Intellectual property as a law of organization

Southern California Law Review

Volumn 84, Issue 4, 2011, Pages 785-858

  Barnett, Jonathan M ^a  

^a UNIVERSITY OF SOUTHERN CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 79958178601     PISSN: 00383910     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (213)

1. This approach builds upon and generalizes arguments set forth in Ashish Arora & Robert P. Merges, Specialized Supply Firms, Property Rights and Firm Boundaries, 13 INDUS. & CORP. CHANGE 451, 470-72 (2004). For other relevant contributions, see sources cited infra note 10.
2. See Paul J. Heald, A Transaction Costs Theory of Patent Law, 66 OHIO ST. L. J. 473, 473-78 (2005) (arguing that, independent of any exclusionary function, patents reduce transaction costs of organizing and monitoring team production of research and development ("R&D") and other innovation assets);
3. Clarisa Long, Patent Signals, 69 U. CHI. L. REV. 625, 643-55 (2002) (arguing that, independent of any exclusionary function, patents perform a signaling function that relieves informational asymmetries, especially between firms and investors).
4. For the leading account, see B. ZORINA KHAN, THE DEMOCRATIZATION OF INVENTION: PATENTS AND COPYRIGHTS IN AMERICAN ECONOMIC DEVELOPMENT, 1790-1920, at 106-27 (2005) (noting the significance of the U. S. patent regime in providing incentives for advancement of technology, as well as equal access to disadvantaged groups such as blacks and women).
5. See, for example, the Patent and Trademark Office's ("PTO's") reduced fee schedule for small entities, 37 C. F. R. § 1.27 (2010);
6. the Small Business Innovation Development Act of 1982, Pub. L. No. 97-219, 96 Stat. 217
7. (codified as amended at 15 U. S. C. §§ 631, 638 (2006));
8. the Small Business Research and Development Enhancement Act of 1992, Pub. L. No. 102-564, 106 Stat. 4249
9. (codified as amended at 15 U. S. C. § 631);
10. and, in the case of academic research institutions, the University and Small Business Patent Procedures (Bayh-Dole) Act of 1980, Pub. L. No. 96-517, 94 Stat. 3015
11. (codified as amended at 35 U. S. C. §§ 200-212 (2006))
12. (implemented by 37 C. F. R. pt. 401).
13. Asish Arora and colleagues in management literature, Robert Merges in legal literature, and Kenneth Sokoloff and Naomi Lamoreaux in economic history literature have pioneered this line of inquiry. See generally ASHISH ARORA, ANDREA FOSFURI & ALFONSO GAMBARDELLA, MARKETS FOR TECHNOLOGY: THE ECONOMICS OF INNOVATION AND CORPORATE STRATEGY (2001) (discussing the role of intellectual property rights in defining markets for technology); Arora & Merges, supra note 5 (discussing the role of intellectual property rights in emerging trends in choices of firm boundaries);
14. Robert P. Merges, A Transactional View of Property Rights, 20 BERKELEY TECH. L. J. 1477 (2005) [hereinafter Merges, Transactional View] (discussing the role of intellectual property rights in facilitating contracting between "legal strangers" by creating both precontractual liability and enforcement flexibility);
15. Robert P. Merges, Intellectual Property Rights, Input Markets, and the Value of Intangible Assets (Feb. 9, 1999) (unpublished manuscript) [hereinafter Merges, Input Markets], available at http://www.law.berkeley.edu/files/iprights. pdf (discussing how strong intellectual property rights incentivize firm specialization and enable the formation of distinct input markets). For contributions in economic history literature, see infra notes 79-86 and accompanying text.
16. For other contributions in legal literature on intellectual property and firm structure, see Oren Bar-Gill & Gideon Parchomovsky, Law and the Boundaries of Technology-Intensive Firms, 157 U. PA. L. REV. 1649, 1650 (2009) (discussing how the availability of intellectual property protection, rather than economic considerations, determine the boundaries of technology-intensive firms)
17. and Dan L. Burk & Brett H. McDonnell, The Goldilocks Hypothesis: Balancing Intellectual Property Rights at the Boundary of the Firm, 2007 U. ILL. L. REV. 575, 633-36 (arguing that intellectual property rights should be calibrated between firms and within firms to address internal and external transaction costs).
18. On intellectual property and market structure, see MARTIN J. ADELMAN ET AL., CASES AND MATERIALS ON PATENT LAW 38 (2d ed. 2003) ("In effect the existence of a patent system makes the industrial structure of a particular industry essentially irrelevant to the innovation process.")
19. and Martin J. Adelman, The Supreme Court, Market Structure, and Innovation: Chakrabarty, Rohm and Haas, 27 ANTITRUST BULL. 457, 460 (1982) (arguing that the relationship between market structure and innovation is unknowable, but that a strong patent regime is valuable because "it reduce[s] any diversity in incentives to innovate with respect to different market structures").
20. Inquiry into the relationship between intellectual property and firm structure traces back to David J. Teece, Firm Organization, Industrial Structure, and Technological Innovation, 31 J. ECON. BEHAV. & ORG. 193, 222 (1996) (recognizing "firm organization... [as] an important determinant of innovation"). This Article advances these bodies of scholarship in several respects: (1) it views specialized R&D suppliers (the focus of much of the existing literature) as a subset of a general case in which patents enable the efficient allocation of innovation and commercialization functions among least-cost providers; (2) it provides a consolidated framework that identifies links between the entry of upstream R&D suppliers, the unraveling of downstream portions of the supply chain, and the formation of secondary markets in supply chain inputs; (3) it exploits these relationships in order to isolate the circumstances in which patent coverage exerts marginal incentive effects, even assuming the existence of alternative instruments by which to capture innovation returns; and (4) it moves beyond theoretical argument by identifying organizational tendencies in technology markets that are consistent with these relationships.
21. See Kenneth J. Arrow, Economic Welfare and the Allocation of Resources for Invention, in NAT'L BUREAU OF ECON. RESEARCH, THE RATE AND DIRECTION OF INVENTIVE ACTIVITY: ECONOMIC AND SOCIAL FACTORS 609, 614-16 (1962).
22. F. Scott Kieff in particular has emphasized this point. See F. Scott Kieff, IP Transactions: On the Theory & Practice of Commercializing Innovation, 42 HOUS. L. REV. 727, 736-37 (2005) (exploring the role of intellectual property rights in crossing between innovation and market release);
23. F. Scott Kieff, Property Rights and Property Rules for Commercializing Inventions, 85 MINN. L. REV. 697, 703-04, 707-12 (2001) (same). (Pubitemid 33648758)
24. The recent focus on commercialization revives themes promoted by Edmund W. Kitch, The Nature and Function of the Patent System, 20 J. L. & ECON. 265, 265-67, 289-90 (1977) (defining the "prospect function" of patents as a claim system through which rights to innovate are auctioned off, thus lowering transaction costs in transmission of information between firms).
25. The leading evidence is found in survey studies covering large U. S. manufacturing firms, finding that, among legal and extralegal mechanisms for appropriating returns from R&D projects, firm managers (outside the pharmaceutical and chemicals industries) usually report that patents are among the least effective instruments and are rarely the "but for" condition for proceeding with an R&D project. See Richard C. Levin et al., Appropriating the Returns from Industrial Research and Development, in 3 BROOKINGS PAPERS ON ECONOMIC ACTIVITY: SPECIAL ISSUE ON MICROECONOMICS 783, 789-91, 793-98 (Martin Neil Baily & Clifford Winston eds., 1987) (surveying R&D managers in all publicly traded firms in the United States with substantial R&D expenses);
26. Wesley M. Cohen, Richard R. Nelson & John P. Walsh, Protecting Their Intellectual Assets: Appropriability Conditions and Why U. S. Manufacturing Firms Patent (or Not) 4, 14-23 (Nat'l Bureau of Econ. Research, Working Paper No. 7552, 2000), available at http://www.nber.org/papers/w7552 (surveying R&D managers randomly drawn from a sample of all R&D labs in the United States operating as part of a manufacturing firm). Note that neither of these studies addresses the value placed by small firms on patent protection; that is an important limitation, as will become apparent in the ensuing discussion. For a survey of small firms that reaches largely contrary results in selected industries, see infra note 57 and accompanying text.
27. Elsewhere I have reviewed in detail this evidence and other related studies, which show substantial industry-specific and firm-specific variation. See Jonathan M. Barnett, Do Patents Matter? Empirical Evidence on the Incentive Thesis, in HANDBOOK ON LAW, INNOVATION AND GROWTH 178, 182 (Robert E. Litan ed., 2011) ("[T]he incentive effects of patent protection operate along marketspecific, firm-specific and order-specific dimensions.");
28. Jonathan M. Barnett, Private Protection of Patentable Goods, 25 CARDOZO L. REV. 1251, 1257-69 (2004) (reviewing extralegal substitutes for patents).
29. See NATHAN ROSENBERG, INSIDE THE BLACK BOX: TECHNOLOGY AND ECONOMICS 19 (1982).
30. See William J. Baumol, Toward Analysis of Capitalism's Unparalleled Growth: Sources and Mechanism, in ENTREPRENEURSHIP, INNOVATION, AND THE GROWTH MECHANISM OF THE FREE-ENTERPRISE ECONOMIES 158, 164-65 (Eytan Sheshinski et al. eds, 2007) (noting the presence of innovation in both China and the Soviet Union, but rarely the widespread use of these innovations);
31. Maurizio Iacopetta, Dissemination of Technology in Market and Planned Economies, B. E. J. MACROECONOMICS 24-26, Feb. 3, 2004, available at http://www.bepress.com/bejm/contributions/vol4/iss1/art2 (emphasizing how in a planned economy loans are only available for specific production goals, but not for consumption, creating an "output rule" that rewards managers who preserve stable outputs and limit innovation).
32. See OFFICE OF TECH. ASSESSMENT, U. S. CONGRESS, NO. OTA-BP-ITC-165, INNOVATION AND COMMERCIALIZATION OF EMERGING TECHNOLOGIES 49-50(1995), available at http://www.fas.org/ota/reports/9539.pdf (noting that R&D costs are significantly lower than the cost of commercialization activities);
33. FREDERIC M. SCHERER ET AL, PATENTS AND THE CORPORATION: A REPORT ON INDUSTRIAL TECHNOLOGY UNDER CHANGING PUBLIC POLICY 35 (1959) ("A wide variety of case examples is available to show that the costs connected with innovation tend to be much greater than those of the original invention. "); id. at 29-32 (illustrating a specific case example and charting the R&D and commercialization expenditures).
34. For similar views, see, for example, JOHN JEWKES, DAVID SAWERS & RICHARD STILLERMAN, THE SOURCES OF INVENTION 152-53 (2d ed. 1969)
35. and JACOB SCHMOOKLER, INVENTION AND ECONOMIC GROWTH 3 (1966).
36. See Joseph DiMasi, Ronald W. Hansen & Henry G. Grabowski, The Price of Innovation: New Estimates of Drug Development Cost, 22 J. HEALTH ECON. 151, 166 (2003) (estimating, for drugs that underwent the Food and Drug Administration approval process in the 1990s, average costs of $800 million from molecule identification through testing (as calculated on a fully capitalized basis in 2000 dollars)). This figure does not include production, distribution, or marketing costs; hence, total capitalized costs almost certainly exceed $1 billion, as stated above.
37. For the classic sources, see Ronald H. Coase, The Nature of the Firm, 4 ECONOMICA (n. s.) 386, 395 (1937)
38. reprinted in THE NATURE OF THE FIRM: ORIGINS, EVOLUTION, AND DEVELOPMENT 18 (Oliver E. Williamson & Sidney G. Winter eds., 1993)
39. and Oliver E. Williamson, The Logic of Economic Organization, in THE NATURE OF THE FIRM: ORIGINS, EVOLUTION, AND DEVELOPMENT, supra, at 90.
40. See THERESE H. MAYNARD & DANA M. WARREN, BUSINESS PLANNING: FINANCING THE START-UP BUSINESS AND VENTURE CAPITAL FINANCING 400 (2010).
41. See, e.g., Nadel v. Play-By-Play Toys & Novelties, Inc., 208 F.3d 368, 378-81 (2d Cir. 2000) (finding a genuine issue of material fact as to whether a toy idea was "original" in absolute terms, for purposes of summary judgment in a misappropriation claim by an inventor against a toy manufacturer, on the ground that the plaintiff failed to establish the idea's novelty);
42. Sellers v. Am. Broad. Co., 668 F.2d 1207, 1210 (11th Cir. 1982) (rejecting a claim that the defendants misappropriated the plaintiff's idea for a television show on the death of Elvis Presley on the ground that the plaintiff's idea was not novel, unique, or original).
43. For the classic case that addresses this defect in earnout mechanisms, see Bloor v. Falstaff Brewing Corp., 601 F.2d 609 (2d Cir. 1979) (affirming a breach of a contractual best efforts clause where the contract required the buyer to promote and sell beer and, due to a variety of managerial and industry factors, little was sold).
44. In the most well known contribution, James Anton and Dennis Yao have proposed that the idea seller can protect against expropriation by the idea buyer, firm A, by threatening to provide its idea to rival firm B, which will then extract rents that would have been enjoyed by firm A. See James J. Anton & Dennis A. Yao, Expropriation and Inventions: Appropriable Rents in the Absence of Property Rights, 84 AM. ECON. REV. 190, 191-92 (1994). This argument relies on two assumptions. First, it must be the case that the idea seller can credibly commit to either firm A or B that it will not subsequently resell the information to other parties. Second, it must be the case that firms A and B can preserve duopoly rents on products embodying the disclosed technology; if that were not the case, the innovator would have no credible threat against A, who would anticipate that B would pay nothing for an innovation that (given A's knowledge) could not deliver a supracompetitive return.
45. See, e.g., James J. Anton & Dennis A. Yao, The Sale of Ideas: Strategic Disclosure, Property Rights, and Contracting, 69 REV. ECON. STUD. 513, 513-14 (2002) (presenting a model in which an idea seller can extract value from the sale of an idea through a partial disclosure mechanism but noting that the result is dependent on seller wealth, which can be used as a bond to signal the value of the tobe-disclosed portion of the idea).
46. DB Riley, Inc. v. AB Eng'g Corp., 977 F. Supp. 84, 90 (D. Mass. 1997)
47. (quoting J. T. Healy & Son, Inc. v. James A. Murphy & Son, Inc., 260 N. E.2d 723, 731 (Mass. 1970)).
48. MELVIN F. JAGER, TRADE SECRETS LAW § 4.2, at 10.1 (2010).
49. For a more complete discussion of cases that illustrate the uncertainty of trade secrecy protections in precontractual negotiation, see generally Merges, Transactional View, supra note 10.
50. See, e.g., Gerda Gemser & Nachoem M. Wijnberg, Effects of Reputational Sanctions on the Competitive Imitation of Design Innovations, 22 ORG. STUD. 563, 576-77 (2001) (documenting the exchange of technical and style information among designers in the luxury European custom furniture industry);
51. Eric von Hippel, Cooperation Between Rivals: Informal Know-How Trading, 16 RES. POL'Y 291, 292-97 (1987) (documenting the reciprocal exchange of know-how among engineers in the steel minimill industry). Elsewhere I have discussed a large number of other examples.
52. See Jonathan M. Barnett, The Illusion of the Commons, 25 BERKELEY TECH. L. J. 1751, 1793-1813 (2010).
53. This opacity often seems to be the case. See D. J. Teece, Technology Transfer by Multinational Firms: The Resource Cost of Transferring Technological Know-How, 87 ECON. J. 242, 247-48 (1977) (studying twenty-six international technology transfer projects and finding that transfer costs vary widely, ranging from 2% to 59% of total project costs, and averaging 19%).
54. Elsewhere I have discussed at length the infirmities of relying on reputation effects to discipline opportunistic behavior. See Jonathan M. Barnett, Certification Drag: The Opinion Puzzle and Other Transactional Curiosities, 33 J. CORP. L. 95, 100-06 (2007). For an optimistic view of the ability of reputation effects to facilitate bargaining over ideas, see Burk & McDonnell, supra note 10, at 602 ("Bargaining in the marketplace is a multi-round game rather than a single shot, and there are likely to be substantial reputational penalties for 'defecting' from the game by misappropriating intellectual property. This is particularly true in concentrated industries, where opportunities for partnerships are limited, and today's competitor may be tomorrow's essential resource.").
55. I am referring to a standard model of indefinite or infinite repeat play in which a single defection results in irrevocable ejection from the game and the loss of all future cooperative gains. For example, in the most well-known formulation, the "Tit for Tat" game, a successful player elects cooperate in the initial round of an iterated sequence and each round thereafter, but then reverts irrevocably to defect if the other player ever elects defect. For further discussion of this and other iterations, see JEAN TIROLE, THE THEORY OF INDUSTRIAL ORGANIZATION 245-47, 258-59 (3d ed. 1989).
56. The second assumption is consistent with empirical evidence showing that the extent to which patent grants facilitate consummation of licensing transactions is strongest in environments in which reputation effects are weakest (or the technology lifecycle is long) and vice versa. See Joshua S. Gans, David H. Hsu & Scott Stern, The Impact of Uncertain Intellectual Property Rights on the Market for Ideas: Evidence from Patent Grant Delays, 54 MGMT. SCI. 982, 994-96 (2008).
57. Arrow noted this possibility, stating that property rights in information may be held through patents or "in the intangible assets of the firm if the information is retained by the firm and used only to increase its profits." Arrow, supra note 11, at 617. Later commentators have made similar observations. See Bar-Gill & Parchomovsky, supra note 10, at 1664 (noting that a "research unit" can protect against expropriation by a "customer" through vertical integration); Richard Zeckhauser, The Challenge of Contracting for Technological Information, 93 PROC. NAT'L ACAD. SCI. USA 12743, 12744 n. e (1996) (noting integration as an alternative means by which to protect against knowledge leakage). It might be objected that integration is an imperfect defense against expropriation risk insofar as entrepreneurs are still exposed to expropriation by employees who can depart for rivals or set up competing operations. See Burk & McDonnell, supra note 10, at 591-92. That is certainly an important contingency, although a firm can use a variety of means, including confidentiality agreements, reputation effects in the labor market, internal organizational practices, deferred compensation and equity-based incentive schemes, acculturation methods, and threats of dismissal by which to constrain employees from expropriating information. Critically, a firm can condition employment on entry by the prospective employee into an invention assignment contract whereby the firm obtains prospective ownership rights in the idea stock generated by the employee during his or her tenure at the firm. That contractual obligation is bolstered by the duty of loyalty owed by employees to employers as a matter of state common law, which prohibits employees from competing with the principal, assisting the principal's competitors, or using the principal's confidential information for the employee's own purposes.
58. See RESTATEMENT (THIRD) OF AGENCY §§ 8.04-.05 (2006). I am grateful to Jennifer Arlen for bringing this point to my attention. These instruments are not perfect but, in the aggregate, would seem to offer a more potent set of tools by which to control expropriation risk relative to arm's-length interactions with unrelated third parties.
59. For similar views, see OLIVER E. WILLIAMSON, MARKETS AND HIERARCHIES: ANALYSIS AND ANTITRUST IMPLICATIONS 10 (1975) (arguing that opportunism risk is reduced within firms, relative to markets, due to superior "incentive and control" mechanisms and more limited opportunities to capture rents); Arora & Merges, supra note 5, at 452 ("[G]reater control over disclosure of internal information is a well-recognized feature of the employment relationship, as compared with independent contractor status....");
60. Julia Porter Liebeskind, Knowledge, Strategy, and the Theory of the Firm, 17 STRATEGIC MGMT. J. (SPECIAL ISSUE) 93, 102 (1996) ("[F]irms have generalized institutional capabilities that may allow them to protect knowledge from expropriation and imitation more effectively than the limited and costly legal protections tha[t] are available in markets.");
61. and Michael H. Riordan & Oliver E. Williamson, Asset Specificity and Economic Organization, 3 INT'L J. INDUS. ORG. 365, 376 (1985) (noting that forward integration from manufacturing into distribution is a strategy available to firms to limit expropriation of knowledge in human assets). Moreover, employees may have reduced incentives to expropriate an employer's intangible assets if they anticipate facing the same "external" expropriation risk in seeking to commercialize those assets independently or, alternatively, if they anticipate competing with their employer (in which case no supracompetitive rents would be available). At a bare minimum, so long as firms can control internal expropriation risk at some lower cost relative to controlling external expropriation risk, then, relative to contract-based outsourcing, integration offers a preferred mechanism by which to accrue innovation returns in the absence of patent protection.
62. Assuming a market in which suppliers exert some bargaining power would introduce two complications. First, it would mean that innovators could not rely on submitted bids as a perfect reflection of supplier cost, which inability in turn could prevent innovators from selecting the costminimizing level of integration. This simplifying assumption does not necessarily impugn the realism of this framework since anecdotal evidence indicates that firms are often well equipped to evaluate supplier costs even in the absence of perfectly competitive supplier markets. See William S. Lovejoy, Conversations with Supply Chain Managers 5 (Ross Sch. of Bus., Working Paper No. 1145, 2010), available at http://ssrn. com/abstract=1621864 (noting that "[m]ost of the interviewees expressed confidence that... they could accurately estimate what the [supplier] costs should be" and describing several methods used). Second, it would mean that uncoordinated pricing by dominant holders of R&D, production, or other inputs into a single consumption bundle might result in double marginalization inefficiencies that compel innovators or other market participations to select Integrate even where Contract would otherwise minimize total innovation plus commercialization costs. In that case, observed supply chain configurations may reflect a trade-off between minimizing product innovation and commercialization costs and minimizing double marginalization inefficiencies. I am grateful to Richard Epstein for emphasizing the latter point. For further discussion, see infra notes 73 and 76.
63. On these three types of economic efficiency, see Joseph F. Brodley, The Economic Goals of Antitrust: Efficiency, Consumer Welfare and Technological Progress, 62 N. Y. U. L. REV. 1020, 1025 (1987) (analyzing the economic problem of antitrust enforcement in terms of three efficiency goals, "production efficiency, innovation efficiency, and allocative efficiency", and concluding that innovation efficiency is the leading goal in terms of social importance);
64. F. M. Scherer, Antitrust, Efficiency, and Progress, 62 N. Y. U. L. REV. 998, 1011 (1987). Note the surprising implication of the allocative efficiency loss mentioned above: the absence of patent protection imposes deadweight losses by preventing efficient transactions with end users that would have taken place under a lower-cost commercialization path.
65. See ADAM SMITH, AN INQUIRY INTO THE NATURE AND CAUSES OF THE WEALTH OF NATIONS 4-8 (Edwin Cannan ed., 1937) (1776); id. at 5 ("The division of labour, however, so far as it can be introduced, occasions, in every art, a proportionable increase of the productive powers of labour.").
66. See George J. Stigler, The Division of Labor Is Limited by the Extent of the Market, 59 J. POL. ECON. 185, 187-93 (1951).
67. George Stigler's thesis builds on ideas set forth in Allyn Young, Increasing Returns and Economic Progress, 38 ECON. J. 527, 529-36 (1928). Smith himself made similar suggestions. See SMITH, supra note 52, at 6-7 (noting the specialization advantages across trades, regions, and countries). For other discussion of the division of labor across technology industries, see ARORA, FOSFURI & GAMBARDELLA, supra note 10, at 6-7 (describing gains from the division of innovative labor);
68. Ashish Arora & Alfonso Gambardella, The Changing Technology of Technological Change: General and Abstract Knowledge and the Division of Innovative Labour, 23 RES. POL'Y 523, 527-31 (1994) (arguing that the increasing reliance on general and abstract information in industrial research and innovation allows for greater specialization among firms).
69. See generally, e.g., Benito Arruñada & Xosé H. Vázquez, When Your Contract Manufacturer Becomes Your Competitor, HARV. BUS. REV., Sept. 2006, at 135 (noting and illustrating by examples the risk of forward integration by suppliers).
70. Examples exist for both cases. As an example of forward integration by suppliers, several Taiwanese manufacturers in the electronics industry have started out as third-party component suppliers and then entered the market independently as manufacturers of branded devices. See DAVID B. YOFFIE & RENEE KIM, HTC CORP. IN 2009, at 1 (Harvard Bus. Sch. Case Study No. 9-709-466, 2009). Apple provides an example of backward integration by a customer. Normally a user of chip technology, Apple, in the case of the iPad device, adopted a backward integration strategy and independently developed a customized semiconductor chip.
71. See Ashlee Vance & Brad Stone, A Little Chip Designed by Apple Itself, N. Y. TIMES, Feb. 1, 2010, available at http://www.nytimes.com/2010/02/ 02/technolog y/business-computing/02chip. html. For further discussion of Apple's strategy, see infra note 159 and accompanying text.
72. See Stuart J. H. Graham et al, High Technology Entrepreneurs and the Patent System: Results of the 2008 Berkeley Patent Survey, 24 BERKELEY TECH. L. J. 1255, 1290 fig.1, 1290 & n. 1 10, 1290-94 (2009). Specifically, the authors report that biotechnology firms rank patents as the most important appropriability device and medical device firms and venture-backed information technology ("IT") hardware firms rank patents as the second most important device after first-mover advantage. Id. at 1290 fig.1, 1290-91. More generally, the authors find that patenting among start-ups and other small entities in these industries is "[widespread but [n]ot [u]biquitous", id. at 1274, although it is more common among venture-backed start-ups and in the biotechnology and medical device industries, and much less common among non-venture-backed start-ups and in the software industry, id. at 1274-78.
73. See generally Paul M. Romer, Growth Based on Increasing Returns Due to Specialization, 77 AM. ECON. REV. (PAPERS & PROC.) 56 (1987) (using a theoretical model to demonstrate that a larger market will exhibit increased product variety because firms can spread the costs of specialized inputs over a larger volume of product units).
74. This is not to say that all large firms are worse off under "excessive" levels of integration. But the reason why large firms may (sometimes) prefer excessive integration is immaterial from the public's point of view. An incumbent may prefer weak patent protection that increases entry barriers for specialist providers that threaten the incumbent's primary market, even if weak protection increases the incumbent's costs. This behavior is equivalent to adoption of a bundling strategy-in which innovation functions are bundled with all other functions in the supply chain-by a generalist firm for the purpose of deterring entry by specialist innovators who cannot bear the costs of that strategy. See Jay Pil Choi & Christodoulos Stefanadis, Bundling, Entry Deterrence, and Specialist Innovators, 79 J. BUS. 2575, 2582-87 (2006). That strategy in turn applies a more general predation rationale: an incumbent will rationally adopt strategies that inflate its costs if doing so makes entry more difficult by raising rivals' costs, thereby reducing short-term profits (in this case, by forfeiting specialization gains through relationships with outside suppliers) but maximizing long-term profits by extending the incumbent's tenure.
75. This qualification addresses the obvious objection that raising the minimum cost of entry makes no difference in deterring efficient entry so long as outside capital markets will fund any positive NPV project. For arguments to this effect, see Robert H. Bork, Vertical Integration and Competitive Processes, in PUBLIC POLICY TOWARD MERGERS 139 (J. Fred Weston & Sam Peltzman eds., 1969). There are (at least) three uncontroversial reasons to believe external capital markets for R&D are substantially imperfect, in which case the qualifying assumption is never satisfied. First, discussions with potential investors and lenders that necessitate disclosure of technological information restore expropriation risk to some extent. This risk may be mitigated to the extent that a financing entity lacks operational expertise or legal capacity to commercialize the underlying innovation or is subject to reputational pressures that discourage expropriation. Second, in the absence of a secure property right, all lending is unsecured, which substantially inflates the cost of capital. Third, this argument requires perfect information on the part of lenders and complete contracts on the part of lenders and borrowers. Otherwise, adverse selection will require that lenders or investors discount all claims by entrepreneurs as to technological quality so as to reflect uncertainty over the entrepreneur's claims. As it turns out, start-ups appear to use patents to alleviate this problem by signaling underlying value to venture capitalist investors.
76. For further discussion, see J. HIRSHLEIFER, INVESTMENT, INTEREST, AND CAPITAL 200-01 (1970);
77. Oliver E. Williamson, The Vertical Integration of Production: Market Failure Considerations, 6 AM. ECON. REV. (PAPERS & PROC.) 112, 119-20 (1971) (discussing the vertical integration incentives that result from incomplete contracting and risks of opportunism in complex technological markets with unavoidable uncertainty). For empirical evidence on the signaling value of patents to start-ups in seeking external funding, see Graham et al., supra note 57, at 1304-07.
78. The concept of entry barrier is much debated and sometimes borders on the metaphysical. The formulation used in the text above adopts the narrowest "Chicago school" definition of entry barrier as any differentially higher cost of new firms, as compared to incumbents in the relevant market. GEORGE J. STIGLER, THE ORGANIZATION OF INDUSTRY 67 (1968) ("A barrier to entry may be defined as a cost of producing... which must be borne by a firm which seeks to enter an industry but is not borne by firms already in the industry."). Consistent with that definition, it is not absolute commercialization costs per se that constitute a barrier to entry; rather, it is the differential cost of capital borne by nonintegrated relative to integrated firms that explains why absolute cost increases can have socially relevant entry-deterrent effects.
79. For the leading expression of this thesis, see Michael A. Heller, The Tragedy of the Anticommons: Property in the Transition from Marx to Markets, 111 HARV. L. REV. 621, 673-88(1998), which states that excessively fragmented property rights can generate net social losses by impeding, rather than facilitating, innovation (or, in a broader real property context, other) investments;
80. and Michael A. Heller & Rebecca S. Eisenberg, Can Patents Deter Innovation? The Anticommons in Biomedical Research, 280 SCIENCE 698(1998), available at http://www.sciencemag.org/content/280/5364/698.full.pdf, which advances the same thesis with respect to gene patents, and concludes that "[p]olicy-makers should seek to ensure coherent boundaries of upstream patents and to minimize restrictive licensing practices that interfere with downstream product development."
81. This result is a subset of a more general set of circumstances in which firms' ability to migrate to alternative appropriability instruments (including but not limited to organizational form) renders ambiguous any effect of changes in intellectual property protection on incentives and access. For further discussion, see Jonathan M. Barnett, Is Intellectual Property Trivial?, 157 U. PA. L. REV. 1691, 1692-97 (2009).
82. Small firms outperform larger firms in R&D performance on several measures. First, small firms obtain, on average, more highly cited patents and more patents per employee. See CHI RESEARCH, INC., SMALL SERIAL INNOVATORS: THE SMALL FIRM CONTRIBUTION TO TECHNICAL CHANGE 3(2003), available at http://archive.sba.gov/advo/research/rs225tot.pdf Second, small firms produce more innovations per employee.
83. Zoltan J. Acs & David B. Audretsch, Innovation in Large and Small Firms: An Empirical Analysis, 78 AM. ECON. REV. 678, 680-82 (1988). Third, small firms otherwise exhibit higher measures of innovative output relative to R&D dollars.
84. See F. M. SCHERER & DAVID ROSS, INDUSTRIAL MARKET STRUCTURE AND ECONOMIC PERFORMANCE 654-56 (1990).
85. For extensive discussion, see 1 CARLISS Y. BALDWIN & KIM B. CLARK, DESIGN RULES: THE POWER OF MODULARITY (2000) (describing the growing role of modular design in the evolution of integrated systems industry structure).
86. The "rationality" qualifier substantially narrows the set of circumstances under which a patent forecloses entry into a technology market. Even in mature markets in which a patent position controls a dominant design or basic process or product technology, the patent holder may have a rational incentive to license the patent widely. First, the holder may recoup some licensing revenue. Second, it may protect its position in a market for platform goods that derive value from third-party suppliers of complementary goods and services. As I have shown elsewhere with respect to markets for operating systems and other platform technologies, even the most dominant firms often give away access at a zero or even negative fee. See Jonathan M. Barnett, The Host's Dilemma: Strategic Forfeiture in Platform Markets for Informational Goods, 124 HARV. L. REV. (forthcoming 2011) (manuscript at 16-17) (on file with author).
87. David Adelman emphasizes this point with respect to the biotechnology market, which he notes is rich in opportunities and therefore not easily susceptible to preclusion by patented positions. See David E. Adelman, A Fallacy of the Commons in Biotech Patent Policy, 20 BERKELEY TECH. L. J. 985, 1018-23 (2005).
88. This result is precisely what occurred in the oil and gas industry, in which specialized engineering firms ("SEFs") have historically licensed patented process technologies to downstream manufacturers (principally,
89. Nathan Rosenberg, Chemical Engineering as a General Purpose Technology, in GENERAL PURPOSE TECHNOLOGIES AND ECONOMIC GROWTH 167, 188 (Elhanan Helpman ed, 1998) (describing SEFs as an "important niche" market that contributed to rapidly diffusing the new technologies of the oil industry).
90. See DANIEL F. SPULBER, MARKET MICROSTRUCTURE: INTERMEDIARIES AND THE THEORY OF THE FIRM, at xiii-xvi (1999) (introducing the intermediation theory of the firm).
91. Under the rubric of transaction costs, we may include the costs entailed in reaching agreement among multiple suppliers of complementary inputs so as to avoid a collectively harmful "royalty stacking" outcome (equivalent to the more general problem of double marginalization), in which each input holder demands excessive licensing fees, resulting in input costs that prevent or impede commercialization by depleting total profits. Market intermediaries and other participants can reduce these costs through several mechanisms that substantially preserve the transactional flexibility of vertically disintegrated organizational structures. One mechanism is the use of patent pools and crosslicensing mechanisms that enable blanket licensing of large packages of patented technologies across large pools of patentees and licensees. This is a common phenomenon in the electronics industry. See Barnett, supra note 40, at 1789 & n. 71; Peter C. Grindley & David J. Teece, Managing Intellectual Capital: Licensing and Cross-Licensing in Semiconductors and Electronics, 39 CAL. MGMT. REV. 8, 8-10 (1997). Another mechanism available to intermediaries is to assemble packages of complementary patented inputs that are sold as a single bundle, an emergent tendency in the fabless industry examined later in this Article. See infra Part IV. C. Last, dominant upstream patent holders may unilaterally adopt self-imposed restraints on pricing. There is anecdotal evidence of these restraints in the mobile handset industry.
92. See QUALCOMM, LICENSING/IPR OVERVIEW 20(2006), available at http://www.qualcomm. com/common/documents/financial/QCOMIPR0621.pdf.
93. On royalty stacking, and the paucity of evidence for its existence, see Damien Geradin, Anne Layne-Farrar & A. Jorge Padilla, The Complements Problem Within Standard Setting: Assessing the Evidence on Royalty Stacking, 14 B. U. J. SCI. & TECH. L. 144, 145-50 (2008).
94. For the most well-known theoretical argument identifying the possibility of royalty stacking, see Mark A. Lemley & Carl Shapiro, Patent Holdup and Royalty Stacking, 85 TEX. L. REV. 1991, 1992-95 (2007). Where these mechanisms do not sufficiently address these pricing inefficiencies, then market participants would have an incentive to revert to more integrated organizational structures.
95. A prominent stream of economic thought once promoted this view. See JEWKES, SAWERS & STILLERMAN, supra note 21, at 185 (noting, as of 1969, the "modern, and by now widely held, opinion that monopoly encourages, and may even be a condition precedent to, innovation"). For the original source for this "Schumpeterian Hypothesis", see JOSEPH A. SCHUMPETER, CAPITALISM, SOCIALISM, AND DEMOCRACY 131-34 (5th ed. 1962).
96. For reviews of the literature, see ZOLTAN J. ACS & DAVID B. AUDRETSCH, INNOVATION AND SMALL FIRMS 38-45 (1990) (reviewing literature and finding it mostly ambiguous, but claiming some support for Schumpeter's hypothesis with regard to R&D activity, which increases more than proportionally to firm size, while patenting activity increases less than proportionally);
97. MORTON I. KAMIEN & NANCY L. SCHWARTZ, MARKET STRUCTURE AND INNOVATION 75-104 (1982) (reviewing literature and empirical studies and finding some evidence of linear correlation between an increase in firm size and either inventive or R&D activity, subject to decreasing returns after a threshold point).
98. Note that this Article's framework does not address (at least) two social costs of patent protection that would be reflected in a net welfare analysis: (1) transaction costs that impede subsequent innovation (provided that any subsequent innovation still would have taken place under weaker levels of intellectual property) and (2) deadweight losses incurred by consumers as a result of supracompetitive pricing. On transaction costs, it must be noted that decreases in patent protection generate another set of transaction costs captured by the disclosure paradox, which, as discussed at length, can also reduce negative effects on innovation incentives. For further discussion, see infra Part IV. D and, in particular, note 160. Deadweight losses are unlikely to change any normative inference in favor of the property rights solution for two reasons. First, where reduced patent protection forces firms to select more costly integrated structures, those costs must be reflected in higher prices, constraining output relative to an environment in which firms could select less costly contract-based structures. Second, even assuming the standard positive correlation between patent strength and deadweight losses, consumers may still be better off: if it is true, as economic commentators widely agree, that dynamic efficiency gains in technological advance are likely to far outweigh any static efficiency losses in the form of constrained output, then (setting aside distributive concerns) consumers should collectively prefer incurring supracompetitive pricing over the short term in order to enjoy an accelerated rate of technological advance over the long term. See, e.g., Phillip Areeda, Antitrust Law as Industrial Policy: Should Judges and Juries Make It?, in ANTITRUST, INNOVATION, AND COMPETITIVENESS 29, 31 (Thomas M. Jorde & David J. Teece eds., 1992) (citation omitted) (noting the widespread view among economists that "innovation has been thought to contribute far more to our well-being than keeping prices closer to costs through competition").
99. For the classic account, see ALFRED D. CHANDLER, JR., THE VISIBLE HAND: THE MANAGERIAL REVOLUTION IN AMERICAN BUSINESS 1-11 (1977).
100. For relevant discussion in economic and management literatures, see, for example, Naomi R. Lamoreaux, Daniel M. G. Raff & Peter Temin, Beyond Markets and Hierarchies: Toward a New Synthesis of American Business History, 108 AM. HIST. REV. 404 (2003) (describing, through historical surveys, progression toward vertical hierarchies until the 1980s, and then away from these hierarchies in the computer era, toward coordination by long-term relationships)
101. and Richard N. Langlois, The Vanishing Hand: The Changing Dynamics of Industrial Capitalism, 12 INDUS. & CORP. CHANGE 351, 376 (2003) (describing the disintegration trend of the knowledge economy, in which firms specialize and have "soft" assets and "technical standards... permit external mechanisms of coordination and reduce the need for rich information transfer" (footnote omitted)).
102. In legal literature, see generally Ronald Gilson, Charles F. Sabel & Robert E. Scott, Contracting for Innovation: Vertical Disintegration and Interfirm Collaboration, 109 COLUM. L. REV. 431, 432 (2009) (describing how vertical disintegration and changes in the boundaries of firms have given rise to a new form of implicit and explicit "contracting" in which the parties deter opportunism by acquiring "transaction-specific investments in knowledge about their collaborators' capacities");
103. Erica Gorga & Michael Halberstam, Knowledge Inputs, Legal Institutions and Firm Structure: Towards a Knowledge-Based Theory of the Firm, 101 NW. U. L. REV. 1123 (2007) (discussing how knowledge-management requirements and relevant legal regimes affect firm structure);
104. Merges, Transactional View, supra note 10, at 1487-88 (framing the role of property rights in a transaction-intensive economy as providing for precontractual liability and enforcement flexibility); and Merges, Input Markets, supra note 10 (discussing the theoretical framework for the role of intellectual property rights in increased licensing activity and, ultimately, industry structure and firm specialization).
105. See Feng Gu & Baruch Lev, Markets in Intangibles: Patent Licensing 3-7 (N. Y. Univ. Stern Sch. of Bus., Working Paper No. 2451/27465, 2001), available at http://ssrn. com/abstract=1280686 (noting that growing offline and online markets trading in intangibles creates complex accounting issues relating to valuation and disclosure);
106. Merges, Input Markets, supra note 10, at 3. Estimates of the value of the market for patent licensing, patent sale, and trading of patent technology vary substantially. See, e.g., ARORA, FOSFURI & GAMBARDELLA, supra note 10, at 40 (estimating a value of $35-$50 billion per year);
107. Ashby H. B. Monk, The Emerging Market for Intellectual Property: Drivers, Restrainers and Implications, 9 J. ECON. GEOGRAPHY 469, 472 (2009) (citing estimates of $500 million in 2006 in the United States but noting that precise estimates are not available); Gu & Lev, supra, at 4 (noting data showing that revenues from patent licensing rose from $15 billion in 1990 to more than $110 billion in 1999).
108. See Naomi R. Lamoreaux & Kenneth L. Sokoloff, Inventors, Firms, and the Market for Technology in the Late Nineteenth and Early Twentieth Centuries, in LEARNING BY DOING IN MARKETS, FIRMS, AND COUNTRIES 19 (Naomi R. Lamoreaux et al. eds., 1999) [hereinafter Lamoreaux & Sokoloff, Inventors] (describing how U. S. intellectual property rights allowed inventors to assign their inventions to intermediaries in various arrangements and facilitated the diffusion of technical knowledge and the development of markets for technology);
109. Naomi R. Lamoreaux & Kenneth L. Sokoloff, The Market for Technology and the Organization of Invention in U. S. History, in ENTREPRENEURSHIP, INNOVATION, AND THE GROWTH MECHANISM OF THE FREE-ENTERPRISE ECONOMIES, supra note 20, at 213, 223-28
110. [hereinafter Lamoreaux & Sokoloff, The Market for Technology] (describing the emergence of patent intermediaries in the late nineteenth century).
111. See Lamoreaux & Sokoloff, Inventors, supra note 85, at 29-30.
112. See Scherer et al., supra note 21, at 75 (citing SUBCOMM. ON PATENTS, TRADEMARKS, & COPYRIGHTS, S. COMM. ON THE JUDICIARY, 85TH CONG., REP. ON PATENTS, TRADEMARKS AND COPYRIGHTS 14 (Comm. Print 1957)) (noting that, from 1941-1957, more than one hundred court decrees were issued ordering compulsory licensing to all applicants).
113. See id. at 130-35 (showing the decline in U. S. patent applications from 1930 through 1955 in absolute values, showing a sharper decline per capita since 1914, and suggesting that the decline may be due to increased judicial invalidation rates); id. at 137-46 (finding a disproportionate decline in patenting during the period from 1939-1956 among large corporations subjected to compulsory licensing remedies, especially if those remedies required licensing of future patents); Schmookler, supra note 21, at 30-31 (observing that, starting in the 1930s, corporations reduced patenting rates, which fact may be attributed to the increased rates at which courts invalidated patents, the increased exposure to antitrust liability for alleged patent misuse, and a general political animus toward patents at the time); Lamoreaux & Sokoloff, The Market for Technology, supra note 85, at 236 (noting how the rise of the corporate R&D department may have contributed to the decline in patenting rates and the decline of the individual inventor).
114. See Scherer et al., supra note 21, at 72 (quoting President Roosevelt as saying that "future patents might be made available for use by anyone upon payment of appropriate royalties" (quoting GEORGE E. FOLK, PATENTS AND INDUSTRIAL PROGRESS 260 (1942))). President Roosevelt's proposal was never enacted but was adopted in the final report of the Temporary National Economic Committee.
115. See TEMPORARY NAT'L ECON. COMM., FINAL REPORT AND RECOMMENDATIONS, S. DOC. NO. 77-35, at 36-37 (1941).
116. See Richard S. Rosenbloom & William J. Spencer, Introduction: Technology's Vanishing Wellspring, in ENGINES OF INNOVATION: U. S. INDUSTRIAL RESEARCH AT THE END OF AN ERA 1, 4 (Richard S. Rosenbloom & William J. Spencer eds., 1996).
117. During the period 1983-2002, the number of patents issued tripled, representing an annual rate of increase of about 5.7 percent per year, which compares to an annual rate of increase of 1 percent per year from 1930 until 1982 (the year in which the Federal Circuit was established). See ADAM B. JAFFE & JOSH LERNER, INNOVATION AND ITS DISCONTENTS: HOW OUR BROKEN PATENT SYSTEM IS ENDANGERING INNOVATION AND PROGRESS, AND WHAT TO DO ABOUT IT 11-12 (2004).
118. See Rosemarie H. Ziedonis, On the Apparent Failure of Patents: A Response to Bessen and Meurer, 22 ACAD. MGMT. PERSP. 21, 25 & fig.1 (2008)
119. (citing Business and Industrial R&D, NAT'L SCI. FOUND., http://www.nsf.gov/statistics/industry/ (last visited Apr. 25, 2011)) (compiling data from surveys of industrial R&D available on the National Science Foundation Web site).
120. See Diamond v. Chakrabarty, 447 U. S. 303, 309-10 (1980) (affirming that microorganisms are patentable subject matter).
121. This description is a simplification. Some biotechnology start-ups have integrated forward to some extent, and all large pharmaceutical firms maintain some upstream R&D capacities in biotechnology. These structures are consistent with strategically anticipating that bargaining leverage would be lost without being able credibly to threaten independently to undertake R&D (in the case of a downstream incumbent) or production and distribution functions (in the case of an upstream R&D firm). For further discussion of these structures, see ARORA, FOSFURI & GAMBARDELLA, supra note 10, at 63-76 (discussing the growing role of dedicated biotechnology firms and analyzing the nature of the alliances and the uses of IT that have arisen in their aftermath), and Gary P. Pisano, Weijian Shan & David J. Teece, Joint Ventures and Collaboration in the Biotechnology Industry, in INTERNATIONAL COLLABORATIVE VENTURES IN U. S. MANUFACTURING 183 (David Mowery ed., 1988) (analyzing empirical evidence on collaborative relationships between established firms and dedicated research firms, and on characteristics, motivating factors, and transactional difficulties peculiar to the market).
122. A pithy description of this turn of events can be found in an account by a founder of Intel. See ANDREW S. GROVE, ONLY THE PARANOID SURVIVE: HOW TO EXPLOIT THE CRISIS POINTS THAT CHALLENGE EVERY COMPANY AND CAREER 39-47 (1996).
123. See Timothy J. Sturgeon, Modular Production Networks: A New American Model of Industrial Organization, 11 INDUS. & CORP. CHANGE 451, 458-59 (2002) (noting the outsourcing trend).
124. For a description of the same phenomenon in a broader range of industries, see Giovanni Dosi et al., The Economics of Systems Integration: Towards an Evolutionary Interpretation, in THE BUSINESS OF SYSTEMS INTEGRATION 95, 109 (Andrea Prencipe et al. eds., 2003) (noting this process in "many industries");
125. Keith Pavitt, Specialization and Systems Integration: Where Manufacture and Services Still Meet, in THE BUSINESS OF SYSTEMS INTEGRATION, supra, at 78, 82-83 (describing how technological breakthroughs in various industries have been related to processes of technological convergence and vertical disintegration).
126. Press Release, IBM, IBM Shatters U. S. Patent Record; Tops Patent List for 18th Consecutive Year (Jan. 10, 2011), available at http://www-03.ibm.com/ press/us/en/pressrelease/33341.wss#release.
127. Deepak Somaya & David J. Teece, Patents, Licensing, and Entrepreneurship: Effectuating Innovation in Multi-invention Contexts, in ENTREPRENEURSHIP, INNOVATION, AND THE GROWTH MECHANISM OF THE FREE-ENTERPRISE ECONOMIES, supra note 20, at 185, 198
128. (citing IBM ANNUAL REPORT 91 (2001);
129. IBM ANNUAL REPORT 7 (2000);
130. IBM ANNUAL REPORT 5 (1994)).
131. See R. Gene Richter, Foreword to DAVE NELSON, PATRICIA E. MOODY & JONATHAN STEGNER, THE PURCHASING MACHINE: HOW THE TOP TEN COMPANIES USE BEST PRACTICES TO MANAGE THEIR SUPPLY CHAINS, at xi (2001) (remarking on IBM's increased reliance on outside suppliers).
132. See QUALCOMM INC., QUALCOMM BUSINESS MODEL: A FORMULA FOR INNOVATION & CHOICE 3-4(2008), available at http://www.qualcomm.com/documents/files/ qualcomm-businessmodel-formula-innovation-choice.pdf.
133. The number of patents is as reported in Qualcomm's annual report for fiscal year 2009. Qualcomm, Inc., Annual Report (Form 10-K), at 13 (Nov. 5, 2009).
134. See Sturgeon, supra note 99, at 459; Timothy J. Sturgeon, Turnkey Production Networks: The Organizational Delinking of Production from Innovation, in NEW PRODUCT DEVELOPMENT AND PRODUCTION NETWORKS: GLOBAL INDUSTRIAL EXPERIENCE 67, 76 (Ulrich Jürgens ed., 2000).
135. This company is sometimes better known by its subsidiary Foxconn International Holdings Ltd. On revenues, see Financial Statements for Hon Hai Precision Industry, BUS. WK., http://investing. businessweek.com/research/ stocks/financials/financials.asp?ticker=2317:TT (last visited Apr. 25, 2011) (citing revenues of nearly 2 trillion new Taiwanese dollars). The revenues figure in the text is calculated using the exchange rate to United States dollars at the end of fiscal year 2009.
136. On the number of employees, see Fortune Global 500 2009: The World's Biggest Companies-Hon Hai Precision Industry, CNNMONEY. COM, http://money.cnn. com/magazines/fortune/global500/2009/snapshots/11204.html (last visited Apr. 25, 2011).
137. See Robert K. Perrons, The Open Kimono: How Intel Balances Trust and Power to Maintain Platform Leadership, 38 RES. POL'Y 1300, 1300-01 (2009) (analyzing and characterizing Intel's supplier relationships as a platform leader).
138. The Agreement on Trade-Related Aspects of Intellectual Property Rights is formally Annex 1C of the Marrakesh Agreement of the World Trade Organization, signed on April 15, 1994, and binds all 144 members of the World Trade Organization. See Agreement on Trade-Related Aspects of Intellectual Property Rights, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 U. N. T. S. 331.
139. See Bharat N. Anand & Tarun Khanna, The Structure of Licensing Contracts, 48 J. INDUS. ECON. 103, 106, 125, 128-30 (2000) (analyzing a sample set of 1612 technology licensing agreements and finding that, in industries with weak intellectual property rights, there was a lower incidence of licensing activity but firms continued to execute technology transfer in the form of joint ventures, crosslicensing, or licensing to known parties);
140. Ashish Arora, Marco Ceccagnoli & Wesley M. Cohen, Trading Knowledge: An Exploration of Patent Protection and Other Determinants of Market Transactions in Technology and R&D, in FINANCING INNOVATION IN THE UNITED STATES, 1870 TO THE PRESENT 365, 376-81 (Naomi Lamoreaux & Kenneth L. Sokoloff eds., 2007) (finding, based on survey of 1478 manufacturing units, that, in industries in which patents are stronger, there is greater licensing of new technological knowledge by smaller firms, or firms that specialize in R&D, but that no such effect is observed in the case of larger firms);
141. Joanne Oxley, Institutional Environment and the Mechanisms of Governance: The Impact of Intellectual Property Protection on the Structure of Inter-Firm Alliances, 38 J. ECON. BEHAV. & ORG. 283 (1999) (finding that firms tend to use joint ventures or similar arrangements in jurisdictions with weak intellectual property rights and arm's-length contractual relationships in jurisdictions with strong intellectual property rights).
142. For further discussion of this point and supporting evidence, see Mikhaelle Schiappacasse, Intellectual Property Rights in China: Technology Transfers and Economic Development, 2 BUFF. INTELL. PROP. L. J. 164, 172-75 (2004) (discussing the role of the relative strength of intellectual property rights in encouraging foreign direct investment, technology transfer, and hence also in bringing about technological advances).
143. See, e.g., Stephen P. Magee, The Appropriability Theory of the Multinational Corporation, 458 ANNALS AM. ACAD. POL. & SOC. SCI. 123, 129 (1981) (stating that innovating firms must expand to solve appropriability problems). On these older views, see Langlois, supra note 83, at 361-62 & n. 16 (recounting the view that integration is necessary to avoid expropriation risk);
144. Gary P. Pisano & Paul Y. Mang, Collaborative Product Development and the Market for Know How: Strategies and Structures in the Biotechnology Industry, in 5 RESEARCH ON TECHNOLOGICAL INNOVATION, MANAGEMENT, AND POLICY 112 (Richard S. Rosenbloom & Robert A. Burgelman eds., 1993).
145. See Naomi R. Lamoreaux & Kenneth L. Sokoloff, Afterword to FINANCING INNOVATION IN THE UNITED STATES, 1870 TO THE PRESENT, supra note 110, at 469, 471 ("Not so long ago... it was common for many in industry and academe to question how useful patents were as a means of encouraging private parties to invest in inventive activity.").
146. See Martin Fransman, Biotechnology: Generation, Diffusion, and Policy, in TECHNOLOGY AND INNOVATION IN THE INTERNATIONAL ECONOMY 41, 66 (Charles Cooper ed., 1994).
147. For prior commentary on this segment by management scholars, see CLAIR BROWN & GREG LINDEN, CHIPS AND CHANGE: HOW CRISIS RESHAPES THE SEMICONDUCTOR INDUSTRY 61-75 (2009) (describing rising design costs);
148. Ludovic Dibiaggio, Design Complexity, Vertical Disintegration and Knowledge Organization in the Semiconductor Industry, 16 INDUS. & CORP. CHANGE 239 (2007) (examining the impact of knowledge integration versus collaborative outsourcing on firm boundaries and organizational capabilities);
149. Greg Linden & Deepak Somaya, System-on-a-Chip Integration in the Semiconductor Industry: Industry Structure and Firm Strategies, 12 INDUS. & CORP. CHANGE 545 (2003) (analyzing the role of design trading in semiconductor industry organizational structures);
150. Jeffrey T. Macher & David C. Mowery, Vertical Specialization and Industry Structure in High Technology Industries, 21 ADVANCES STRATEGIC MGMT. 317, 330-37 (2004) (describing the progression toward fabless firms and vertical specialization);
151. and Jeffrey T. Macher, David C. Mowery & Timothy S. Simcoe, E-Business and the Semiconductor Industry Value Chain: Implications for Vertical Specialization and Integrated Semiconductor Manufacturers (East-West Ctr. Working Papers, Econ. Series, Paper No. 47, 2002), available at http://www.eastwestcenter.org/fileadmin/stored/pdfs/ECONwp047.pdf (describing the progression and implications for policy).
152. Press Release, Gartner, Inc., Gartner Says Worldwide Semiconductor Revenue Increased 31.5 Percent in 2010 to Exceed $300 Billion (Dec. 8, 2010), http://www.gartner.com/it/page.jsp?id= 1487916.
153. See Semiconductors and Related Devices Market Report, HIGHBEAM BUS., http://business.highbeam.com/industry-reports/equipment/semiconductors-related- devices (last visited Apr. 26, 2011).
154. Note that ASIC devices can be divided into two categories: (1) off-the-shelf devices that can be programmed by the user to implement certain functions as desired and (2) customized devices supplied by an integrated circuit manufacturer. See RAKESH KUMAR, FABLESS SEMICONDUCTOR IMPLEMENTATION 67 (2008).
155. See 13 Fabless IC Suppliers Forecast to Top $1.0 Billion in Sales in 2010!, RES. BULL. 1 (Dec. 21, 2009), http://www.icinsights.com/data/articles/ documents/204.pdf [hereinafter Fabless IC Suppliers].
156. See Macher, Mowery & Simcoe, supra note 114, at 3. Distribution was sometimes outsourced with the manufacturer retaining some "captive" distribution capacities. Some distributors also provided basic support services. For an extensive history of the industry, see generally BO LOJEK, HISTORY OF SEMICONDUCTOR ENGINEERING (2007).
157. See DAVID P. ANGEL, RESTRUCTURING FOR INNOVATION: THE REMAKING OF THE U. S. SEMICONDUCTOR INDUSTRY 37-43 & n. 4 (1994) (describing how in postwar years most firms made no attempt to enforce patents and instead cross-licensing became the norm, thus contributing to the openness of the U. S. technological environment);
158. CHRISTOPHE LÉCUYER, MAKING SILICON VALLEY: INNOVATION AND THE GROWTH OF HIGH TECH, 1930-1970, at 253-94 (2006) (recounting the growth and structure of the semiconductor industry in the Silicon Valley in the 1960s and 1970s).
159. Semiconductor Chip Protection Act of 1984, Pub. L. No. 98-620, 98 Stat. 3347
160. (codified as amended at 17 U. S. C. §§ 901-914 (2006)).
161. Adjusted relative to R&D dollars, this rate (that is, the propensity to patent) doubled during 1982-1992. See Rosemarie Ham Ziedonis & Bronwyn H. Hall, The Effects of Strengthening Patents on Firms Engaged in Cumulative Innovation: Insights from the Semiconductor Industry, in 13 ENTREPRENEURIAL INPUTS AND OUTCOMES 133 (2001) (analyzing the patenting behavior of ninety-five U. S. semiconductor firms during 1979-1995).
162. In particular, Texas Instruments is infamous for having broken from the industry norm of underenforcement of patents. See Bronwyn H. Hall, Exploring the Patent Explosion, in ESSAYS IN HONOR OF EDWIN MANSFIELD: THE ECONOMICS OF R&D, INNOVATION, AND TECHNOLOGICAL CHANGE 195, 201-02 (Albert N. Link & F. M. Scherer eds., 2005).
163. See Ziedonis & Hall, supra note 122. Note that figure 6 also depicts registration rates for "mask works" covered under the Semiconductor Chip Protection Act. As is evident, the Act has been underused. This underuse is generally attributed to technological developments that have frustrated third-party imitation that relies solely on reverse engineering layout designs. See Leon Radomsky, Sixteen Years After the Passage of the U. S. Semiconductor Chip Protection Act: Is International Protection Working?, 15 BERKELEY TECH. L. J. 1049, 1077-80 (2000).
164. See Fabless IC Suppliers, supra note 118, at 1.
165. See Rajà Attia, Isabelle Davy & Roland Rizoulières, Innovative Labor and Intellectual Property Market in the Semiconductor Industry, in TECHNOLOGY AND MARKETS FOR KNOWLEDGE: KNOWLEDGE CREATION, DIFFUSION AND EXCHANGE WITHIN A GROWING ECONOMY 137, 145-46 (Bernard Guilhon ed., 2001) (describing the maturation of the fabless model); Ziedonis & Hall, supra note 122, at 136.
166. Two developments were of particular importance. First, in 1979, a technical achievement in semiconductor design methodology, known as very large scale integration ("VLSI"), enabled the assembly of working prototypes of chip design at relatively low cost and without any involvement in the far more costly fabrication process. For further discussion, see BALDWIN & CLARK, supra note 69, at 77-88 (describing Carver Mead and Lynn Conway's breakthrough to dimensionless, scalable design rules and their impact on chip designs); Nathan Rosenberg & W. Edward Steinmueller, The Economic Implications of the VLSI Revolution, in INSIDE THE BLACK BOX: TECHNOLOGY AND ECONOMICS, supra note 19, at 178 (projecting the impact of VLSI as depending on economic factors). Second, in the 1980s, the industry converged on silicon-based complementary metal oxide semiconductors as the dominant design in semiconductor process technology. This convergence facilitated standardization of the interfaces that allow design modules to be designed independently by multiple providers. See Linden & Somaya, supra note 114, at 555.
167. As of 2007, a new plant required a $3.5 billion investment, as illustrated by Intel's newest plant in Israel. See ILKKA TUOMI, JOINT RESEARCH CTR. SCIENTIFIC & TECHNOLOGICAL REPORTS, THE FUTURE OF SEMICONDUCTOR INTELLECTUAL PROPERTY ARCHITECTURAL BLOCKS IN EUROPE 74 n. 109 (Marc Bogdanowicz ed., 2009), available at http://publications.jrc.ec.europa.eu/repository/ bitstream/111111111/6946/1/jrc52422.pdf. These costs are compounded by the fact that plant technology is usually obsolete within a few years due to technological advances.
168. See G. Dan Hutcheson, Economics of Semiconductor Manufacturing, in HANDBOOK OF SEMICONDUCTOR MANUFACTURING TECHNOLOGY 35-1, 35-14 (Robert Doering & Yoshio Nishi eds., 2d ed. 2008).
169. In total, design costs of an ASIC device have been estimated as high as $45 million. See KUMAR, supra note 117, at 235. Other sources give estimates of up to $80 million for a highly customized design. See Dieter Ernst, Internationalisation of Innovation: Why Is Chip Design Moving to Asia? 8 n. 21 (East-West Ctr. Working Papers, Econ. Series, Paper No. 64, 2004), available at http://www.eastwestcenter.org/fileadmin/stored/pdfs/ECONwp064.pdf. Higher estimates rise to $100 million for the latest-generation chip designs. See TUOMI, supra note 129, at 75 fig.23.
170. See Ziedonis & Hall, supra note 122, at 137, 159 (finding that firms that entered the semiconductor industry after 1982 patent more intensively than pre-1982 entrants, where 1982 is used as a "marker" for strengthened patents based on creation of Federal Circuit, and, in particular, finding that small firms are five times more likely to patent than all other firms in the sample, which excludes, however, some of the largest diversified semiconductor manufacturers); Adam B. Jaffe, The U. S. Patent System in Transition: Policy Innovation and the Innovation Process, 29 RES. POL'Y 531, 540 (2000) (stating that semiconductor patents held by small design firms are disproportionately the subject of patent litigation);
171. Rosemarie H. Ziedonis, Don't Fence Me In: Fragmented Markets for Technology and the Patent Acquisition Strategies of Firms, 50 MGMT. SCI. 804, 817-18 (2004) (finding that large vertically integrated semiconductor firms tend to cross-license patents while small design firms tend to adopt more litigious and exclusionary strategies).
172. See DAVID J. TEECE, MANAGING INTELLECTUAL CAPITAL: ORGANIZATIONAL, STRATEGIC, AND POLICY DIMENSIONS, app. A at 193-94, 203-07 (2000).
173. Some of the leading integrated firms are members in the SEMATECH research consortium, which pools member resources to develop manufacturing technologies. See SEMATECH, 2005 ANNUAL REPORT: ACCELERATING THE NEXT TECHNOLOGY REVOLUTION 3-5(2006), available at http://www.sematech.org/corporate/annual/annual 05.pdf.
174. The precise historical sequence is somewhat more complex. The leading foundry, Taiwan Semiconductor Manufacturing Company, Ltd. ("TSMC"), was founded as a combination of existing Taiwanese government initiatives to promote local technology industries and demands by integrated circuit design start-up firms, some of which had codeveloped technology with a Taiwanese government R&D institute, which required a local specialized chip foundry in order to avoid sending production overseas to Korean firms. In 1988 TSMC was founded through an investment and transfer of intellectual property by Philips, as well as investments by local investors and the Taiwanese government. See ALICE H. AMSDEN & WAN-WEN CHU, BEYOND LATE DEVELOPMENT: TAIWAN'S UPGRADING POLICIES 166-67 (2003) (discussing the Taiwanese government's role in the establishment of foundries, the integrated circuit mask industry, and the training of human resources in integrated circuit design);
175. WILLY SHIH & JYUN-CHENG WANG, UPGRADING THE ECONOMY: INDUSTRIAL POLICY AND TAIWAN'S SEMICONDUCTOR INDUSTRY 8 (Harvard Bus. Sch. Case Study No. 9-609-089, 2010).
176. Industry Data: Semiconductor & Fabless Facts, GLOBAL SEMICONDUCTOR ALLIANCE, http://www.gsaglobal.org/resources/industrydata/facts.asp (last visited Apr. 27, 2011) [hereinafter Industry Data]. Based on forecasted 2010 results, thirteen fabless firms are expected to have in excess of $1 billion in sales in 2010 and together account for about 70 percent of total fabless company sales.
177. Fabless IC Suppliers, supra note 118, at 1. Those firms are Qualcomm, Broadcom, Advanced Micro Devices Inc. ("AMD"), MediaTek, Marvell Semiconductor, Nvidia, Xilinx, Altera, LSI Corp., Avago, Novatek, ST-Ericsson, and MStar. Id. at 2 fig.1.
178. See Jeffery T. Macher, David C. Mowery & Alberto Di Minin, The "Non-Globalization" of Innovation in the Semiconductor Industry, 50 CAL. MGMT. REV. 217, 223 (2007).
179. On the geographic distribution of semiconductor design firms and foundries, see Macher & Mowery, supra note 114, at 334-35. Some readers have expressed surprise that design houses would transfer technology to Asian jurisdictions in which patent protections are generally thought to be insecure. Two observations largely moot this concern. First, even if patent protections are insecure, expropriation opportunities are limited by the fact that the target markets for the ultimate consumption goods would bar entry of any products made using unlicensed patented components. Second, contrary to common belief, Asian jurisdictions do not uniformly have insecure patent rights. Taiwan, the chief location of the largest foundries, explicitly adopted a policy of strongly enforced patents in 1986, consisting principally of increased infringement awards and creation of a specialized court to hear patent disputes. That change almost precisely coincides with the rise of the foundry industry and provides highly suggestive evidence consistent with this Article's core thesis: strong intellectual property rights, both as a formal and effective matter, enabled Taiwanese foundries to commit credibly against expropriation, thus also enabling mutually efficient technology-transfer transactions with Western (mostly U. S.-based) design houses. This reform process is extensively detailed by Shih-tse Lo, Strengthening Intellectual Property Rights: Experience from the 1986 Taiwanese Patent Reforms (Concordia Univ., Dep't of Econ., Working Paper No. 04004, 2004), available at http://economics.concordia.ca/documents/ working-papers/04004sl.pdf. Shih-tse Lo describes the reforms and documents the positive effects both on domestic innovation by R&D-intensive Taiwanese firms (as measured by R&D investment and patenting in the United States) and on foreign direct investment into Taiwan. These legal reforms have translated into concrete enforcement effects, as reflected by dramatically lower software piracy rates in Taiwan (43 percent) relative to China (86 percent).
180. See Chun-Hsien Chen, Explaining Different Enforcement Rates of Intellectual Property Protection in the United States, Taiwan, and People's Republic of China, 10 TUL. J. TECH. & INTELL. PROP. 211, 213 (2007). As further confirmation, it appears that weaker, but far from nonexistent, protections of intellectual property in China are impeding further outsourcing of wafer fabrication to China, despite lower costs relative to Taiwan and other East Asian jurisdictions.
181. See SHIRI SHNEORSON, TAIWAN SEMICONDUCTOR MANUFACTURING COMPANY: THE SEMICONDUCTOR SERVICES COMPANY 18 (Stanford Graduate Sch. of Bus., Case Study No. GS-40, 2006), available at http://www.gsb.stanford.edu/scforum/login/pdfs/ Taiwan-Semiconductor-2006.pdf ("[F]oreign fabless companies were worried about the country's lack of respect to intellectual property.").
182. See TSMC Holds More Than 50% Share of Top Ten Semiconductor Foundries, EE HERALD (Jan. 31, 2010), http://www.eeherald.com/section/news/nw100010277.html (summarizing IC Insights' 2010 report, and listing figures for 2009). The leading firms and market share are TSMC (based in Taiwan; 47% share); UMC (based in Taiwan; 14.7% share); Chartered (based in Singapore; 8% share); Globalfoundries (based in the United States; 5.8% share); and SMIC (based in China, 5.6% share). Id. (calculating market shares based on revenue figures). In the fourth quarter of 2009, Globalfoundries acquired Chartered.
183. See Foundries Play Semiconductor Survivor in 2010, ISUPPLI (Sept. 21, 2009), http://www.isuppli.com/Semiconductor-Value-Chain/MarketWatch/Pages/ Foundries-Play-Semiconductor-Survivor-in-2010.aspx.
184. See JEORGE S. HURTARTE, EVERT A. WOLSHEIMER & LISA M. TAFOYA, UNDERSTANDING FABLESS IC TECHNOLOGY, at xvii-xviii (2007) (noting that most major integrated manufacturers today have adopted outsourcing to some extent).
185. BANK OF AM. MERRILL LYNCH, TECHNOLOGY: FINDING VALUE IN SMARTPHONES 2-4(2009), available at http://thebln. com/wp-content/uploads/2009/12/ Smartphone-market-2010.pdf.
186. ORG. FOR ECON. CO-OPERATION & DEV., OECD INFORMATION TECHNOLOGY OUTLOOK 2008, at 161, 162 tbls.3 & 4(2008), available at http://www.oecd-ilibrary.org/docserver/download/full text/9308041e.pdf?expires= 1303976079&id=id&accname=ocid177219&checksum=75D22929C1F0A9 419ECEF6D390E1DBBA.
187. See Industry Data, supra note 137.
188. For further information, see Manufacturing Alliance Partners, COMMON PLATFORM, http://www.commonplatform.com/about/manufacturing-partners.asp (last visited Apr. 28, 2011).
189. See Mark LaPedus, AMD Foundry Spinoff Open for Business, EE TIMES (Mar. 4, 2009, 12:01 AM), http://www.eetimes.com/electronics-news/4081523/AMD-foundry- spinoff-open-forbusiness.
190. Almost concurrently with the emergence of the SoC market, software companies, fabless chip companies, and other entities established the Virtual Socket Interface ("VSI") Alliance (disbanded in 2008) in order to establish standardized interfaces for the transmission of design modules from fabless firms to foundries and the circulation of design modules among chipless and fabless firms. See Grant Martin, The History of the SoC Revolution: The Rise and Transformation of IP Reuse, in WINNING THE SOC REVOLUTION: EXPERIENCES IN REAL DESIGN 1, 4-6 (Grant Martin & Henry Chang eds, 2003). Other leading standardization initiatives are as follows: (1) OCP International Partnership, which provides an openly licensed socket ("IP core interface") (OCP 2.2), OCP-IP, http://www.ocpip.org/ (last visited May 4, 2011); (2) Accellera (merged with SPIRIT in 2010), which provides design and verification standards, including hardware design language (System Verilog), Accellera, www.accellera.org (last visited May 4, 2011); (3) Open System Initiative, which provides an open industry standard for system-level modeling, design, and verification (SystemC 2.2), and an interface standard enabling interoperability of models at the transaction level (TLM Standard 2.0)
191. OPEN SYSTEMC INITIATIVE (OSCI), www.systemc.org (last visited May 4, 2011);
192. and (4) Silicon Integration Initiative, which provides open interface standards for producing integrated silicon systems, SI2 (SILICON INTEGRATION INITIATIVE), www.systemc.org (last visited May 4, 2011). On obstacles to achieving greater standardization in the semiconductor market, see Macher, Mowery & Simcoe, supra note 114, at 11-16 (describing the development of Web-enabled software to connect fabs and foundries, and describing the founding of the VSI Alliance).
193. See ARORA, FOSFURI & GAMBARDELLA, supra note 10, at 76-77 (describing the development of "chipless" firms); Attia, Davy & Rizoulières, supra note 127, at 146, 165-67 (discussing industry structure and economic implications); Linden & Somaya, supra note 114, at 568-69 (discussing the role of EDA). By licensing design modules from third-party suppliers, foundries can offer clients both production and design input functions (as in the case of the market leader, TSMC). See MARCO IANSITI & ROY LEVIEN, THE KEYSTONE ADVANTAGE: WHAT THE NEW DYNAMICS OF BUSINESS ECOSYSTEMS MEAN FOR STRATEGY, INNOVATION, AND SUSTAINABILITY 132-33 (2004) (discussing the NVIDIA example of using third-party design tools and building blocks, as well as testing standards).
194. ARM HOLDINGS PLC, 20/20 VISION: 20 YEARS OF GROWTH DELIVERING FUTURE OPPORTUNITIES: ANNUAL REPORT AND ACCOUNTS 2010 31 (2010) [hereinafter ARM ANNUAL REPORT], available at http://media.corporate-ir.net/media-files/IROL/19/197211/ 626-1-ARM-AR-040311.pdf (using average dollar exchange rate for 2010 of $1.55 to £1).
195. Market capitalization is based on the company's NASDAQ share price as of May 3, 2011. ARM Holdings plc (ARMH), YAHOO! FINANCE, http://finance.yahoo.com/ q?s=ARMH (last visited May 4, 2011).
196. ARM, ARM HOLDINGS PLC Q1 2011 RESULTS: ROADSHOW SLIDES 10(2011), available at http://phx.corporate-ir.net/External. File?item= UGFyZW50SUQ9OTA4MTJ8Q2hpbGRJRD0tMXxUe XBlPTM=&t=1.
197. See Ernst, supra note 130, at 10 (citing industry sources stating that systematic design reuse can cut chip development costs in half in three years and by more than 70 percent in six years). This observation confirms an intuition advanced by Mark Lemley and Julie Cohen with respect to the software market. See Julie E. Cohen & Mark A. Lemley, Patent Scope and Innovation in the Software Industry, 89 CALIF. L. REV. 1, 5-6 (2001) (arguing that patent rights over software may promote reuse of software by allowing firms to earn returns on programming concepts that would otherwise not be disclosed).
198. See Vance & Stone, supra note 56. To further this endeavor, in 2008 Apple also purchased P. A. Semi, a semiconductor design company, for $278 million. See Erika Brown, Elizabeth Corcoran & Brian Caulfield, Apple Buys Chip Designer, FORBES. COM (Apr. 23, 2008, 12:15 AM), http://www.forbes.com/ 2008/04/23/apple-buys-pasemi-tech-ebiz-cz-eb-0422apple.html.
199. These transaction costs fall into two categories. In both cases, there are cogent theoretical arguments that these patent-related costs could impede innovation, but scant empirical evidence that these costs actually do impede innovation. The first category encompasses administrative and disputeresolution costs associated with negotiating access to intellectual inputs from third parties. Empirical inquiries have had difficulties confirming the predicted adverse effects of patent-related transaction costs on innovative activity. See David E. Adelman & Kathryn L. DeAngelis, Patent Metrics: The Mismeasure of Innovation in the Biotech Patent Debate, 85 TEX. L. REV. 1677, 1680 (2007) (finding "little evidence that the recent growth in biotechnology patenting is threatening innovation", based on a dataset of 52, 000 biotechnology patents from January 1990 through December 2004);
200. John P. Walsh, Ashish Arora & Wesley M. Cohen, Effects of Research Tool Patents and Licensing on Biomedical Innovation, in PATENTS IN THE KNOWLEDGE-BASED ECONOMY 285, 286-89, 292-94 (Wesley M. Cohen & Stephen A. Merrill eds., 2003) (finding, based on a sample of seventy interviews, that patents on inputs to drug discovery generally have not halted research projects due to potentially conflicting patent claims held by other parties, although there is evidence of some delays in negotiating access to research tools or other valuable information or methodologies).
201. But see Fiona Murray & Scott Stern, Do Formal Intellectual Property Rights Hinder the Free Flow of Scientific Knowledge? An Empirical Test of the Anti-Commons Hypothesis, 63 J. ECON. BEHAV. & ORG. 648, 649-51, 672-74 (2007) (hypothesizing that anticommons effects would predict a lower citation rate for papers that contained ideas that were subsequently patented, and finding evidence of a modest effect). Second, this category encompasses "double marginalization" costs that result from uncoordinated pricing by holders of patents over nonsubstitutable components of a single technological bundle. As I note elsewhere, there exists little affirmative evidence supporting this thesis as a practical matter and considerable evidence that markets anticipate this risk and often take measures to address it. See supra note 76. I recognize, however, that these are unsettled empirical questions; my analytical framework, therefore, is fully open to the possibility that collateral social costs may sometimes recommend incomplete levels of patent protection, notwithstanding the fact that doing so inherently constrains innovators' transactional opportunities.
202. For an attempt to address this difficult causality question, see Jeff Thurk, Market Effects of Patent Reform in the U. S. Semiconductor Industry (Oct. 31, 2009) (unpublished manuscript), available at https://webspace.utexas.edu/ jmt597/www/papers/patent-pro-US-semi.pdf (running a policy simulation to estimate the effects of increased patent protection and increased market size (demand shock) on R&D investment and finding the contribution of market size is greater, but also finding there are still significant effects of patent protection on licensing revenue and the number of fabless firms).
203. On Qualcomm's aggressive patent litigation strategy, see Annabelle Gawer & Michael A. Cusamano, How Companies Become Platform Leaders, MIT SLOAN MGMT. REV., Winter 2008, at 28, 31-32.
204. This latter strategy refers to the so-called trolls phenomenon. The extent of the phenomenon remains unclear. For relevant studies, see John R. Allison, Mark A. Lemley & Joshua Walker, Extreme Value or Trolls on Top? The Characteristics of the Most-Litigated Patents, 158 U. PA. L. REV. 1, 32 (2009) (finding that nonpracticing entities own a large segment of the most-litigated patents and file many of the suits, but that the issue of whether there is a "flood of patent trolls" depends on the definition);
205. Gwendolyn G. Ball & Jay P. Kesan, Transaction Costs and Trolls: Strategic Behavior by Individual Inventors, Small Firms and Entrepreneurs in Patent Litigation 14-15 (Ill. Law & Econ. Paper Series, Paper No. LE09-005, 2009), available at http://ssrn. com/abstract=1337166 (finding inconclusively that small plaintiffs sue more often than licensing firms, and that licensing firms have less propensity to go to trial or seek judgment when suing large plaintiffs).
206. As Steven Cheung once pondered, "[E]conomists tend to overlook the crucial question: What would an inventor or innovator do if there were no patent protection?" Stephen N. S. Cheung, Property Rights in Trade Secrets, 20 ECON. INQUIRY 40, 40 (1982). Possible answer: not invent or invent subject to secrecy. See id. 40-41; Adelman, supra note 10, at 458, 463, 466 (observing that the absence of patents necessitates using other entry barriers in order to block imitation).
207. For the latest proposed bills, see Patent Reform Act of 2009, S. 515, 111th Cong. (2009);
208. H. R. 1260, 111th Cong. (2009).
209. For a fuller discussion of these tendencies, see Jonathan M. Barnett, Property as Process: How Innovation Markets Select Innovation Regimes, 119 YALE L. J. 384 (2009).
210. For an example of support for the proposed reforms by a technology trade association that includes larger technology firms, including leading integrated chip manufacturers such as Intel and Micron, see Letter from the Coalition for Patent Fairness to the President of the United States (Mar. 23, 2009), available at http://www.patentfairness.org/pdf/CEO-letter.pdf The Innovation Alliance (which tends to represent smaller technology firms, including fabless firms Qualcomm and LSI Logic) and the National Venture Capital Association (which represents venture capitalists) had previously opposed the reform effort but are now prepared to accept the legislation after considerable modifications. The new amended proposal is now opposed as being too weak by the Coalition for Patent Fairness, which tends to represent larger technology firms.
211. On prior opposition by smaller technology companies, see Anne Broache, Patent Law Overhaul: Bad for Start-Ups?, CNET NEWS (Sept. 20, 2007, 12:19 PM), http://news.cnet.com/Patentlaw-overhaul-Bad-for-start-ups/2100-1028-3-6209223. html.
212. On opposition by venture capitalists, see Anne Broache, Tech Companies, Investors Clash over Patent Law, CNET NEWS (Mar. 29, 2007, 3:19 PM), http://news.cnet.com/Tech-companies%2C-investors-clash-over-patent-law/ 2100-1028-36171866.html.
213. On the changes in position with respect to the modified bill, see Kim Hart, Tech Industry Splinters over Patent Reform Proposal, HILLICON VALLEY (Mar. 9, 2010, 7:10 AM) http://thehill.com/blogs/hillicon-valley/technology/85515- tech-industry-splinters-over-patent-reformproposal.