Appropriating Returns From Innovation PDF

CHAPTER 1 APPROPRIATING THE RETURNS FROM INNOVATION Marco Ceccagnoli and Frank T. Rothaermel ABSTRACT This chapter explores the extent to which an innovator is able to capture innovation rents. After examining the two main drivers of such rents, the strength of the appropriability regime and the ownership of specialized complementary assets, the chapter examines how their interaction is so critical in affecting imitation, commercialization options, and firm performance. After reviewing the underlying conceptual framework and empirical evidence, and using a perspective that cuts across both time and industries, the authors then discuss the implications of innovation profits for the resources to be devoted to the discovery of new or improved product and processes. 1. INTRODUCTION Although signiﬁcant science and engineering competencies are needed to invent new processes and products, technological prowess that underlies process and product innovations is simply not enough to beneﬁt from innovation. While invention is a necessary ﬁrst step to innovation, it is not Technological Innovation: Generating Economic Results Advances in the Study of Entrepreneurship, Innovation and Economic Growth, Volume 18, 11–34 Copyright r 2008 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1048-4736/doi:10.1016/S1048-4736(07)00001-X 11 12 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL sufﬁcient for commercial success (Teece, 1986). Innovators frequently fail to appropriate the returns to their innovations. This implies that protecting the returns to innovation is a key strategic challenge in technology-intensive industries. Commercially successful innovations create temporary mono- polies, which in turn enable ﬁrms to extract transitory Schumpeterian rents. In high-technology industries, competitive advantage can be sustained only through a string of continuous innovations.1 Thus, a ﬁrm’s ability to appropriate rents from innovation determines its performance and continued survival. Table 1 depicts several high-proﬁle examples in which innovators lost to imitators, because the innovators were unable to appropriate the returns to their own innovation(s). Why does this happen so frequently? To answer this question, we focus on two factors highlighted by Teece’s seminal treatise on proﬁting from technological innovation: the appropriability regime and the complementary assets (Teece, 1986; Abernathy & Utterback, 1978; Anderson & Tushman, 1990). Today, it is widely accepted that innovators seeking to proﬁt from their inventions must understand the strength of the appropriability regime and the nature of the complementary assets required to commercialize their inventions. The commercialization of the CAT scanner provides a well-known example in which the innovator, Electrical Musical Instruments (EMI), lost to the imitator, GE Medical Systems.2 In the 1970s, EMI was a widely diversiﬁed British multinational corporation holding, for example, the rights to The Beatles records and competing in phonographic records, movies, and advanced electronics. Based on breakthrough research conducted in the 1960s by Godfrey Hounsﬁeld, a senior research engineer at EMI (and 1976 Nobel Laureate in Medicine), EMI developed the CAT scanner, originally Table 1. Innovators Failing to Appropriate the Returns to Innovation. Innovator Innovation Lost to Imitator EMI CAT scanner GE Medical Systems RC Cola Diet cola Coca-Cola and Pepsi Bowmar Pocket calculator TI, HP DeHavilland Commercial jet Boeing Ampex Video recorder Matsushita MITS PC Apple, IBM Xerox GUI interface Apple, Microsoft Prodigy Online service AOL, EarthLink, other ISPs Appropriating the Returns from Innovation 13 designed to take three-dimensional pictures of the brain, and later of the entire human body. This invention is hailed as the most signiﬁcant technological breakthrough in radiology since William Conrad Röntgen (Nobel Laureate in Physics, 1901) discovered the use of X-rays for imaging in 1895. Moreover, the invention of the CAT scanner paved the way for follow-up innovations like nuclear magnetic resonance tomography. In spring 1972, EMI launched the CAT scanner. Despite being aware that it lacked some of the requisite manufacturing and distribution assets, EMI decided to go it alone rather than to license its technology to a strong incumbent in medical devices like GE Medical Systems. Prior to market entry, the EMI management had little understanding of how effectively their patents would protect their innovation, and of the importance of complementary assets such as large-scale manufacturing and a distribution and marketing network. Still, EMI was able to create enormous excitement about its breakthrough innovation, which was ﬁrst demonstrated at a medical conference. Based on its technological lead, EMI became the market leader worldwide early on, with a strong position in the United States. Yet EMI was unable to sustain this lead because it could not satisfy the surge in demand, given its limited production capabilities, or solve the technological problems it encountered when setting up a production facility in the United States. By the mid-1970s, entrants into the CAT scanner business, like GE Medical Systems and others with strong technological and complementary assets, began to capture signiﬁcant market share from EMI. It is important to recall that GE did not invent the CAT scanner, but GE soon became the market leader because it possessed the requisite complementary assets necessary to succeed in this new market – especially large-scale manufactur- ing and a distribution network combined with a strong technical maintenance force. On the other hand, the innovator, EMI, was unable to acquire or develop the needed complementary assets to sustain its initial lead. This deﬁciency eventually led EMI to exit the market. To add insult to injury, the poor-performing unit EMI Medical was acquired by GE Medical Systems. The CAT scanner example clearly highlights the competitive race between innovators and imitators. The innovator (EMI) races to acquire complementary assets, and the imitator (GE Medical Systems) races to build the technological assets necessary to create the innovation, frequently through reverse engineering, and without infringing on the innovator’s patents. 14 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL 2. APPROPRIABILITY REGIME AND COMPLEMENTARY ASSETS: THE TEECE FRAMEWORK EMI’s strategy neglected the two most important determinants of innovation proﬁts: the appropriability regime and the specialized comple- mentary assets. The appropriability regime mainly depends on legal and technological factors. On one hand, the realization of rents from innovation depends on strong, or effective, intellectual property rights (IPR) protection by the legal system. On the other hand, characteristics of technology, such as degree of codiﬁcation, complexity, and ease of reverse engineering, determine the height of barriers to imitation, which in turn affect the ease with which rivals can imitate the innovation. In the EMI case, while the CAT scanner was a remarkable advance in medical technology, it only re-combined simple and well-known computing, X-ray, and imaging technologies in a new, albeit revolutionary, way. Once the idea of re-combining the different elements became widely known, it was difﬁcult to protect because it was easy to replicate through reverse engineering. Moreover, patents were not effectively enforced by both companies and courts, partly due to fears of their anti-competitive effects and partly due to the view that patents were a cost- rather than proﬁt-generating activity. As a result, the appropriability regime that EMI faced when commercializing the CAT scanner was weak. The second fundamental component of appropriability is the ownership of specialized complementary assets. Teece (1986) highlighted the importance of complementary assets in understanding the performance implications of a new technology when he examined the reason many innovators were unable to capture the economic rents ﬂowing from their innovations. He argued that the commercialization of an innovation ‘requires that the know-how in question be utilized in conjunction with other capabilities or assets. Services such as marketing, competitive manufacturing, and after-sales support are almost always needed. These services are obtained from complementary assets, which are specialized’ (Teece, 1986, p. 288). The commercialization of the CAT scanner provides a compelling example: the innovator, EMI, lost to the follower, GE Medical Systems, because EMI lacked specialized complementary assets. In his conceptual framework, Teece (1986) differentiated among three different types of complementary assets: generic, specialized, and cospecia- lized. Complementary assets that are generic need not be adjusted to the Appropriating the Returns from Innovation 15 innovation, because they can frequently be contracted for in the market on competitive terms. General purpose manufacturing equipment falls into this category. Specialized complementary assets exhibit unilateral dependence between the innovation and the complementary assets, and cospecialized complementary assets are characterized by a bilateral dependence. GE Medical System’s stellar reputation for quality and service in hospital equipment is considered a specialized complementary asset, whereas specialized repair facilities for Mazda’s rotary engine would be a cospecialized complementary asset. Because the distinction between unilateral and bilateral dependence of the complementary assets and the innovation in question is not critical to our analysis, we use the term specialized complementary assets here to denote both specialized and cospecialized complementary assets. Why are complementary assets so critical in commercializing innovation? When large-scale and high-quality manufacturing capabilities are necessary complementary assets, the owner of such assets is in a position to satisfy a large surge in customer demand, while maintaining product quality. A lack of large-scale manufacturing capabilities was the reason, for example, that innovator Immunex, a biotechnology ﬁrm, lost out to second-mover Johnson & Johnson, a healthcare conglomerate, in commercializing a biotechnology- based drug for rheumatoid arthritis. Immunex was the innovator in this market through its breakthrough development of the drug Enbrel in 1998, and its sales reached quickly $750 million in 2001. Surprised by the large demand for its highly successful new drug, Immunex had not created the necessary large-scale manufacturing capabilities to satisfy such an exponen- tial surge in demand. This strategic oversight provided Johnson & Johnson an opportunity to enter the market for biotechnology-based rheumatoid arthritis drugs with its own product (Remicade), developed by its fully owned subsidiary Centocor, which by 2002 had closed the lead held by Immunex’s Enbrel. Immunex’s innovative advantage dissipated due to a lack of the necessary complementary assets in manufacturing (Hill & Jones, 2007). Moreover, a large-scale manufacturing capability allows the company to ride down the experience curve faster due to learning effects and scale economies, and thus reach a low-cost position that is not attainable by competitors lacking such manufacturing capabilities. This was precisely one of the problems EMI faced. While it built a manufacturing plant in the United States to supply the largest market for medical devices in the world, EMI was unable to create a manufacturing capability necessary to produce the quantity and quality that would satisfy demand in the United States. 16 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL This problem exacerbated another problem faced by EMI: its lack of knowledge of the U.S. market for medical equipment and the way hospitals purchase and maintain such high-ticket items. As a case in point, in the U.K., due to the nature of its socialized healthcare system, only a few regional hospitals are equipped with expensive medical devices like CAT scanners or MRI systems; thus, the market for such high-ticket items is relatively thin. This was the mental mindset of EMI’s managers when entering the U.S. market. They did not realize that in the United States, competition among hospitals is decentralized, and hospitals compete with one another precisely by providing the latest advances in medical devices. Thus, most hospitals in the United States are equipped with a set of high- ticket items such as CAT scanners. Overall, the delivery of healthcare in the United States is much more capital-intensive than in Europe, where it tends to be more labor-intensive. This difference has signiﬁcant implications for the demand for medical devices and explains the surge in U.S. demand for CAT scanners, which EMI had not anticipated. In summary, strategy scholars have highlighted the importance of ownership of specialized complementary assets in proﬁting from innovation. These assets are frequently built over long periods of time and thus are path dependent and idiosyncratic (Teece, Pisano, & Shuen, 1997). Their market availability is limited because ﬁrms tend to gain control over them to avoid potential bargaining problems. Overall, specialized complementary assets constitute the bulk of a ﬁrm’s resources and capabilities that are valuable and difﬁcult to imitate, and they can therefore be a source of sustainable competitive advantage (Barney, 1991). 2.1. Interaction between Appropriability Regime and Complementary Assets In this section, we discuss who – the innovator or imitator – is more likely to extract innovation rents. In Section 2.2, we discuss in more detail the strategic options on which an innovator can draw when attempting to commercialize its innovations. The interaction between the strength of the appropriability regime and the ownership of specialized complementary assets determines the degree to which ﬁrms proﬁt from their innovations. A strong appropriability regime is typically sufﬁcient to capture at least a positive fraction of the innovation rents. But even in such a case, a greater degree of specialization in complementary assets corresponds to greater rents for its owner. When the Appropriating the Returns from Innovation 17 innovator owns such assets, it can capture almost all of the value associated with its innovation. When assets are specialized and owned by a different ﬁrm, rents have to be shared through an alliance, which in high-tech industries typically takes the form of technology licensing agreements (discussed in Chapter 8), such as in the pharmaceutical industry after the emergence of biotechnology (Rothaermel, 2001a, 2001b; Rothaermel & Hill, 2005). Teece’s (1986) conceptual framework depicting the interaction between the appropriability regime and the complementary assets is summarized in Fig. 1. The case of weak appropriability is analyzed by Teece in greater detail, most likely because during the decades preceding his work, courts typically provided weak protection to patent holders. Weak appropriability and generic complementary assets seem to be the unfortunate case of many entrepreneurial ventures seeking to ‘build a better mousetrap.’ Think about simple toys, for example, where entrepreneurial inventors often introduce tiny improvements from which they hope to generate quick revenues. Such simple inventions, however, are easily imitated and complementary assets are easily acquired, with customers appropriating most of the value created by the innovations. Appropriability Regime Weak Strong Complementary Complementary Assets Assets Generic Specialized Generic Specialized Consumers capture Innovator captures minority Innovator captures Innovator and owners of most of the value share of value. most of the value specialized assets share (Ex. simple toys, Owners of specialized assets (Ex. Semiconductor value (Ex. Biopharma); consumer products) capture largest share design firms may own both IP/comp. (Ex. EMI-CAT Scanner) outsourcing to assets (Eli Lilly/Prozac) foundries) Fig. 1. The Teece Framework. 18 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL The combination of a weak appropriability regime and specialized complementary assets typically allows the owners of such assets to capture the lion’s share of the value created by the innovation. This is the case exempliﬁed by the EMI/GE race. With stronger appropriability, the innovator usually captures a greater share of the proﬁts. It may be able to capture most of the proﬁts if it is able to easily acquire the necessary complementary assets. When specialized assets are required, an alliance should allow the parties to earn a return commensurate with the assets they bring to the table and with their respective bargaining power. A strong appropriability regime typically safeguards the innovator, which can disclose and protect its inventions to its potential alliance partners without fear of imitation. A strong appropria- bility regime does not simply happen by coincidence, but can be strategically enacted by the innovator not only through patenting, but also through following up with aggressive patent litigation. The U.S. semiconductor ﬁrm Intel is said to follow such a legal strategy (Somaya, 2003). 2.2. Appropriability Regime, Complementary Assets, and Commercialization Strategies Innovations create opportunities for companies to capture ﬁrst-mover advantages and thus temporary monopolies (Hill & Jones, 2007). But how should the innovator leverage its innovation towards commercial success and sustained competitive advantage? While we have focused more on theoretical descriptions by highlighting who captures the rents to innovation above, in this section we focus on the strategies available to the innovator in a more normative fashion: answering the question what an innovator should do given certain scenarios. An innovator basically has three strategic options at its disposal: (1) develop and commercialize the innovation itself, if necessary, through forward vertical integration; (2) develop and commercialize the innovation jointly with a partner through strategic alliances or a joint venture; and (3) license the innovation to another company or companies, and let them develop and market the innovation in exchange for royalties. The optimal strategy to be pursued depends upon (1) the availability and the type of complementary assets; (2) the height of imitation barriers, addressing the degree of difﬁculty of imitating the innovation by competitors (which is determined by the appropriability regime); and (3) the number of capable Appropriating the Returns from Innovation 19 competitors, which interact with the strength of the appropriability regime in determining the likelihood of imitation. The ﬁrst question the innovator must answer is whether it possesses the necessary complementary assets to commercialize the new technology. We discussed different types of complementary assets above. Assuming the innovator possesses specialized complementary assets to commercialize the innovation, the next question to consider is the height of barriers to imitation. These barriers deﬁne the degree of difﬁculty competitors face when attempting to imitate the innovation. Assuming the barriers to imitation are high, due to a strong appropriability regime, and the number of capable competitors is low, the inventor should go it alone – i.e., pursue a forward vertical integration strategy. The innovator will then be in a position to leverage its complementary assets to extract monopoly rents from the innovation, and barriers to imitation will delay entry. If the number of capable competitors remains low, the innovator might be able to build a sustained competitive advantage. More often than not, however, the innovator does not possess the required complementary assets to commercialize the innovation. If the barriers to imitation remain high (due to a strong appropriability regime) and the number of capable competitors is not too large, the innovator may proﬁt from the innovation through developing it jointly with the holder of complementary assets through an alliance or joint venture. While an alliance is a contractual agreement between two independent parties to share knowledge and resources and to co-develop product and processes, joint ventures are newly established third entities generally created by two parent companies to accomplish certain tasks, such as developing a new product or process. Alliances tend to be non-equity, contract-based cooperative agreements, whereas joint ventures are equity-based through setting up a third organization. As a consequence, non-equity alliances are much more frequent, although joint ventures are considered to establish stronger ties between ﬁrms. Intensive inter-ﬁrm cooperation based on alliances and joint ventures is a scenario that has played out in the pharmaceutical industry after the emergence of biotechnology; thus, one can now observe a (temporary) cooperative equilibrium between the innovating biotechnology ﬁrms and the large incumbent pharmaceutical companies (Teece, 1992; Gans & Stern, 2000; Rothaermel, 2000). In this industry, thousands of alliances and joint ventures have been documented in which the returns to innovation are shared by biotechnology and pharmaceutical companies (Rothaermel & Deeds, 2004). The distribution of rents, in turn, depends on the relative bargaining power of each party. 20 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL If the innovator lacks the necessary complementary assets and the barriers to imitation are low due to a weak appropriability regime combined with a large number of capable competitors, then the innovator should license the innovation to at least capture some of the innovation rents. Not only does imitation generally cost only 40–60% of the innovation, but imitation of an innovation through reverse engineering, for example, also is frequently possible within a few short years. While EMI held patents on its CAT scanner, the barriers to imitation were low, because GE Medical Systems (one of the ﬁrst customers to purchase a CAT scanner from EMI) was able to quickly imitate the CAT scanner through reverse engineering and thus invent around EMI’s thicket of patents. Thus, the barriers to imitation for the CAT scanner were low. Moreover, EMI faced a very capable and aggressive competitor in GE Medical Systems. Given this situation, EMI probably would have been better off either to enter into an alliance or joint venture or to license the commercialization of the CAT scanner directly to GE Medical Systems. The decision between these two remaining strategic options depends on the appropriability regime, which appeared to be weak initially when the CAT scanner was ﬁrst commercialized. This implies that EMI probably would have done best to license its innovation to GE Medical Systems. This is exactly the strategy Microsoft followed when faced with the question of how to commercialize its MS-DOS operation system. It opted for a non- exclusive license to IBM, which (involuntarily) aided Microsoft in making MS-DOS the ﬁrst and only industry standard for operating systems in the PC industry. Microsoft was able to defend this lead for over 25 years, through continuing innovations that leveraged the standard created through widespread adoption of MS-DOS. Microsoft’s innovation strategy thus resulted in a sustainable competitive advantage. Finding an appropriate partner to leverage the partner’s complementary assets to commercialize an innovation may not always be this straightfor- ward, because alliances often enable one partner to learn more than the other, and thus capabilities are frequently transferred. Here, the holder of complementary assets would be interested in obtaining the R&D capabilities of the innovator, while protecting its complementary assets. The innovator has the opposite motivation. The result is that learning races frequently ensue in alliances, especially in alliances initiated to commercialize innovations (Hamel, 1991). Note that often the holder of specialized complementary assets is more advantageously positioned to learn, and thus to appropriate innovation capabilities, because these ﬁrms tend to be larger and thus have more resources at their proposal, combined with an existing Appropriating the Returns from Innovation 21 R&D capability. In contrast, innovators frequently lack any competence in complementary assets, especially if those assets are downstream value chain activities like large-scale manufacturing, distribution, and after-sales service. Innovative ﬁrms tend to be small research-intensive outﬁts that exclusively focus on discovery and early-stage development of new products and processes. Going it alone through vertical integration may have to be achieved, absent any appropriate partners. Not infrequently, major innovations require complementary assets that are unavailable in the market, yet their nature is specialized and requires signiﬁcant sunk investments to be successfully commercialized. Downstream integration frequently takes substantial time if the capabilities are to be built from scratch. In such a case, both the demand for licensing and the potential rents to be realized are very low or absent, whereas the potential commercial success could be high. The key challenge, here, is to ﬁnd a partner willing to share the ﬁnancial risks of developing the cospecialized assets. With weak appropriability, however, partners may well be unwilling to share such risks, which are exacerbated by the high likelihood of imitation. Downstream integration remains the only alternative left. This option should be pursued only if the investment is expected to yield positive net returns to the innovator, a principle that should always guide rational investment decisions. 3. PROFITING FROM INNOVATION: EMPIRICAL EVIDENCE While we discussed theoretical decision points on how to commercialize an innovation, what does the empirical literature tell us about how well these theoretical conjectures hold up? Overall, the theoretical model presented holds up pretty well. Rothaermel and Hill (2005), for example, found support for the notion that the type of complementary assets (generic versus specialized) needed to commercialize a new technology is critical in determining the industry- and ﬁrm-level performance implications of a competence-destroying technological discontinuity. Competence-destroying technological discontinuities are radical innovations that emerge exogenous to incumbent industries, and to which established ﬁrms must respond to ensure continued survival. At the industry level, Rothaermel and Hill (2005) hypothesized, incumbent industry performance declines if the new technology can be 22 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL commercialized through generic complementary assets, whereas incumbent industry performance improves if the new technology can be commercia- lized through specialized complementary assets. At the ﬁrm level, they posited, an incumbent ﬁrm’s ﬁnancial strength has a stronger positive impact on ﬁrm performance in the post-discontinuity time period if the new technology can be commercialized through generic complementary assets. They further hypothesized, however, that an incumbent ﬁrm’s R&D capability has a stronger positive impact on ﬁrm performance in the post- discontinuity time period if the new technology has to be commercialized through specialized complementary assets. Drawing on multi-industry, time series, and panel data over a 26-year period to analyze pre- and post- discontinuity industry and ﬁrm performance, they found broad support for their theoretical model. Their ﬁndings are summarized in Table 2. Further, several empirical studies ﬁnd evidence for the innovation framework described on the right-hand side of Fig. 1 (Rothaermel, 2001a; Rothaermel, 2001b). Most of these studies have focused on the pharma- ceutical industry after the emergence of biotechnology. Here, the appro- priability regime is relatively strong, especially after the Supreme Court decision in 1980 that new life forms can be patented (Diamond v. Chakrabarty, 447 U.S. 303 (1980)). Moreover, specialized complementary assets (in the form of large-scale manufacturing, clinical trial, and regulatory management) as well as large sales forces are critical in commercializing new biotechnology drugs. Since the scientiﬁc breakthrough of genetic engineer- ing in the mid-1970s, numerous new biotechnology entrants demand access to the market for pharmaceuticals, which is controlled by a few incumbent pharmaceutical ﬁrms. These incumbent pharmaceutical ﬁrms have devel- oped path-dependent, ﬁrm-speciﬁc competencies with respect to certain drug and disease areas that are valuable, rare, and difﬁcult to imitate; thus, these competencies may, according to the resource-based view of the ﬁrm, form a basis of a competitive advantage (Barney, 1991). For example, Eli Lilly enjoys a dominant position in human insulin and growth hormones, while Hoffman-La Roche has developed a strong hold in anti-anxiety drugs. This degree of specialization reduces the number of potential strategic alliance partners for new biotechnology ﬁrms and further accentuates the value of the incumbents’ downstream, market-related value chain activities – i.e., specialization enhances the value of their complementary assets. Hence, these incumbents can beneﬁt from the technological breakthrough in biotechnology to the extent it enables them to create and extract innovation rents based on their specialized complementary assets, through strategic alliances, joint ventures, and licensing agreements with new Table 2. Technological Discontinuities, Complementary Assets, and Incumbent Industry and Firm Performance (Rothaermel & Hill, 2005). Technological Industry Examples Impact on Type of Impact on Effect on Stronger Effect on Discontinuity Incumbent Complementary Downstream Incumbent Incumbent Firm Upstream Assets Needed to Complementary Industry Performance Technological Commercialize Assets Performance Appropriating the Returns from Innovation Competencies New Technology PC, Electric arc Computer, steel Destroying Generic Destroying Decline Financial strength furnace Biotechnology, Pharmaceutical, Destroying Specialized Enhancing Improvement R&D capability wireless telecommunications telephony 23 24 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL Table 3. Top-Ten Biotechnology Drugs, 2001. Product Indication 2001 Sales Developer Marketer (millions) Procrit Red blood cell 3,430 Amgen Johnson & Johnson enhancement Epogen Red blood cell 2,109 Amgen Amgen enhancement Intron A Hepatitis C, certain 1,447 Biogen Schering-Plough cancers Neupogen Restoration of white 1,346 Amgen Amgen blood cells Humulin Diabetes 1,061 Genentech Lilly Avonex Multiple sclerosis 972 Biogen Biogen Rituxan B-cell non-Hodgkin’s 819 IDEC Genentech, IDEC lymphoma Enbrel Rheumatoid arthritis 762 Immunex AHP, Immunex Remicade Rheumatoid arthritis, 721 MedImmune Johnson & Johnson Chron’s disease Cerezyme Enzyme replacement 570 Genzyme Genzyme therapy Source: Standard and Poor’s Biotechnology Industry Report, May 2002. biotechnology ﬁrms. The emergence of a cooperative equilibrium in the biopharmaceutical industry has also been highlighted by other researchers and is exempliﬁed in Table 3, which depicts the top-ten selling biotechnology drugs in 2001. What is interesting to note is that none of the top-ten selling drugs was developed by the incumbent pharmaceutical companies. Thus, all new biotechnology drugs were discovered and developed by new biotechnology ﬁrms leveraging their R&D competencies in the new biotechnology paradigm. About half of these drugs were commercialized by incumbent pharmaceutical companies. This empirical outcome is in line with Teece’s theoretical predictions (Teece, 1986, 1992). It is important to note, however, that more recently several more biotechnology companies were able to integrate downstream, as there are now fewer cooperative arrangements between biotechnology ventures and large pharmaceutical companies to commercialize new drugs. Rothaermel and Deeds (2004) document a new product development process based on an alliance system orchestrated by biotechnology companies, by which the biotechnology ﬁrms reach upstream to universities for basic knowledge, and Appropriating the Returns from Innovation 25 then downstream to pharmaceutical companies to commercialize their innovations. While this integrated new product development process resonates with Teece’s framework, Rothaermel and Deeds also demonstrate that the new biotechnology companies withdraw from this integrated product development process in a discriminate fashion, as the new venture accrues more resources to discover, develop, and commercialize promising projects through vertical integration. They empirically tested their model on a sample of 325 biotechnology ﬁrms that entered into 2,565 alliances over a 25-year period; they found broad support for the hypothesized product development system and the negative moderating effect of ﬁrm size. Thus, the effect of complementary assets on ﬁrm performance is likely to change over time. This ﬁnding also resonates with the recent study of Rothaermel and Boeker (2008), who found, through studying over 32,000 dyads (i.e., pairs) between pharmaceutical and biotechnology companies over time, that a pharmaceutical company and a biotechnology ﬁrm are more likely to enter into an alliance based on complementarities when the biotechnology ﬁrm is younger. This ﬁnding echoes the theoretical conjecture above that the holder of complementary assets (e.g., a large pharmaceutical ﬁrm) is more likely to acquire the R&D skills necessary to create the innovation from the innovator (e.g., a biotech start-up) than the other way around. Evidence from litigation provides further support for this notion. For example, the ﬁrst biotechnology drug to be commercialized was Humulin, a human insulin, which was discovered and developed by the biotechnology ﬁrm Genentech and commercialized by the pharmaceutical company Eli Lilly in 1982. Later, however, Genentech sued Lilly, accusing it of misusing materials provided by Genentech to commercialize recombinant human insulin. In other words, Lilly was concerned that Genentech had appropriated relevant R&D skills through their alliance to commercialize Humulin. Differences in the strength of complementary assets have also been documented across different industries. Both changes in appropriability over time and across industries are analyzed in Section 3.1. 3.1. Degree of Appropriability and Inter-Industry Differences: Empirical Evidence Systematic empirical evidence on the effectiveness of different appropria- bility strategies for the U.S. manufacturing sector is available from the 1983 26 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL Yale University survey and the 1994 Carnegie Mellon University (CMU) survey, summarized for selected industries in Table 4 (Levin, Klevorick, Nelson, & Winter, 1987; Cohen, Nelson, & Walsh, 2000). Focusing on the more recent CMU survey ﬁrst, Table 4 suggests that the most effective mechanisms to protect product innovations across a wide number of industries is secrecy, closely followed by being ﬁrst to market, which captures the effectiveness of ﬁrst mover advantages (Lieberman & Montgomery, 1988). The ownership of specialized complementary assets represents the third most effective mechanism, whereas patent protection is rated as the least effective relative to these other mechanisms. Several policy and management changes lead us to expect that the relative strength of different appropriability strategies has changed since the early 1980s, about the time during which the earlier Yale survey was conducted. In particular, belief in the importance of patents and intellectual property (IP) protection in stimulating innovation is the main economic rationale underpinning the trend towards a strengthening of IP protection that has characterized the last two decades, particularly in the United States. In 1982, the Court of Appeals for the Federal Circuit was established to make patent protection more uniform. Indirectly, this also strengthened patent protec- tion. Plaintiff success rates, as well as damages in infringement, have also risen. In the early 1980s, we also witnessed an expansion of what can be patented, when the courts decided that life forms and software were both patentable. Patent coverage has been extended recently to business methods as well. Patents have also become a growing preoccupation of management. The comparison between the earlier 1983 Yale survey and the 1994 CMU survey, shown in Table 4, conﬁrms that substantial changes in appropria- bility conditions have taken place in the United States over time and across industries, as perceived by survey respondents.3 The data highlight that patents are more recently perceived as signiﬁcantly more important, with almost a 30% increase in the percentage of ﬁrms within industries ranking patents as the ﬁrst or second most important mechanism of appropriation. Being ﬁrst to market is also slightly more important, whereas ownership of complementary assets is slightly less important to protect the competitive advantage from an innovation. The sharper difference is related to the effectiveness of secrecy, with a change in the perceived effectiveness of over 90%. Sharper differences across time characterize some industries, such as computers, machinery, and controlling devices. Such variations reﬂect the fact that the strength of appropriability has an important endogenous component: exogenous changes in the appropriability regime may have a Table 4. Comparing the 1983 Yale and 1994 CMU Appropriability Surveys: Selected High-Tech Industriesa. % Firms within Industries Ranking Appropriability Strategy as First or Second Most Important Number of Patent Protection Secrecy Being First to Complementary Observations Market Assets Yale CMU Yale CMU % Yale CMU % Yale CMU % Yale CMU % Change Change Change Change Industrial chemicals 73 52 0.75 0.78 4% 0.59 0.98 66% 0.80 0.68 15% 0.79 0.78 1% Drugs and medicines 17 47 0.94 0.80 15% 0.53 0.91 72% 0.71 0.71 1% 0.71 0.51 28% General industrial 32 18 0.47 0.78 66% 0.41 0.94 132% 0.78 0.89 14% 0.81 0.83 3% machinery Appropriating the Returns from Innovation Computers 21 28 0.29 0.64 125% 0.43 0.79 83% 0.86 0.89 4% 0.62 0.61 2% Communication 17 22 0.41 0.62 50% 0.53 0.81 53% 0.88 1.00 13% 0.94 0.81 14% equipment Semiconductors 10 17 0.50 0.63 25% 0.20 0.94 369% 0.90 0.94 4% 0.70 0.75 7% Motor vehicles 24 27 0.63 0.76 22% 0.33 0.76 128% 0.71 0.92 30% 0.79 0.60 24% Aircraft and missiles 21 41 0.38 0.54 41% 0.48 0.95 99% 1.00 0.92 8% 0.71 0.62 14% Search and navigation 9 29 0.44 0.66 47% 0.67 0.97 45% 1.00 0.86 14% 0.89 0.83 7% equipment Measuring and 18 25 0.33 0.65 96% 0.28 0.87 213% 0.94 0.96 1% 0.78 0.74 5% controlling device Medical instruments 12 60 0.58 0.73 26% 0.50 0.83 67% 1.00 0.90 10% 0.83 0.72 14% Total manufacturing 650 852 0.53 0.67 28% 0.47 0.89 91% 0.84 0.87 4% 0.80 0.73 8% a Based on own computation using original respondent-level Yale and CMU surveys data. 27 28 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL different effect on ﬁrms’ use of different strategies in different industries within the same country (Hall & Ziedonis, 2001). In particular, the increase in ﬁrms’ propensity to patent, as a consequence of a stronger appropriability regime in industries such as electronics and semiconductors, has spawned patent portfolio races whose main objectives are both to discourage infringement suits and to strengthen incumbents’ bargaining positions in cross-licensing negotiations. Overall, considering that both patent protection and trade secrecy are knowledge-related proprietary strategies, the strength of the appropriability regime seems to have increased over time in the United States. Teece’s (1986) framework implies that we increasingly observe cases falling on the right-hand side of the tree represented in Fig. 1, where innovators capture a greater share of innovation rents due to a strengthened appropriability regime. This is consistent with the widespread belief that innovation is increasingly the key source of competitiveness and economic growth. Interestingly enough, changes in the appropriability regime that have taken place since the early 1980s likely have affected the evolution of the degree to which EMI has proﬁted from its CAT scanner itself. Indeed, EMI started to enforce its patents between the end of the 1970s and the beginning of the 1980s. Legal documents and company interviews suggest that the company, despite suffering imitation and bankruptcy of its medical device operation, was able to succeed against infringers and extract a substantial fraction of rents through ‘stick’ licensing.4 In a recent interview, IP managers familiar with these matters estimated that EMI realized over $100 million in ‘stick’ licensing revenues related to the CAT scanner.5 The strengthening of patent protection is expected to have a profound impact on the way a ﬁrm proﬁts from innovation as well. Consistent with Teece’s framework, Arora and Ceccagnoli (2006) provide systematic empirical evidence suggesting that ﬁrms lacking the specialized complemen- tary assets required to commercialize innovation typically license more when patent protection is strong, in contrast to ﬁrms that have specialized complementary assets, which license less. Their work also suggests that in a world of strong IPR, although technology buyers enjoy lower transactions costs and gain from trading technology, they also lose some bargaining power in technology alliances in favor of IP owners and therefore realize lower returns on the ownership of specialized complementary assets. This may in part explain the increasing downward pressure on the proﬁtability of ‘big Pharma,’ which seems to suffer in a world placing increasing rewards on the owners of upstream proprietary knowledge.6 Appropriating the Returns from Innovation 29 3.2. Quantifying the Returns Provided by Patent Strategies Assuming ﬁrms apply for patents if net beneﬁts of doing so are positive, Arora, Ceccagnoli, and Cohen (2007) have used survey-based responses on a ﬁrm’s propensity to patent (% of innovations for which a ﬁrm applies for patent protection) to compute an unobservable concept, the patent premium – i.e., the proportional increment to the value of an innovation realized by patenting. Results indicate that patents provide a positive expected premium only for a small fraction of innovations. In fact, on average, the relative magnitude of beneﬁts and costs suggests that ﬁrms expect to lose about 50% of an innovation’s value by patenting it in a broad set of manufacturing industries. Put differently, patenting the typical invention is not proﬁtable in most industries because the opportunity costs of patenting (including the cost of information disclosure, the likelihood of inventing around, and the cost of enforcement) are substantial. The patent premium is around unity for the typical patent portfolio of the average ﬁrm in biotechnology and pharmaceuticals, meaning that a ﬁrm expects no difference, on average, between payoffs realized by patenting or not. In medical instruments, instead, patenting the typical innovation is worthwhile. Only innovations for which there is a premium greater than unity are eventually patented. Indeed, the average expected premium for the innovations that ﬁrms choose to patent is about 1.5, suggesting that ﬁrms expect to earn, on average, a 50% premium over the no-patenting case. Such a premium, conditional on patenting, is about 1.6 in the health-related industries and 1.4 in electronics and semiconductors (see Fig. 2). Overall, these results suggest that even in All innovations Patented innovations 1.42 1.43 1.45 1.45 1.48 1.49 1.49 1.57 1.58 1.62 1.38 1.41 1.43 1.11.96.99.66.72.73.55.56.40.46.46.47.28 ts on les s rs ip ls y rs es h s ent ent ner tec nen ica cto ute o qu cin ati ssi acc rum rum Bio ne chi em du mp ort edi po mi ob con tio Ma l ch om nsp t Co nst m and Ins ,T ica nd ic c mi al i Tra ria od aft un sa Se ust dic Fo on mm rcr ug Ind ctr Me Ai Dr Co Ele Fig. 2. The Patent Premium (Arora et al., 2007). 30 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL those industries where patenting is not proﬁtable on average, some inventions are proﬁtable to patent and may actually provide large payoffs from doing so. This, however, does not mean that patenting is a sufﬁcient condition to proﬁt from innovation. Indeed, in most cases, patent strategies must be integrated with appropriate strategies to leverage or acquire complementary assets, exploit lead times, or maintain secrecy over other aspects of an innovation. In a recent study, Ceccagnoli (2008) directly links the degree of appropriability achieved through different strategies and the way ﬁrms enforce their patents to ﬁrm performance. He ﬁnds that among the various appropriability strategies considered in the previous section (see Table 4), the strength of a ﬁrm’s patent protection strategy and the ownership of specialized complementary assets are associated with a substantial increase in the stock market valuation of a ﬁrm’s R&D assets relative to tangible assets. He also ﬁnds that among the patent strategies that are increasingly and purposefully used by technology-intensive companies, patent preemp- tion – deﬁned as the patenting of substitute or complements of other innovations owned by the ﬁrm – tends to remarkably improve the rate of return to R&D investments, as valued by the stock market. Consistent with existing theories, his empirical ﬁndings also indicate that patent preemption tends to improve the proﬁts due to a ﬁrm’s R&D and ﬁrm performance, and this effect is higher for innovating incumbents with higher market power and those facing the threat of entry and it is lower when R&D competition is characterized by the discovery of drastic innovations. 4. THE TWO FACES OF APPROPRIABILITY: PROTECTION VERSUS INCENTIVES The degree to which a ﬁrm captures the value created through the introduction of its innovations has a dual function. It increases an innovating ﬁrm’s proﬁts and market power, which has been the focus of much of the previous discussion, but it also affects its inventive efforts. Previous empirical studies on the impact of appropriability have mostly focused on the ﬁrst effect. Theoretical work, in particular economic analysis of the impact of patent protection, has instead focused on the incentive effect. The main rationale of patent protection is indeed to stimulate innovative investments, while at the same time promoting the diffusion of technological knowledge. By providing restrictions to the use of patented Appropriating the Returns from Innovation 31 inventions, patent law provides the ability to recover the investment needed to introduce technological innovations, in exchange for disclosure of the technical details of the patented inventions to the public. The main social cost is the restriction in use, and thus the inefﬁciencies associated with monopoly protection. The empirical work presented above, and in particular the results of Table 4, have been interpreted as suggesting that the inducement provided by patents for innovation is small in most industries. However, these results do not imply that patents provide little incentives to invest in R&D. Indeed, the estimates of the patent premium suggest that patents could be effective for a small fraction of innovations and still provide substantial average returns. Moreover, incentive effects depend on the impact of appropriability on the marginal beneﬁts of R&D investments. Indeed, there is still no clear empirical consensus on the idea that greater appropriability of proﬁts due to innovation, conferred by patents or any other mechanism, actually stimulates investments in innovation. To address this point, recent economic studies have attempted to quantify the incentive effect of patents. In particular, Arora et al. (2003) provided robust evidence of a positive incentive effect of the strength of patent protection, using ﬁrm-level data from the CMU survey. They estimate an economic model in which ﬁrms’ R&D decisions depend on expected returns, which are conditioned by the effectiveness of patent protection. The study further recognizes that if one ﬁrm beneﬁts from stronger patent protection in a speciﬁc area, its competitors will also beneﬁt from it. Their quantitative estimates suggest that a 10% increase in the strength of appropriability provided by patent protection would increase R&D investments by 7%, the ﬁrm’s propensity to patent by 17%, and the number of patents applied for by each ﬁrm by 15%. Moreover, their results indicate that the incentive effect of patents varies substantially across industries, with the largest effect in pharmaceuticals, biotechnology, medical instruments, and computers. In semiconductors and communications equipment, the incentive effect of patent protection is much lower, although still positive and not negligible. 5. CONCLUSION Strategies used to capture the value created by innovative investment are a fundamental source of a ﬁrm’s competitive advantage. The degree to which ﬁrms proﬁt from innovation is critically affected by the interplay of imitation-related factors, such as ownership and strength of IP and the 32 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL number of capable innovators, and the ownership of specialized comple- mentary assets required for successful product and process market introduction. During the last two decades we have witnessed economic and legal changes, as well as evolving managerial practices, related to the strength of available appropriation strategies. These changes have affected the propensity to use different appropriation strategies, ﬁrm performance, and the division of labor and proﬁts from the value created by innovation, in a world that places increasing importance on innovation for ﬁrms’ competitiveness, productivity, and economic growth. Within this evolving competitive environment, the understanding of the relationship between appropriability and innovation investments is particu- larly important, not only for policy, but also for strategy and entrepreneurship. Appropriability conditions and the effective management of IP should indeed guide entrepreneurs and companies alike in their choices about allocating resources for the creation of value through technological innovations. NOTES 1. See discussion in Chapter 7. 2. See EMI and CT Scanner (A) and (B), Harvard Business Cases 9-383-194 and 383-195. 3. A sample of comparable ﬁrms was built using original respondent-level Yale and CMU surveys data, i.e., only using public ﬁrms operating in comparable industries. Each ﬁrm’s responses on both product and process appropriability mechanisms were used to compute a dummy equal to one if any mechanism was rated as the ﬁrst or second most effective in protecting the competitive advantage from its innovations. Table 2 shows the percentage of ﬁrms rating each mechanism as ﬁrst or second most effective. 4. In a 1983 legal document we ﬁnd clear and ofﬁcial evidence of the ‘stick’ licensing activities: ‘The CT manufacturing industry is a decade old. EMI, a corporation organized under the laws of England, started production of the CT scanners in the United States, in Chicago, Illinois, and made delivery of its ﬁrst orders in 1973. Several years ago, it discontinued making the machines and closed its Chicago plant. It licenses others to manufacture under these patents; in all, there are approximately ten licensees in the United States, including General Electric Company, Johnson & Johnson, and Pﬁzer, Inc. The latter three, in contested litigation, challenged the validity of plaintiff’s patents; however, those separate litigations, which extended over a two-year period, were settled before trial in return for licenses issued by EMI to the contestants.’ (EMI Limited, Plaintiff, v. Picker International, Inc., Defendant; No. 83,Civ. 0759, U.S. District Court for the Southern District of New York; 565 F. Supp. 905, 1983). Appropriating the Returns from Innovation 33 5. Interview by Marco Ceccagnoli with Dr Stephen Potter, Commercial Director of QED (patent licensing division of EMI, subject of a management buy-out in 1996: http://www.qed-ip.com), June 1st, 2003, INSEAD (Fontainebleau, France). 6. An increasing fraction of R&D expenses of large pharmaceutical companies includes the cost of developing drugs that are in-licensed from smaller biopharma- ceutical ﬁrms, which in turn aggressively safeguard their proprietary knowledge and are able to extract a signiﬁcant fraction of rents associated with their innovations. ACKNOWLEDGMENTS Rothaermel gratefully acknowledges ﬁnancial support from National Science Foundation Grant SES 0545544 (CAREER Award). Any opinions, ﬁndings, conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reﬂect the views of the National Science Foundation. REFERENCES Abernathy, W. I., & Utterback, J. M. (1978). Patterns of industrial innovation. Technology Review, 80, 40–47. Anderson, P., & Tushman, M. L. (1990). Technological discontinuities and dominant design: A cyclical model of technological change. Administrative Science Quarterly, 35, 604–633. Arora, A., & Ceccagnoli, M. (2006). Patent protection, complementary assets, and ﬁrms’ incentives for technology licensing. Management Science, 52, 292–308. Arora, A., Ceccagnoli, M., & Cohen, W. M. (2007). R&D and the patent premium. International Journal of Industrial Organization (forthcoming). Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17, 99–120. Ceccagnoli, M. (2008). Appropriability, preemption, and ﬁrm performance. Strategic Manage- ment Journal (forthcoming). Cohen, W. M., Nelson, R. R., & Walsh, J. P. (2000). Protecting their intellectual assets: Appropriability conditions and why U.S. manufacturing firms patent or not. NBER Working Paper 7552. Retrieved March 31, 2007, from http://www.nber.org/papers/w7552 Diamond v. Chakrabarty, 447 U.S. 303 (1980). EMI Ltd. v. Picker International, Inc., 565 F. Supp. 905 (S.D.N.Y. 1983). Gans, J. S., & Stern, S. (2000). Incumbency and R&D incentives: Licensing the gale of creative destruction. Journal of Economics and Management Strategy, 9, 485–511. Hall, B., & Ziedonis, R. H. (2001). The patent paradox revisited: An empirical study of patenting in the U.S. semiconductor industry, 1979–1995. RAND Journal of Economics, 32, 101–128. Hamel, G. (1991). Competition for competence and inter-partner learning within international strategic alliances [summer special issue]. Strategic Management Journal, 12, 83–103. 34 MARCO CECCAGNOLI AND FRANK T. ROTHAERMEL Hill, C. W. L., & Jones, G. R. (2007). Strategic management theory: An integrated approach (7th ed.). Boston: Houghton Mifﬂin. Levin, R. C., Klevorick, A. K., Nelson, R. R., & Winter, S. G. (1987). Appropriating the returns from industrial R&D. Brookings Papers on Economic Activity, 1987(3), 783–820. Lieberman, M. B., & Montgomery, D. B. (1988). First-mover advantages [summer special issue]. Strategic Management Journal, 9, 41–58. Rothaermel, F. T. (2000). Technological discontinuities and the nature of competition. Technology Analysis & Strategic Management, 12(2), 149–160. Rothaermel, F. T. (2001a). Incumbent’s advantage through exploiting complementary assets via interﬁrm cooperation. Strategic Management Journal, 22(6–7), 687–699. Rothaermel, F. T. (2001b). Complementary assets, strategic alliances, and the incumbent’s advantage: An empirical study of industry and ﬁrm effects in the biopharmaceutical industry. Research Policy, 30(8), 1235–1251. Rothaermel, F. T., & Boeker, W. (2008). Old technology meets new technology: Complementarities, similarities, and alliance formation. Strategic Management Journal, 29(1), 47–77. Rothaermel, F. T., & Deeds, D. L. (2004). Exploration and exploitation alliances in biotechnology: A system of new product development. Strategic Management Journal, 25, 201–221. Rothaermel, F. T., & Hill, C. W. L. (2005). Technological discontinuities and complementary assets: A longitudinal study of industry and ﬁrm performance. Organization Science, 16, 52–70. Somaya, D. (2003). Strategic determinants of decisions not to settle patent litigation. Strategic Management Journal, 24, 17–38. Teece, D. J. (1986). Proﬁting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15, 285–305. Teece, D. J. (1992). Competition, cooperation, and innovation: Organizational arrangements for regimes of rapid technological progress. Journal of Economic Behavior & Organization, 18, 1–25. Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18, 509–533.

Appropriating Returns From Innovation PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue