt, Bierman, Uhlenbruck and Shimizu (2006). In the graphs presented in Figure 6, we show the effects of technology complementarity on innovation quantity for two levels of science complementarity, low and high (minus one standard deviation from the mean and plus one standard deviation from the mean respectively). We plotted innovation quantity regressed on different levels of technology complementarity. As can be seen in Figure 6, the highest level of innovation quantity is achieved when both science and technology complementarity are high. [Insert Figure 6 here] In addition, in Figure 6 we show the effects of technology complementarity on innovation quality for two levels of science complementarity, low and high (minus one standard deviation from the mean and plus one standard deviation from the mean respectively). We plotted innovation quality regressed on different levels of technology complementarity. As can be seen in Figure 6, the highest level of innovation quality can be reached when both science and technology complementarity are high. However, it is also worthy of note that innovation quality is higher at all levels when science complementarity is high regardless of the level of technology complementarity, which is in line with the idea that science is the enabling force in the innovation process. [Insert Figure 6 here] Hypothesis 2 addresses the effects of science similarity and technology complementarity 27

30 on innovation quantity and quality and Table 4 summarizes the results of the analyses. The results show that science similarity and technology complementarity have a positive and moderately statistically significant interactive effect on innovation quantity (p<0.10) but none on innovation quality. Thus, Hypothesis 2 receives partial support. [Insert Table 4 about here] Hypothesis 3 addresses the effects of product/market similarity and technology complementarity on innovation quantity and quality and Table 5 summarizes the results of the analyses. The results show that product/market similarity and technology complementarity have a positive statistically significant interactive effect on innovation quantity (p<0.001) but a negative statistically significant interactive effect on innovation quality (p<0.05). Thus, Hypothesis 3 receives partial support. [Insert Table 5 about here] Hypothesis 4 addresses the effects of acquirer diversification pre-m&a and technology complementarity on innovation quantity. As suggested by the results in Table 6, the interaction of technology complementarity and acquirer diversification does not have a statistically significant effect on innovation quantity but it has a positive statistically significant interactive effect on innovation quality (p<0.05). Thus, Hypothesis 4 receives partial support. [Insert Table 6 about here] In the graphs presented in Figure 7, we show the effects of technology complementarity on innovation quantity and quality for two levels of product/market similarity, low and high. As can be seen in Figure 7 the highest level of innovation quantity is achieved when both the target and the acquirer are in the general industries (high product/market similarity) and technology complementarity is high. However, when product/market similarity and technology complementarity are high, innovation quality suffers. Interestingly, innovation quality is higher 28

31 when technology complementarity and the acquirer and target are in different industries because possibly that allows for more unique recombinations to emerge. In addition, in Figure 7 we show the effects of technology complementarity on innovation quality for two levels of acquirer product diversification. As can be seen, the highest level of innovation quality can be reached when both acquirer diversification and technology complementarity are high because when the acquirer operates in multiple product-markets, it can more easily find ways to utilize acquired complementarities than when it is in a single line of business. [Insert Figure 7 here] DISCUSSION While prior research has yielded equivocal results regarding the relationship between relatedness and performance in M&As, the research reported herein extends our understanding of the relationship (e.g., more depth and clarity), especially in high-technology industries. We find that knowledge relatedness is complex and that technological complementarity must be integrated with other forms of relatedness to enhance innovation. Prior strategic management research suggested that firms engaging in M&As may produce less innovation (Hitt, Hoskisson, Johnson & Moesel, 1996). Our study shows that firms acquiring others with complementary science and technology knowledge can produce a greater number of innovations and more novel innovations. The Hitt et al (1996) results are likely representative of acquisitions in general, especially for industrial firms operating in medium and low technology markets and firm in high technology industries that do not have the science and technology complementarities. Additionally, we find that two types of product market relatedness can help to focus the breadth provided by complementary science and technology. If the target and the acquirer are in the same general industries, it is more likely that the ideas developed from the complementarities will be relevant to the acquirer s strategic objectives. Conversely, when the acquirer itself is in 29

32 multiple product-markets it is more likely to find ways to utilize acquired complementarities than is it is in a single line of business. Simply put, the more diversified the acquirer, the greater the influence of complementary science and technology on post-merger innovation. Overall the results of this study support the important and pervasive influence of knowledge complementarities, and strongly suggest that firms in high-technology industries interested in making acquisitions should search for, identify and acquire businesses that have complementary scientific and technological knowledge. If they do so, they have a higher likelihood of experiencing positive outcomes from the acquisition. Specifically, our findings suggest that the integration of technology complementarities with science similarities can help to produce a greater number of innovations because technology complementarities increase the number of potential recombinations and thereby contribute to higher innovation productivity. In addition to creating fruitful avenues for future research on M&As, this study has implications for research in science and innovation, strategic alliances, interorganizational learning, and for management practice as well. Implications for Theory and Research M&A Research. Evaluating the performance of high technology acquisitions has received only limited attention (Ahuja & Katila, 2001; Grandstrand & Sjolander, 1990; Gerpott, 1995) and the existing research has mainly focused on the effect of acquisitions on the size of the acquirer s knowledge base. The framework introduced herein allows evaluation of the effects of acquisitions on the quality of the acquirer s knowledge base in addition to size, thereby integrating perspectives from the innovation literature and studies in the corporate control tradition. It allows future research to extend and enrich the prior research on the relationship between acquisitions and innovation (Hitt et al., 1991, 1996; Hoskisson et al., 1994). We can better understand the boundary conditions for the positive and negative relationships posed by 30

33 others previously. Recently, Harrison, Hitt, Hoskisson and Ireland (2001) and Hitt et al (2001) suggested that the two viewpoints are complementary. They indicate that while an active acquisition strategy does not necessarily reduce managerial commitment to innovation, firms should be careful and vigilant when trying to increase innovation by acquiring other companies. They argue that taking great care in selecting a complementary target followed by an emphasis on innovation after the acquisition can produce greater success. Assessments of technology and science knowledge relatedness among high-technology firms can help identify the best targets to acquire for effective integration and enhancements of innovation quantity and quality. The differential effects of knowledge relatedness on innovation emphasize the importance of target selection, the resource-picking mechanisms of rent creation (Barney, 1986; Makadok, 2001) and the effect os such selections on capability-building. Makadok (2001) provided an integrative framework that synthesized the resource-based view (e.g. Barney, 1986) and dynamic capabilities view (e.g., Teece, Pisano & Shuen, 1997). Makadok s (2001) synthesis suggests that firms generate rents via two distinct mechanisms, resource-picking and capability building. The former, which occurs before a resource is acquired, creates economic rents when the firm is able to obtain unique and valuable resources that can be used in combination with its existing stock of resources. The latter creates economic profit when a firm has the ability to leverage its existing resources and improve their productivity. Makadok (2001) suggested that the two mechanisms complement each other; however he did not explain how they do this. The theory developed herein sheds some light on this process by exploring how choosing targets with complementary knowledge (resource-picking) affects innovation outcomes (capability-building). Alliance Research. Although our study examined M&As, the findings provide 31

34 implications for understanding high technology alliances as well. Prior research has suggested when the anticipated interdependence between alliance partners is high, there is a greater the magnitude of expected coordination costs. As a result a more hierarchical governance structure is used to manage the alliance (Gulati & Singh, 1998). Specifically, Gulati and Singh (1998) argued that technology alliances are likely to be governed with a hierarchical structure due to the uncertainty arising from task coordination and the complexity of decomposing the division of tasks (coupled with managers bounded rationality). The theory introduced herein advances our understanding of the governance structure in alliances. For example, knowledge relatedness between the partner firms facilitates coordination. Additionally, a distinction between similar and complementary knowledge stocks among partners in high technology alliances contributes to understanding the potential value that can be gained from the alliance. Future research could examine whether alliances with complementary knowledge, due to higher interdependence, require a different governance structure than alliances based on more similar knowledge. Interorganizational Learning. This study extends our understanding of knowledge complementarity and interorganizational learning (knowledge transfer), specifically the notion of relative absorptive capacity. Lane and Lubatkin (1998) suggested that a firm s ability to learn from its partner in an alliance depends on the similarity of firms knowledge bases, organizational structures, compensation policies, and dominant logics. Our research suggests that the relative absorptive capacity construct should be expanded to include complementarity in firms knowledge bases. Future research could examine how relative complementarity affects knowledge transfer between alliance partners. For instance, the degree of relative complementarity can affect the level of private benefits accrued to each partner (Khanna, 1998) and subsequently the duration of the alliance. Therefore, the idea of knowledge complementarity can be employed to examine how relative complementarity affects the mix of private and 32

35 common benefits (Khanna, 1998), and explore how the combination of benefits, in turn, affects alliance partners' incentives to invest in learning. Implications for Management Practice This study suggests that managers and members of the board of directors can better evaluate potential acquisition targets and alliance partners using their understanding of the knowledge similarity and complementarity between the two firms to assess the probability of achieving unique and valuable synergy (e.g., innovation outcomes). Further, it suggests that evaluating a target based on technology complementarities alone is inadequate. Assessing knowledge complementarity in both technology and science is important because the acquirer has to integrate both complementary science knowledge bases and complementary technology knowledge bases. This is because their integration is likely to have an increasingly positive effect on innovation productivity and quality. Accordingly, managers should be encouraged via incentive programs to seek targets for acquisition that hold complementary knowledge. As such, using short-term objective financial criteria to reward those managers (i.e. fiscal performance results) may negate the potentially positive effects of acquiring complementary knowledge by encouraging them to stay within familiar areas of knowledge that provide more immediate results (e.g., gaining economies of scale). On the other hand, using long-term criteria to reward executives can provide incentives for them to search for targets with complementary knowledge and to explore new realms of knowledge. Study Limitations Several limitations should be noted. First, the relevance of these findings may be limited to industries where patents are meaningful indicators of innovation. In addition to the drugs, chemicals, and electronics industries, these findings are relevant to industries that are emerging 33

36 users of scientific research such as soaps, plastics, and food (Lane & Makri, 2006). Second, while patent classification schemas provide a clearer picture of a firm s knowledge domain than SIC codes, similar to SIC codes they were developed for another purpose. However, coding conventions facilitate the study of innovation-related topics using large samples. Third, while patents generally correlate well with new products (Comanor & Scherer, 1969), not all are commercialized. Thus, patent quality only partially captures the novelty of commercialized inventions. Future research could examine the novelty of the commercialized products to enable a more thorough evaluation of the effects of knowledge relatedness on innovation. Concluding Remarks Many of the theories and constructs used in prior research and practice to evaluate the effectiveness of M&As are entrenched in the old, manufacturing-dominated competitive environment. Yet, in high technology industries and an increasing number of other industries (e.g., professional services), knowledge is essential for success. Thus, new constructs and approaches are needed to understand what is required for success in the new competitive landscape (Bettis & Hitt, 1995). This study suggests that in high technology M&As, firms integration of science and technology knowledge may serve as a better indicator of private synergy than assessing relatedness in terms of firms market and product portfolios. As such, it provides a base for future research and changes in managerial practices regarding the M&A strategy. 34

37 REFERENCES Abernathy, W. J., & Clark, K Innovation: Mapping the winds of creative destruction. Research Policy, 14: Acs, Z. & Audretsch, D Innovation and Technological Change. In Handbook of entrepreneurship research, Springer, Acs and Audretch (Eds.). Ahuja, G., & Katila, R Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strategic Management Journal. 22: Ahuja, G., & Lampert, C. M Entrepreneurship in the large corporation: A longitudinal study of how established firms create breakthrough inventions. Strategic Management Journal, 22: Allen, T. J Managing the flow of technology. Cambridge: MIT Press. Balmer, B., & Sharp, M The battle for biotechnology: Scientific and technological paradigms and the management of biotechnology in Britain in the 1980s. Research Policy, 22: Baltagi, B.H., 1995, Econometric Analysis of panel data, Chichester: John Wiley & Sons Ltd. Barney, J. B Strategic factor markets: expectations, luck and business strategy. Management Science, 32: Barnes, B. & Edge, D., The interaction of science and technology, In B. Barnes & D. Edge (Eds.), Science in context: Readings in the sociology of science: Cambridge, MA: MIT Press. Barney, J. B Returns to bidding firms in mergers and acquisitions: Reconsidering the Relatedness Hypothesis. Strategic Management Journal, 9: Baumol, W The free-market innovation machine, Princeton University Press. Baysinger, B., & Hoskisson, R. E Diversification strategy and R&D intensity in multiproduct. Academy of Management Journal, 32: Bettis, R. A. & Hitt, M. A The new competitive landscape. Strategic Management Journal, 16(Special Issue): Bettis, R. A., & Mahajan, V Risk/return performance of diversified firms. Management Science, 31: Boyd, B. K., Gove, S. & Hitt, M. A Construct measurement in strategic management research: illusion or reality? Strategic Management Journal, 26: Burgelman, R., Madique, M. & Wheelwright, S Strategic management of technology and 35

38 innovation. Chicago: Irwin. Chatterjee, S., Lubatkin, M. H Corporate mergers, stockholder diversification, and changes in systematic risk. Strategic Management Journal, 11: Chatterjee, S., Lubatkin, M. H., Schweiger, D. M., & Weber, Y Cultural differences and shareholder value in related mergers: Linking equity and human capital. Strategic Management Journal, 13: Chung, K. H., & Pruitt, S. W A simple approximation of Tobin's q. Financial Management, 2: Cimilluca, D Buyout firms fuel a record; Pace of M&A pacts could start to slow amid tighter credit, Wall Street Journal, July 2: C8. Clark, N Similarities and differences between scientific and technological paradigms. Futures, (February): Cohen, W. M., & Levinthal, D. A Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35: Comanor, W. & Scherer, F Patent statistics as a measure of technical change. Journal of Political Economy, 77: Crane, D Invisible colleges: The diffusion of knowledge in scientific communities. Chicago: Chicago University Press. Cyert R., & March, J A behavioral theory of the firm. Prentice-Hall: Englewood Cliffs, NJ. Datta, D. K. Pinches, G. P., & Narayanan, V. K Factors influencing wealth creation from mergers and acquisitions: A meta-analysis. Strategic Management Journal, 13: Das, S., & Van de Ven, A Competing with new product technologies: A process model of strategy. Management Science. v 46: Davis, P. S., Robinson, R. B., Pearce, J. A., & Park, S. H Business unit relatedness and performance: A look at the pulp and paper industry. Strategic Management Journal, 13: DeCarolis, D. M., & Deeds, D The impact of stocks and flows of organizational knowledge on firm performance: An empirical investigation of the biotechnology industry. Strategic Management Journal, 20: Deeds, D The role of R&D intensity, technical development and absorptive capacity in creating entrepreneurial wealth in high technology start-ups. Journal of Engineering and Technology Management. 18: p Deng, Z., Lev, B., & Narin, F Science and technology as predictors of stock performance. Financial Analysts Journal, 20:

39 Dess, G. & Beard, D Dimensions of Organizational Task Environments. Administrative Science Quarterly, 29: Dosi, G Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the Determinants and Directions of Technical Change. Research Policy, 11: Edwards, K. & Gordon, T Characterization of innovations introduced on the US market in The Futures Group, prepared for the U.S. Small Business Administration under Contract No. SBA-6050-OA82. Elgers, P. T., & Clark, J. J Merger types and shareholder returns: Additional evidence. Financial Management, 9(2): Evenson, R. & Kislev, Y A Stochastic Model of Applied Research. Journal of Political Economy, 84: Farjoun, M The independent and joint effects of the skill and physical bases of relatedness in diversification. Strategic Management Journal, 19, Finkelstein, S. & Haleblian, J Understanding acquistion performance: The role of transfer effects. Organization Science. 13: Fiol, C. M., & Lyles, M Organizational learning. Academy of Management Review, 10: Fleming, L Recombinant uncertainty in technological search. Management Science, 47: Fleming, L. & Sorenson, O Science as map in technological search. Strategic Management Journal, 25: Garcia, R., & Calantone, R A critical look at technological innovation typology and innovativeness terminology: A literature review. Product Innovation Management, 19: Garfield, E Citation indexing, its theory and application in science, technology, and humanities. New York: Wiley. Garfield, E., Shere, I. H., & Torpie, R. J The use of citation data for writing the history of science. Philadelphia: Institute for Scientific Information. Gatignon, H., Tushman, L. M., Smith, W., & Anderson, P A structural approach to assessing innovation: Construct development of innovation locus, type, and characteristics. Management Science, 48: Gerpott, T Successful integration of R&D functions after acquisitions: An exploratory empirical study. R&D Management, 25: Granstrand, O., & Sjolander, S The acquisition of technology and small firms by large 37

40 firms. Journal of Economic Behavior and Organization, 13: Grant, M. R Toward a Knowledge-Based Theory of the Firm. Strategic Management Journal 17 (Winter Special Issue): Greene, W.H., Econometric Analysis, New York: MacMillan Publishing Company. Griliches, Z Patent statistics as economic indicators: A survey. Journal of Economic Literature, 28: Griliches, Z R&D and productivity: Econometric results and measurement issues. Oxford: Blackwell. Griliches, Z R&D and productivity: The econometric evidence. Chicago: University of Chicago Press. Gulati, R., & Singh, H The architecture of cooperation: Managing coordination costs and appropriation concerns in strategic alliances. Administrative Science Quarterly. 43: Hagedoorn, J Inter-firm R&D partnerships: an overview of major trends and patterns since Research Policy, 31: Haleblian, J., & Finkelstein, S The influence of organizational acquisition experience on acquisition performance: A behavioral learning perspective. Administrative Science Quarterly, 44: Hall, B., Jaffe A., & Trajtenberg, M Market value and patent citations. The Rand Journal of Economics, 36: Hall, B, Jaffe, A., & Trajtenberg, M NBER Patent citations data file: Lessons, insights and methodological tools (NBER working paper, No. 8498). National Bureau of Economic Research. Harhoff, D., Narin, F., Scherer, F. M., & Vopel, K Citation frequency and the value of patented inventions. Review of Economics and Statistics, 81: Harrison, J., Hitt, M. A., Hoskisson, R. E., & Ireland, R. D Synergies and postacquisition performance: Differences versus similarities in resource allocations. Journal of Management, 17: Harrison, J., Hitt, M., Hoskisson, R., & Ireland, D Resource complementarity in business combinations: Extending the logic to organizational alliances. Journal of Management, 27: Hayward, M. L When do firms learn from their acquisition experience? Evidence from Strategic Management Journal, 23, Healy, P., Rothman, H., & Hock, P.K An experiment in science mapping for research planning. Research Policy, 15:

41 Helfat, C Know-how and asset complementarity and dynamic capability accumulation: The case of R&D. Strategic Management Journal, 18, Henderson, R. M., & Clark, K. B Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms. Administrative Science Quarterly, 35: Hill, C.W.L., & Hoskisson, R. E Strategy and Structure in the Multiproduct Firm. Academy of Management Review, 12: Hitt, M. A., Bierman, L., Uhlenbruck, K., & Shimizu, K The importance of resources in the internationalization of professional service firms: The good, the bad, and the ugly. Academy of Management Journal, in press. Hitt, M. A., Harrison, J.S., & Ireland, R.D Mergers & acquisitions: A guide to creating value for stakeholders, New York: Oxford University Press. Hitt, M. A., Harrison, J.S., Ireland, R.D., & Best, A Attributes of successful and unsuccessful acquisitions of U.S. firms. British Journal of Management, 9: Hitt, M.A., Hoskisson, R.E., Kim, H International diversification: Effects on innovation and firm performance in product-diversified firms, Academy of Management Journal, 40: Hitt, M. A., Hoskisson, R. E., Ireland, D. R, Harrison, S.J Effects of Acquisitions on R&D Inputs and Outputs. Academy of Management Journal. 34: Hitt, M. A., Hoskisson, R. E., Johnson, R. A. & Moesel, D. D The market for corporate control and firm innovation. Academy of Management Journal, 39: Hoskisson & Busenitz, Market uncertainty and learning distance in corporate entrepreneurship entry mode choice, in Hitt, Ireland, Camp & Sexton (eds.). Strategic entrepreneurship: Creating a new mindset, Oxford, UK: Blackwell Publishers. Hoskisson, R. E., & Hitt, M. A Antecedents and performance outcomes of diversification: A review and critique of theoretical perspectives. Journal of Management, 16: Hoskisson, R., Johnson, R., & Moesel, D Corporate divestiture intensity in restructuring firms: Effects of governance, strategy, and performance. Academy of Management Journal. 37: Hsiao, C., Analysis of panel data, Cambridge: Press Syndicate of the University of Cambridge. Hughes, T Networks of Power: Electrification in Western Societies, Johns Hopkins University Press, Baltimore, MD. Judge, G., Hill C., Griffiths W., & Lee, T., The theory and practice of econometrics, New York: John Wiley & Sons. 39

42 Kale, P. Singh, H., & Perlmutter, H Learning and protection of proprietary assets in strategic alliances: Building relational capital. Strategic Management Journal, 21: Kannan, R The performance impact of strategic similarity in horizontal mergers: Evidence from the U.S. banking industry. Academy of Management Journal, 40: Katila, R., & Ahuja, G Something old, something new: A longitudinal study of search behavior and new product introduction. Academy of Management Journal, 45: Khanna, T The scope of alliances. Organization Science. 9: King, D.R., Dalton, D.R., Daily, C.M., & Covin, J Meta-analyses of post-acquisition performance: Indicators of unidentified moderators. Strategic Management Journal, 25: Klavans, R. & Boyack, K. W Identifying a better measure of relatedness for mapping science. Journal of the American Society for Information Science and Technology. 57(2): Kogut, B., & Zander, U Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science, 3: Kuhn, T. S The structure of scientific revolutions. Chicago: University of Chicago Press. Lane, P. J., & Lubatkin, M Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19: Lane, P. J., & Makri, M Responding to diminishing technological opportunities: A sociocognitive view of science and innovation. Manuscript submitted for publication. Larsson, R., & Finkelstein, S Integrating strategic, organizational and human resource perspectives on mergers and acquisitions: A case survey of synergy realization. Organization Science, 10: Latour, L., Science in action. Cambridge, MA: Harvard University Press. Levinthal, D., & March, J The myopia of learning. Strategic Management Journal, 14: Link, A On the classification of industrial R&D. Research Policy, 25(3), Lubatkin, M., Florin, J. & Lane, P Learning together and apart: A model of reciprocal interfirm learning. Human Relations, 54: Lubatkin, M Merger Strategies and Stockholder Value. Strategic Management Journal, 8: Mahajan, V., & Wind, Y Business Synergy Does Not Always Pay off. Long Range 40

43 Planning, 21(1): Makadok, R Toward a synthesis of the resource-based and dynamic capability views of rent creation. Strategic Management Journal, 22: Malerba, F., & Orsenigo, L Technological entry, exit and survival: An empirical analysis of patent data. Research Policy, 28: March J.G Exploration and exploitation in organizational learning. Organization Science, 2: Martin, X., & Mitchell, W The influence of local search and performance heuristics on new design introduction in a new product market. Research Policy. 26: McGrath, R Exploratory learning, innovative capacity, and managerial oversight. Academy of Management Journal, 44, Meeks, G., & Meeks, J Profitability measures as indicators of post-merger efficiency. Journal of Industrial Economics, 29: Megna, P., & Klock, M The impact of intangible capital on Tobin's q in the semiconductor industry. American Economic Review, 83: Merchant, H., & Schendel, D How do international JVs create shareholder value? Strategic Management Journal, 21: Meyer-Krahmer, F., & Schmoch, U Science-based technologies: University-industry interactions in four fields. Research Policy, 27, Miner, A., Bassoff, P., & Moorman, C Organizational improvisation and learning: A Field study. Administrative Science Quarterly, 46: Morck, R., Shleifer, A., & Vishny, R. W Management ownership and market variation: An empirical analysis. Journal of Financial Economics, 20: Morck, R., & Yeung, B Why investors value multinationality. Journal of Business, 64: Mowery, D. C., Oxley, J. E., & Silverman, B. S Technological overlap and interfirm cooperation: implications for the resource view of the firm. Research Policy, 27: Mulkay, M.J Sociology of the scientific research community. In I. Spiegel-Rosing & D. de SollaPrice (Eds.), Science, technology, and society: A cross disciplinary perspective, Beverly Hills, CA: Sage Publications. Narin, F., Hamilton, K. & Olivastro, D The increasing linkage between U.S.technology and publicscience. Research Policy, 26: Nightingale, P A Cognitive Model of Innovation. Research Policy, 27:

44 Nonaka, I., Takeuchi, H., & Umemoto, K A theory of organizational knowledge creation. International Journal of Technology Management, v11(7,8): Pavitt, What makes basic research economically useful?. Research Policy, 20(2): Ranft, A.L, & Lord, M.D Acquiring new technologies and capabilities: A grounded model of acquisition implementation. Organization Science, 13: Rip, A Science and technology as dancing partners. In P. Kroes & M. Bakker (Eds.), Technological development and science in the industrial age (pp ). Netherlands: Kluwer. Robins, J., & Wiersma, M A resource based approach to the multibusiness firm. Strategic Management Journal, 16: Rosenkopf, L., & Nerkar, A Beyond local search: Boundary-spanning, exploration, and impact in the optical disk industry. Strategic Management Journal, 22: Rothaermel, F. & Hill, C Technological discontinuities and complementary assets: A longitudinal study of industry and firm performance. Organization Science, 16: Rothaermel, F. T., Hitt, M. A. & Jobe, L. A Balancing vertical integration and strategic outsourcing: Effects on product portfolio, product success, and firm performance. Strategic Management Journal, 27( 11): Rumelt, R. P Strategy, structure, and economic performance. Boston, MA: Harvard University Press. Saikat, C., & Tabrizi, B Capturing the real value in high-tech acquisitions. Harvard Business Review, 77: Shelton, L. M Strategic business fits and corporate acquisitions: Empirical evidence. Strategic Management Journal, 9: Silverman, B Technological resources and the direction of corporate diversification: Toward an integration of the resource-based view and transaction cost economics. Management Science, 45: Simonin, B The importance of collaborative know-how: An empirical test of the learning Organization. Academy of Management Journal. 40: Singh, H., & Montgomery, C. A Corporate acquisition strategies and economic performance. Strategic Management Journal, 8: Sorensen, B. J., & Stuart, E. T., Aging, obsolescence, and organizational innovation. Administrative Science Quarterly, 20: Stuart, T. E., & Podolny, J. M Local search and the evolution of technological capabilities. Strategic Management Journal, 17:

45 Tanriverdi, H. & Venkatraman, N Knowledge relatedness and the performance of multibusiness firms. Strategic Management Journal, 26: Teece, D., Pisano, G. & Shuen, A Dynamic capabilities and Strategic Management. Strategic Management Journal, 18: Teece, D Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15, Thirtle, C. G. & Ruttan, V. W The Role of Demand and Supply in the Generation and Diffusion of Technical Change. Harwood Academic Publishers, New York. Travlos, N Corporate takeover bids, methods of payment, and bidding firms stock returns. Journal of Finance, 42: Trajtenberg, M A penny for your quotes: Patent citations and the value of innovations. Rand Journal of Economics, 21: Tripsas, M Unraveling the process of creative destruction: Complementary assets and incumbent survival in the typesetter industry. Strategic Management Journal, 18: Tushman, M. L., & Rosenkopf, L Organizational determinants of technological change: Toward a sociology of technological evolution. Research of Organizational Behavior, 14: Tushman, M. L., & Anderson, P Technological Discontinuities and Organizational Environments. Administrative Science Quarterly, 31: Uhlenbruck, K., Hitt, M. A., & Semadeni, M Market value effects of acquisitions involving Internet firms: A resource-based analysis. Strategic Management Journal, Vol. 27(10): Utterback, J.M., & Abernathy, W.J A dynamic model of process and product innovation, Omega, 3: Vermeulen F. & Barkema H Learning through acquisitions. Academy of Management Journal, 44: Von Hippel, E Lead Users: A Source of Novel Product Concepts. Management Science. 32:

46 Table 1 Science & Technology Similarity and Complementarity Measures Technology similarity is calculated as degree of patent overlap between the target and acquirer, weighted by the importance of each patent class for the Technology Similarity acquirer. Technology Similarity = 417 OverlapClassi NumberOfAcquirerPatentsClassi i= 1 TotalPatentsA & T AcquirerPatents Technology Complementarity Technology complementarity is operationalized as the overlap in patents in the same Subcategory but in a different Class. 36 OverlapSubcategory NumberOfAcquirerPatentsSubcategory OverlapClass Technology Complementarity = ( i i ) i= 1 TotalPatentsA& T AcquirerPatents TotalPatentsA& T The degree of science similarity is calculated as the overlap in the two firms Research Communities (RCs). Science Similarity 6 Science Similarity = (Number of Overlapping RCs) / (Total RCs A&T participate in) Science Complementarity Science Complementarity = [(Number of Non-Overlapping RCs in Overlapping Disc1) / (Total RCs(A&T) in Disc1)]+.+ [(Number of Non-Overlapping RCs in Overlapping Disc n) / (Total RCs(A&T) in Disc n)] Science complementarity calculated as the extent to which the target and acquirer build on the same science disciplines (Disc) but not on the same research communities thus capturing the extent to which they draw from related but different areas of science. It is based on the number of unique (nonoverlapping) research communities each firm is in within the science disciplines that they share. 6 Science similarity is calculated differently than technology similarity (without the weights) because each RC is distinct for each firm whereas for technology similarity a firm may have many patents in the same patent class. 44

47 Table 2 Descriptive Statistics Variable Mean s.d Deal Value R&D Intensity Cost Reductions *** Prior acquisition 4 experience # of Post M&A Acquisitions * 6 Acquirer Diversification ** Relative Size Technology Similarity * ** 9 Science Similarity Technology 10 Complemenatrity ** 0.38 ** * Science Complementarity ** Innovation 12 Quantity * ** 0.40 ** 0.37 ** ** Innovation 13 Quality * ** * * Acquirer 14 ROA(Industry Weighted) ** N = 95 The dependent variables are reported for p< 0.10 * p< 0.05 ** p< 0.01 ***p<

48 Table 3 The Effect of Knowledge Complementarities on Innovation Quantity and Quality (Hypothesis 1) Variables β/ Std. Error β/ Std. Error β/ Std. Error β/ Std. Error Innovation Quantity Innovation Quality Constant -2.66** -2.85** -2.30* -2.32* Product/Market Similarity 2.21* *** 4.91*** Industry No of acquisitions post M&A 3.58*** 3.66*** 5.52*** 5.51*** Prior Acquisition Experience 6.03*** 6.02*** 2.06* 2.04* Acquirer ROA Relative size (T/A) ** -3.12** R&D Intensity 1.92* 2.08* 4.38*** 4.40*** Acquirer Diversification ** -2.58* Cost Reductions Technology Complementarity Science Complementarity 2.89** * Technology Complementarity X Science Complementarity 2.53* 1.82 Adj- R R x *** *** 94.52*** 94.73*** 46

49 Table 4 The Effects of Technology Complementarities and Science Similarities on Innovation Quantity and Quality (Hypothesis Table 5 2) The Effects of Technology Complementarities and Product/Market Similarities on Innovation Quantity and Variables β/ Std. Error β/ Std. Error Quality (Hypothesis β/ Std. Error 3) β/ Std. Error Innovation Quantity Innovation Quality Constant ** -2.57* Product/Market Similarity * 5.15*** 4.92*** Industry No of acquisitions post M&A 2.86** 2.89** 5.99*** 5.97*** Prior Acquisition 6.54*** 6.56*** 2.09* 2.09* Experience Acquirer ROA Relative size (T/A) -2.19* *** -3.75*** R&D Intensity 2.54* 2.32* 4.51*** 4.49*** Acquirer Diversification * -2.43* Cost Reductions Technology Complementarity Science Similarity * 2.26* Technology Complementarity X Science Similarity Adj- R R x *** 93.31*** N=218 p< 0.10 * p< 0.05 ** p< 0.01 *** p<

50 Variables β/ Std. Error β/ Std. Error β/ Std. Error β/ Std. Error Innovation Quantity Innovation Quality Constant * -2.92** Industry *** No of acquisitions post M&A 2.92** 2.96** 5.61*** 5.72*** Table 6 Prior Acquisition 6.56*** 7.02*** 2.17* 1.91* Experience The Effects of Technology Complementarities and Acquirer Diversification on Innovation Quantity and Acquirer ROA Quality (Hypothesis 4) Relative size (T/A) -2.20* *** -4.21*** R&D Intensity 2.55* *** 5.16*** Acquirer Diversification * -2.55* -2.83** Cost Reductions Technology Complementarity *** * Product/Market Similarity * 4.96*** 4.60*** Technology Complementarity X Product/Market Similarity 3.30*** -2.73** Adj- R R x N=218 p< 0.10 * p< 0.05 ** p< 0.01 *** p<

51 Variables β/ Std. Error β/ Std. Error β/ Std. Error β/ Std. Error Innovation Quantity Innovation Quality Constant -1.99* -2.03* Product/Market Similarity *** Industry * No of acquisitions post M&A 2.92** 2.93** 5.61*** 5.68*** Prior Acquisition 6.56*** 6.59*** 2.17* 2.21* Experience Acquirer ROA Relative size (T/A) -2.20* -2.08* -3.89*** -3.74*** R&D Intensity 2.55* 2.66** 4.56*** 4.78*** Cost Reductions Technology Complementarity * Acquirer Product Diversification * -3.48*** Technology Complementarity X * Acquirer Product Diversification Adj- R R x 49

52 FIGURE 1 Knowledge Breadth, Product-Market Focus and Post M&A Innovative Performance Technology Complementarity Science Complementarity Innovation Outcomes - Quantity Product-Market Relatedness Acquirer Diversification - Quality 50

53 FIGURE 2 Science/Technology Relatedness in One Science/Technology Area Complementarity Unrelatedness Unrelatedness Similarity 51

54 FIGURE 3 S NBER Data Overall science # of # of Similar Area Complementa Are 52

55 FIG URE 4 Patent Class 424 Sub-Category 31 Drugs Patent Class 514 Category 3 Drugs & Medical CRP Data Patent Class 435 Sub-Category 33 Biotechnology Patent Class 800 Exa mple s of Nest ed Rela tions hips withi n the NBE R and CRP Data Science Discipline 2 Science Discipline 1 Science Discipline 3 RC1 RC1 RC2 RC2 RC2 RC3 RC3 53

56 FIGURE 5 Science & Technology Relatedness Similar Areas Firm A Biotechnology Firm B Category: Drugs & Molecular Microbiology Medical Unrelated Complementary Electrical Devices Areas Areas Organic Compounds Firm A Firm B Total Patents: 20 Total Patents: 20 Biotechnology Patents: 12 Biotechnology Patents: 12 54

57 Molecular Biology Patents: 2 Microbiology Patents: 2 Electrical Devices Patents: 6 Organic Compounds Patents: 6 55

58 FIGURE 6 Interaction Effects of Technology Complementarity and Science Complementarity on Innovation Quantity Innovation Quantity Low High SC Low SC Technology Complementarity High SC = Science Complementarity Interaction effects of Technology Complementarity with Science Complementarity on Innovation Quality 1 Innovation Quality 0.5 High SC 0 Low Technology Complementarity Low SC High 56

59 FIGURE 7 Interaction Effects of Technology Complementarity and Product/Market Similarity On Innovation Quantity Innovation Quantity 100 High PS 50 Low PS 0 Low Technology Complementarity High PS = Product/Market Similarity High PS = same 4-digit primary SIC APD=Acquirer Product Diversification High APD Low PS Low APD High PS Low High 57