The danger of underdeveloped patent prospects

Cornell Law Review

Volumn 92, Issue 6, 2007, Pages 1065-1121

  Abramowicz, Michael ^a,b  

^a UNIVERSITY OF CHICAGO LAW SCHOOL   (United States)

^b GEORGE WASHINGTON UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 35548966117     PISSN: 00108847     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Review

References (234)

1. Life Estate: Termination Because of Waste, REAL EST. L. REP., Mar. 1997, at 5, 5 (citing McIntyre v. Scarbrough, 471 S.E.2d 199 (Ga. 1996) as providing a useful illustration of the doctrine);
2. see Thomas W. Merrill & Henry E. Smith, The Property/Contract Interface, 101 COLUM. L. REV. 773, 822-24 (2001) (discussing the evolution of the law of waste in the landlord-tenant context). I am grateful to John Duffy for first pointing me to the analogy between nondevelopment of patents and waste of real property.
3. While some commentators have considered whether intellectual property law may tolerate waste, they have not elaborated on the nature of that waste or recognized that the finite patent term makes the danger of waste particularly severe. See Edwin C. Hettinger, Justifying Intellectual Property, 18 PHIL. & PUB. APT. 31, 42-44 (1989) (noting that the incentive to charge high prices for intellectual property might create waste, but not considering the danger of underdevelopment of the patent right);
4. Justin Hughes, The Philosophy of Intellectual Property, 77 GEO. L.J. 287, 328-29 (1988) (noting that patent law may risk waste by not requiring the owner to commercialize the property, but not indicating why such failure to commercialize is a problem); see also supra note 1 and accompanying text (discussing the concept of waste).
5. 35 U.S.C. § 154(a)(2) (2000).
6. See id. § 103(a).
7. See generally F. Scott Kieff, IP Transactions: On the Theory & Practice of Commercializing Innovation, 42 HOUS. L. REV. 727 (2005) (considering the implications of new institutional economics on the commercialization of innovation).
8. See, e.g., id. at 735, 752.
9. See infra Part I.A.1; Part II.B.1-2.
10. See infra Part II.B.3.
11. See 35 U.S.C. § 101 (allowing patents for "new and useful improvement[s]" of inventions). See generally Mark A. Lemley, The Economics of Improvement in Intellectual Property Law, 75 TEX. L. REV. 989, 1000-13 (1997) (outlining patent law's approach to improvements).
12. But see infra Part II.C (suggesting that the Finiteness of a patent term might suboptimally reduce patentees' incentives to make improvements - even patentable improvements - to their inventions).
13. European law permits compulsory licensing when a patentee has abandoned an invention. See Ronald E. Myrick, Influences Affecting the Licensing of Rights in a Unitary European Market, 4 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 81, 95-96 (1993). Requiring outright abandonment to punish underdevelopment makes sense because such a binary test does not require courts to determine what constitutes "underdevelopment."
14. For an argument that patent rights should not be surrendered even when a patentee fails to practice the underlying invention, see F. Scott Kieff, Coordination, Property, and Intellectual Property: An Unconventional Approach to Anticompetitive Effects and Downstream Access, 56 EMORY L.J. 327, 426-27 (2006).
15. See, e.g., 35 U.S.C. §§ 101-103 (delineating the requirements for patenting inventions in the United States).
16. See Edmund W. Kitch, The Nature and Function of the Patent System, 20 J.L. & ECON. 265, 266 (1977).
17. For a brief statement of the prospect theory and an overview of its criticisms, see A. Samuel Oddi, Un-Unifled Economic Theories of Patents - The Not-Quite-Holy Grail, 71 NOTRE DAME L. REV. 267, 281-82 (1996).
18. Even Kitch assumes that there generally will be sufficient patent term for commercializing an invention. See Kitch, supra note 13, at 284-85.
19. See id. at 272 (listing numerous inventions' patent and commercialization dates).
20. See Yoram Barzel, Optimal Timing of Innovations, 50 REV. ECON. & STAT. 348 (1968).
21. Id. at 348, 354-55 (noting that invention might occur inefficiently late because an inventor cannot capture the full social benefit of an invention).
22. Each entrepreneur hoping to sell or improve the patented product after the patent term will have an incentive to free ride on the patentee's development activities and on any subsidy to the patentee provided by other similarly situated entrepreneurs.
23. See generally William T. Robinson et al., First-Mover Advantages from Pioneering New Markets: A Survey of Empirical Evidence, 9 REV. INDUS. ORG. 1 (1994) (providing an overview of the econometric literature analyzing the extent of first-mover advantages). More recent work has emphasized the possibility that first movers may suffer from some systematic disadvantages.
24. See generally Marvin B. Lieberman & David B. Montgomery, First-Mover (Dis)Advantages: Retrospective and Link with the Resource-Based View, 19 STRATEGIC MGMT. J. 1111 (1998) (summarizing the existing literature and research theories discussing first-mover advantages).
25. For a discussion of the implication of first-mover disadvantages for the patent system, see F. Scott Kieff, Property Rights and Property Rules for Commercializing Inventions, 85 MINN. L. REV. 697, 708-09 (2001), which notes that second movers often benefit from a first mover's success.
26. See supra notes 12-13 and accompanying text.
27. Requiring commercialization before an inventor can obtain a patent could greatly reduce this problem, but such requirements would produce considerable costs. See infra Part I.B.2 (illustrating the benefit of granting patents at a relatively early stage).
28. Additional rules would ensure that a patentee who values the patent far more than third parties but not because of the potential benefits of ownership continuity will not call for an auction. In particular, the auction system would allow a third-party winner at the auction to sell the patent back to the patentee but only at the end of the original patent term. See infra Part III.B.2(a). As a result, a third party who recognizes the special value of the patent extension to the patentee would bid the full value of the patent extension to the patentee (ignoring bargaining costs). Anticipating this, the patentee would not call for an auction because, absent some benefit of ownership continuity, the patentee would be unable to pay a profitable markup above the third party's bid. However, a third party will not factor the benefit of ownership continuity into its bid because the rule preventing sale during the patent term will prevent the third party and the patentee from reaching an agreement guaranteeing such continuity.
29. Because Martin and Partnoy have not yet released the paper in which they develop this theory, the summary given here is based on a presentation they gave at Washington University on November 4, 2005. See Video recording: Presentation by Shaun Martin & Frank Partnoy at Conference on Commercializing Innovation, held by Washington University School of Law, Center for Research on Innovation & Entrepreneurship (Nov. 4, 2005) [hereinafter Martin & Partnoy Presentation] (available at http://law.wustl.edu/CRIE/ videos/CI11.4-5.06/03martinpartnoy.ram).
30. See id.
31. See John F. Duffy, Rethinking the Prospect Theory of Patents, 71 U. CHI. L. REV. 439, 445 (2004).
32. See id.
33. Mark Lemley roughly estimates that only about five percent of issued patents are ever licensed. Mark A. Lemley, Essay, Rational Ignorance at the Patent Office, 95 Nw. U. L. REV. 1495, 1507 (2001).
34. See Clarisa Long, Patent Signah, 69 U. CHI. L. REV. 625, 645-46, 648-49 (2002) (arguing that because the number of patents a firm obtains correlates with less measurable attributes of a firm that are of interest to potential investors, firms may even choose to acquire patents with little economic value).
35. For a critique of the signals hypothesis, see Gideon Parchomovsky & R. Polk Wagner, Patent Portfolios, 154 U. PA. L. REV. 1, 21-22 (2005).
36. Commentators before Martin and Partnoy likened a patent to a stock option in that it may or may not be exercised or developed. See F. Russell Denton & Paul J. Heald, Random Walks, Non-Cooperative Games, and the Complex Mathematics of Patent Pricing, 55 RUTGERS L. REV. 1175, 1194-95 (2003). Commentators have also analogized patents to lottery tickets.
37. See, e.g., Jonathan A. Barney, A Study of Patent Mortality Rates: Using Statistical Survival Analysis to Rate and Value Patent Assets, 30 AIPLA Q.J. 317, 328 n.30 (2002);
38. F.M. Scherer, The Innovation Lottery, in EXPANDING THE BOUNDARIES OF INTELLECTUAL PROPERTY: INNOVATION POLICY FOR THE KNOWLEDGE SOCIETY 3, 4-7, 15 (Rochelle Cooper Dreyfuss et al. eds., 2001) (observing that patent returns may fit a Pareto distribution, which has a longer and thicker tail than a log normal distribution, and thus that owning large patent portfolios may not entirely diversify risk). Such analogies and observations underscore the conception that patent holders have the option to develop patents as they see fit and may thus choose not to develop the patents at all. These commentators have not, however, noted that the option characteristics of patents may lead to their acquisition at such an early time as to inhibit development within the patent term.
39. See Lemley, supra note 27, at 1498-99 (estimating the cost of a patent prosecution to be between $10,000 and $30,000 per patent).
40. See Duffy, supra note 25; Martin & Partnoy Presentation, supra note 23. In a subsequent article (written after Duffy read a draft of this Article), Duffy recognizes the nondevelopment problem. See John F. Duffy, Embryonic Patents: Prospects, Prophesies and Pedis Possessio (2005) (unpublished manuscript, on file with author). Duffy notes that patent law tries to avoid the problem through the "abandoned experiment" doctrine, which excludes abandoned experiments from the prior art and thus allows others to obtain patents on the technology. See id. at 33-35. The doctrine, however, does not apply to a technology that is patented prior to abandonment, and so it does not address the patent nondevelopment problem. Duffy also suggests that patent law might allow a third party to repatent the commercialized realization of an abandoned patented invention. See id. at 36-38. As Duffy recognizes, however, "[c]urrent U.S. patent law has no clear doctrine permitting" such repatenting. Id. at 36.
41. See Duffy, supra note 25, at 465-66.
42. See Martin & Partnoy Presentation, supra note 23.
43. See id.
44. See generally REAL OPTIONS AND INVESTMENT UNDER UNCERTAINTY: CLASSICAL READINGS AND RECENT CONTRIBUTIONS (Eduardo S. Schwartz & Lenos Trigeorgis eds., 2001) (containing many important contributions to the field).
45. See, e.g., Avinash Dixit, Investment and Hysteresis, in REAL OPTIONS AND INVESTMENT UNDER UNCERTAINTY: CLASSICAL READINGS AND RECENT CONTRIBUTIONS, supra note 36, at 153, 154;
46. Robert McDonald & Daniel Siegel, The Value of Waiting to Invest, in REAL OPTIONS AND INVESTMENT UNDER UNCERTAINTY: CLASSICAL READINGS AND RECENT CONTRIBUTIONS, supra note 36, at 253-54.
47. See Martin & Partnoy Presentation, supra note 23 (discussing patents as including the options to abandon the patent or delay development). The numerical example in the following paragraph also concretely illustrates how the right to wait until the optimal moment to develop a patent increases a patent's value.
48. See Martin & Partnoy Presentation, supra note 23.
49. See id.
50. See id.
51. See id.
52. See Duffy, supra note 25, at 493-96.
53. See generally id. (discussing various theories explaining patent races, their outcomes, and the social utility of the patent system).
54. See, e.g., Wolfgang Leininger, Escalation and Cooperation in Conflict Situations, 33 J. CONFLICT RESOL. 231, 233 (1989).
55. See id.; see also Michael R. Baye et al., Rigging the Lobbying Process: An Application of the All-Pay Auction, 83 AM. EGON. REV. 289 (1993) (providing an additional application of the all-pay auction model).
56. See sources cited supra note 46.
57. See Duffy, supra note 25, at 443-45 (discussing patentees' incentives to invest in patents as early as possible to gain exclusive rights to the invention).
58. See id. at 493-96.
59. See id. at 475-80; see also Harold Demsetz, Why Regulate Utilities?, 11 J.L. & ECON. 55, 63 (1968).
60. See Demsetz, supra note 50.
61. See Duffy, supra note 25, at 443-44. Duffy recognizes that research is probabilistic, and in extending his model, he assumes that inventors will only have some probability in each period of successfully meeting the requirements for patentability. See id. at 480-83.
62. See id. at 477 ("Where patent rights are limited in time, a competition to patent earlier will resemble a Demsetzian auction in that, by trying to be first to patent, the competing inventors are also vying to diminish their rents by placing the patent in the public domain sooner.").
63. See id. at 481 (assuming that "firms' expected profits are zero").
64. See id. at 465-66 (making assumptions about the costs and benefits of patenting without allowing for uncertainty).
65. See id. at 466 ("[P]rivate firms race to capture the rewards implicit in a patent grant. If the rewards are too great . . . then the firms will 'over-race' . . . .").
66. See supra Part I.A.1.
67. See John F. Duffy, A Minimum Optimal Patent Term (Jan. 9, 2003) (unpublished manuscript, available at http://www.ssrn.com/abstract=354282) (offering a model suggesting that an optimal patent term is at least as long as the existing patent term).
68. See id. at 4 ("The increased patent term causes innovation to occur earlier but the change in the time of innovation is less than the increase in patent term.").
69. See id.
70. Duffy has argued against providing windfall retroactive term extensions for intellectual property rights. See infra note 190 and accompanying text.
71. See Duffy, supra note 58, at 29-31; Duffy, supra note 25, at 494-96.
72. This assumption might be unjustified. For example, it might not make sense to commercialize a particular invention until another, complementary invention is developed. However, increasing the length of the patent term could also lead to earlier development of that invention.
73. See generally WILLIAM D. NORDHAUS, INVENTION, GROWTH, AND WELFARE 76-86 (1969) (providing the leading account of the competing effects of a longer patent term).
74. For instance, if inventors anticipate that they will be unable to commercialize their inventions until a particular date, precisely the opposite result may occur: a reduced risk of nondevelopment, but a delayed introduction of the invention into the public domain. Suppose that it will not be feasible to commercialize a particular invention, not yet created, until the year 2025. Suppose further that under the existing patent system, invention and patenting would occur in 2015, and the patent therefore would expire in 2035. If the patent term is lengthened by five years, but commercialization still cannot occur until 2025, invention will take place no earlier than 2010 with patent expiration still occurring in 2035. Because the invention will likely be more expensive to create in 2010 than in 2015, at least in present discounted value terms, invention is likely to occur after 2010 - say, perhaps, in 2011. As a result, the patent will expire in 2036.
75. See Martin & Partnoy Presentation, supra note 23; supra Part I.A.1.
76. See Duffy, supra note 25; supra Part I.A.2.
77. See Martin & Partnoy Presentation, supra note 23.
78. See Duffy, supra note 25, at 471-72 (illustrating the thesis).
79. See id. at 465-75.
80. See id.
81. See id.
82. See id.
83. See id.
84. See, e.g., Robert P. Merges & Richard R. Nelson, On the Complex Economics of Patent Scope, 90 COLUM. L. REV. 839, 908 (1990) ("Public policy . . . ought to encourage inventive rivalry, and not hinder it. . . . [A] rivalrous structure surely has its inefficiencies. But such a structure . . . seems a much better social bet than a regime where only one or a few organizations control the development of any given technology.");
85. see also Mark A. Lemley, Ex Ante Versus Ex Post Justifications for Intellectual Property, 71 U. CHI. L. REV. 129, 141-42 (2004) (concluding that the empirical literature supports the proposition that research will progress faster when no initial inventor is able to control it);
86. Tim Wu, Intellectual Property, Innovation, and Decentralized Decisions, 92 VA. L. REV. 123, 127-31 (2006) (arguing that decentralized decisionmaking, as occurs in the absence of patent protection, might produce more breakthrough inventions).
87. See Kitch, supra note 13, at 276 ("[A] patent 'prospect' increases the efficiency with which investment in innovation can be managed. . . . [T] he patent owner [is] in a position to coordinate the search for technological and market enhancement of the patent's value so that duplicative investments are not made . . . .").
88. See Duffy, supra note 25, at 465-75.
89. See supra note 1 and accompanying text (discussing how real property law addresses the problem of waste).
90. See Duffy, supra note 25, at 465-66. Duffy does not explicitiy calculate the social benefit of a prospect system versus that of a nonprospect system. While he notes that the prospect system produces the social gain of inventions entering the public domain sooner, see id. at 469, he does not note that this gain is partly offset by the innovation's being introduced later in a prospect system than in a nonprospect system.
91. See id. at 465-66.
92. See id.
93. See id.
94. See id.
95. See supra Part I.A.1.
96. See Duffy, supra note 25, at 469.
97. See Barzel, supra note 16, at 349.
98. See id. at 349-50; see also Duffy, supra note 25, at 440-41 (citing Barzel, supra note 16, at 352 n.11) ("Barzel had suggested that . . . the social surplus associated with the innovation [could be] preserved[ ] if the government assigned or auctioned off exclusive claims to develop technological opportunities at a very early time - before any resources were expended on developing the technology.").
99. See Barzel, supra note 16, at 354-55, 355 tb1.2.
100. See Duffy, supra note 25, at 469.
101. Id. at 466 (emphasis added).
102. The numbers used to derive the data in this and subsequent charts were obtained from a computer program developed in C++. This program, which is available from the author, calculates benefits and costs from the year 2000 to the year 2250 and discounts those benefits and costs to the year 2000. Because the discount rate is higher than the rate of growth, the effect of years beyond 2250 on the present discounted value of social welfare in the year 2000 is trivial.
103. Note, again, that in a prospect system, the inventor can delay completion of an invention until after she obtains a patent, whereas in a nonprospect system, the inventor must complete development of the invention in order to obtain a patent.
104. See Duffy, supra note 25, at 465-75.
105. For studies suggesting that there are considerable spillovers from patents as well as from other research and development, see Timothy F. Bresnahan, Measuring the Spillovers from Technical Advance: Mainframe Computers in Financial Services, 76 AM. ECON. REV. 742, 752-54 (1986),
106. and Edwin Mansfield et al., Social and Private Rates of Return from Industrial Innovations, 91 Q.J. ECON. 221, 233-35 (1977).
107. For example, it would be more realistic to assume that once the invention enters the public domain, social benefits increase somewhat because deadweight loss no longer exists. This seems likely to have only a slight effect in Figure 1, however, because the dates of patenting in the prospect and nonprospect systems appear to be quite close.
108. See Duffy, supra note 25, at 469-75; see also id. at 443-44 (noting that this is not necessarily counterbalanced by increased competition to obtain the patent prospect because earlier patenting can also dissipate rents).
109. See Kitch, supra note 13, at 267-70.
110. See Duffy, supra note 25, at 499-500.
111. See Which SSRI?, MED. LETTER ON DRUGS & THERAPEUTICS (The Med. Letter, Inc., New Rochelle, N.Y.), Nov. 24, 2003, at 93 (providing an overview of different SSRIs). William Landes and Richard Posner also use SSRIs to demonstrate that a patent race may have multiple winners.
112. See WILLIAM M. LANDES & RICHARD A. POSNER, THE ECONOMIC STRUCTURE OF INTELLECTUAL PROPERTY LAW 301-02 (2003).
113. See Which SSRI?, supra note 99, at 93 ("Some patients who fail to respond to one SSRI may respond to another, possibly because of differences in tolerability.").
114. These benefits partiy justify a recent proposal for a nonexclusive patent system. See John S. Leibovitz, Note, Inventing a Nonexclusive Patent System, 111 YALE L.J. 2251 (2002) (arguing for a patent system that would grant patents to more than one participant in the patent race).
115. This failure of a market entrant to care that some of its business comes at the expense of existing market participants is known in the industrial organization literature as "demand diversion" or "business stealing." See, e.g., N. Gregory Mankiw & Michael D. Whinston, Free Entry and Social Inefficiency, 17 RAND J. ECON. 48, 49 (1986) (explaining how this failure may lead to economically efficient entry).
116. Note that the economic literature on product differentiation suggests that it is possible to have too many or too few differentiated products in an imperfectly competitive market. See, e.g., Avinash K. Dixit & Joseph E. Stiglitz, Monopolistic Competition and Optimum Product Diversity, 67 AM. ECON. REV. 297 (1977).
117. In Figure 3, the prospect invention year line reflects only those cases in which the patent is in fact developed. In a more realistic model, almost every potential invention will eventually be developed even if it is in the public domain. Such a model, however, would only make the results here stronger by providing a longer average delay between invention and development.
118. For example, when the appropriability factor is 0.5, the invention is developed in the prospect system only 5.4% of the time. Note again that in Figure 3, the prospect invention year refers to the average year of invention only in those cases in which the invention is in fact developed. The development rate predictably increases with higher levels of appropriability. Note that this simulation begins in 1980, not 2000, because uncertainty made it profitable to acquire patents earlier. The prospect patent year line, which is cut off in the graph, descends to 1996 with appropriability equal to 1.0.
119. See supra note 19 and accompanying text.
120. Cf. Thomas L. Irving & Michael D. Kaminski, Double Patenting: One Way, Two Way; Whose Delay?, 1 U. BALT. INTELL. PROP. L.J. 180, 185-87 (1993) (considering policy justifications in support of the doctrine preventing "double patenting" of the same invention as a means of extending the patent term).
121. Further simulations performed by the present author (but not worth reporting in detail here) reveal that if eligibility for a patent requires 10% of the cost of the invention process, the development rate rises to about 82% if half of the social benefit is appropriable, and with 50% of the cost of the invention process demanded, all inventions are developed. But this should not lead to complacency that nondevelopment is not a problem because the assumed shock distribution, with an absolute value averaging 12.5%, is very conservative. Many patented products will have substantial uncertainty about their cost of completion and especially about their benefit since many developed products turn out to be commercial disasters. With 50% of the cost of the invention process demanded in the prospect system and an absolute average shock of 37.5%, only about 25% of inventions will be developed; thus, a nonprospect system will produce greater social value. As these numbers suggest, confident assessments about the relative merits of prospect and nonprospect systems require more empirical analysis of the information available to inventors. A more complete model would also need to consider the effects of asymmetric information among racers and the possibility that different racers may have information about the relative success of other racers' efforts.
122. For an argument that the patent system does not in fact give as much control over future development as Kitch claims because the patent statutes limit the patentee's monopoly to his claim and provide no basis for monopolizing future technology, see Roger L. Beck, The Prospect Theory of the Patent System and Unproductive Competition, 5 RES. L. & ECON. 193, 194-95 (1983). The requirements for a patent, moreover, are nontrivial so it may be unrealistic to assume that patents are granted when a patentee completes only a tiny fraction of the work.
123. For examples of works that seek to overcome these difficulties, see John R. Allison et al., Valuable Patents, 92 GEO. L.J. 435, 437 (2004), which analyzes results from a detailed study of litigated patents to determine what makes a patent valuable,
124. and James Bessen & Eric Maskin, Sequential Innovation, Patents, and Imitation 1-2 (Mar. 2006) (unpublished manuscript, available at http://www.sss.ias.edu/publications/papers/econpaper25.pdf), which analyzes the impact of patent protection on the software industry.
125. See, e.g., Denton & Heald, supra note 30, at 1180, 1182 & n.26 (citing various methodologies and proposing a new patent valuation method using the Black-Scholes equation for stock options pricing).
126. See sources cited supra note 94 (seeking to measure spillovers from research activity).
127. See 35 U.S.C. §§ 101-103 (2000) (requiring the patentee to satisfy subject matter requirements and standards of novelty, utility, and nonobviousness).
128. Id. § 103(a).
129. See, e.g., ADAM B. JAFFE & JOSH LERNER, INNOVATION AND ITS DISCONTENTS 32-34 (2004) (citing patents granted for a peanut butter and jelly sandwich, the process of toasting bread, and a method of swinging on a swing).
130. See 35 U.S.C. §§ 101-103.
131. See id. § 112. Some debate whether the written description and enablement requirements are distinct or are one and the same. See Univ. of Rochester v. G.D. Searle & Co., 375 F.3d 1303, 1304 (Fed. Cir. 2004) (Newman, J., dissenting) (noting a "burgeoning conflict in pronouncements of this court" regarding § 112).
132. See 35 U.S.C. § 112; Teleflex, Inc. v. Ficosa N. Am. Corp., 299 F.3d 1313, 1330 (Fed. Cir. 2002) (setting forth a two-pronged test for compliance with the best mode requirement). See generally Kenneth R. Adamo, What's Better, What's Best - The Best Mode Requirement in U.S. Patent Practice, 73 J. PAT. & TRADEMARK OFF. SOC'Y 811 (1991) (discussing policies behind the best mode requirement and how courts typically apply the doctrine).
133. See 35 U.S.C. § 112; W.L. Gore & Assocs. v. Garlock, Inc., 721 F.2d 1540, 1557 (Fed. Cir. 1983) ("A patent is invalid only when those skilled in the art are required to engage in undue experimentation to practice the invention."). As early as 1888, the Supreme Court held that, to qualify for a patent, an inventor need only "describe[ ] his method with sufficient clearness and precision to enable those skilled in the matter to understand what the process is, and . . . point[ ] out some practicable way of putting it into operation." The Telephone Cases, 126 U.S. 1, 536 (1888).
134. See Duffy, supra note 32, at 6.
135. Id.; see In re Strahilevitz, 668 F.2d 1229, 1232 (C.C.P.A. 1982) (holding that § 112 does not require the patentee to submit working models to obtain a patent).
136. Duffy, supra note 32, at 6.
137. See supra Part I.B.2.
138. See supra Part I.B.
139. See Martin & Portnoy Presentation, supra note 23.
140. See Lemley, supra note 75, at 131-32.
141. Id. at 141 ("Prospect theory is needed when control over subsequent development is a necessary part of the incentive to produce the pioneering invention in the first place, as is arguably true with pharmaceuticals.").
142. Orphan Drug Act, Pub. L. No. 97-414, 96 Stat. 2049 (1983) (codified as amended at 21 U.S.C. §§ 360aa-360ee (2000)). The seven-year exclusivity provision is found at § 360cc(a).
143. See 21 U.S.C. § 360bb(a) (2) (A).
144. See id. § 360bb(a) (2) (B).
145. See Orphan Drug Amendments of 1985, Pub. L. No. 99-91, 99 Stat. 387 (codified as amended at 21 U.S.C. § 360cc (2000)).
146. Drug Price Competition and Patent Term Restoration (Hatch-Waxman) Act of 1984, Pub. L. No. 98-417, 98 Stat. 1585 (codified as amended in scattered sections of 15, 21, 28, & 35 U.S.C.).
147. For an overview of Hatch-Waxman and its legislative history, see Gerald J. Mossinghoff, Overview of the Hatch-Waxman Act and Its Impact on the Drug Development Process, 54 FOOD & DRUG L.J. 187 (1999).
148. See 35 U.S.C. § 156(c) (2) (2000). The statute, however, allows for a maximum extension period of five years and no more than a fourteen-year period of market exclusivity. See id. § 156(c) (3), (d) (5) (E).
149. See id. §156(c) (1).
150. For a general discussion of the science of genomics and recent developments in the field, see Helen M. Berman & Rochelle C. Dreyfuss, Reflections on the Science and Law of Structural Biology, Genomics, and Drug Development, 53 UCLA L. REV. 871, 880-83 (2006).
151. Cf. id. at 884-85 (noting that "lead optimization," just one stage in the drug development process, can take "several years" to finish and costs between two and four million dollars). The genomics revolution also presents a related problem: once it becomes feasible to test many people for individual genetic variations, researchers will need to test many existing treatments to determine which groups of the population can benefit from those treatments. But, without protection for the underlying treatments, there will be litde incentive to engage in such testing. See Arti K. Rai, Pharmacogenetic Interventions, Orphan Drugs, and Distributive Justice: The Role of Cost-Benefit Analysis, 19 SOC. PHIL. & POL'Y 246, 249-51 (2002).
152. See Utility Examination Guidelines, 66 Fed. Reg. 1092, 1093 (Jan. 5, 2001). The concern animating the Guidelines was not the problem of patent nondevelopment, but simply the concern that patents were being granted without sufficient inventive accomplishment. See generally Lawrence T. Kass & Michael N. Nitabach, A Roadmap for Biotechnology Patents? Federal Circuit Precedent and the PTO's New Examination Guidelines, 30 AIPLA Q.J. 233, 248-65 (2002) (providing an overview of the Guidelines).
153. Under patent-claiming rules, "[s]imulated or predicted test results and prophetical examples . . . are permitted in patent applications." U.S. PATENT & TRADEMARK OFFICE, U.S. DEP'T OF COMMERCE, MANUAL OF PATENT EXAMINING PROCEDURE § 608.01 (p), at 600-99 (8fh ed., 5th rev. 2006), available at http://www.uspto.gov/web/offices/pac/mpep/mpep_e8r5_0600. pdf.
154. See, e.g., Harold C. Wegner & Stephen B. Maebius, The Global Biotech Patent Application, in BIOTECHNOLOGY LAW: BIOTECHNOLOGY PATENTS & BUSINESS STRATEGIES IN THE NEW MILLENNIUM 87, 96-97 (PLI Intellectual Prop., Course Handbook Series No. G-666, 2001) (discussing the tradition of prophetic claiming of biotech patents).
155. See, e.g., Energy Absorption Sys., Inc. v. Roadway Safety Servs., Inc., No. 96-1264, 1997 WL 368379, at *5 (Fed. Cir. July 3, 1997) (discussing the prophetic examples doctrine in the context of a mechanical invention).
156. See supra notes 132-35 and accompanying text.
157. See supra note 9 and accompanying text.
158. See Lemley, supra note 9, at 1010 ("The original patent owner can prevent the improver from using his patented technology, but the improver can also prevent the original patent owner from using the improvement. Unless the parties bargain, no one gets the benefit of the improvement"). See generally Michael S. Mireles, An Examination of Patents, Licensing, Research Tools, and the Tragedy of the Anticommons in Biotechnology Innovation, 38 U. MICH. J.L. REFORM 141, 168 (2004) (explaining the phenomenon of blocking patents).
159. An additional reason is that improvement patents may provide suboptimal incentives even for patentable development activities. See infra Part II.C.
160. Edmund Kitch recognized how important it is for patentees to engage in scientific testing of their inventions but assumed that the underlying patent would provide them with sufficient incentives to do so. See Kitch, supra note 13, at 277 ("Absent a patent, firms have less than the optimal incentive to invest in providing information about and techniques for using the new technology."). But, because patents only protect an invention for a finite period, they too provide less than optimal incentives for patentees to invest in scientific testing.
161. See supra notes 128-30 and accompanying text.
162. See 35 U.S.C. § 101 (2000).
163. See, e.g., Rebecca S. Eisenberg, The Problem of New Uses, 5 YALE J. HEALTH POL'Y L. & ETHICS 717, 720 (2005).
164. See, e.g., Allegheny Drop Forge Co. v. Portec, Inc., 541 F.2d 383, 386 (3d Cir. 1976) ("A new use for an old process or product is patentable if the new use or application is itself not 'obvious' to one skilled in the art").
165. See, e.g., Eisenberg, supra note 149, at 720 ("The discovery of a new use for an old drug might support a patent on a method of treatment, but such a patent offers little effective protection against generic competition once the drug itself is off-patent and may lawfully be sold for an older, unpatented use.").
166. See In re Cruciferous Sprout Litigation, 301 F.3d 1343 (Fed. Cir. 2002).
167. Id. at 1351.
168. Id. at 1352.
169. See Editorial, The First Race-Based Medicine, N.Y. TIMES, June 19, 2005, § 4, at 11.
170. See id. NitroMed received an improvement patent for using the drug on African-American patients. See U.S. Patent No. 6,784,177 (filed Aug. 2, 2002) (issued Aug. 31, 2004).
171. See Kieff, supra note 19, at 708-09 (noting the advantage enjoyed by a second-mover since "mere knowledge of a first mover's success eliminates a great deal of risk from the second mover's decision whether to embark on the same enterprise").
172. See Gideon Parchomovsky & Peter Siegelman, Towards an Integrated Theory of Intellectual Property, 88 VA. L. REV. 1455, 1473-74 (2002). Gideon Parchomovsky and Peter Siegelman have argued that "leveraging" a patent dirough a trademark is socially optimal because it increases a patentee's incentive to invent without increasing deadweight loss. See id. at 1473-86.
173. See, e.g., Kitch, supra note 13, at 272 tbl.1.
174. See Timothy J. Muris, California Dental Association v. Federal Trade Commission: The Revenge of Footnote 17, 8 SUP. CT. ECON. REV. 265, 293-97 (2000) (surveying reports on the effects of advertising on the cost and quality of products).
175. See, e.g., Ernst R. Berndt et al., The Long Shadow of Patent Expiration: Generic Entry and RX to OTC Switches, in SCANNER DATA AND PRICE INDEXES 229, 261 (Robert C. Feenstra & Matthew D. Shapiro eds., 2003), available at http://faculty.london. edu/mkyle/Rx%20to%20OTC%20paper.pdf (reporting that pharmaceutical companies decrease their spending on drug advertising in the years before the patent expires).
176. See, e.g., Nicholas Kaldor, The Economic Aspects of Advertising, 18 REV. ECON. STUD. 1, 5-6 (1950-51).
177. See id.; see also Avinash Dixit & Victor Norman, Advertising and Welfare: Another Reply, 11 BELL J. ECON. 753, 753 (1980) (suggesting the possibility of socially inefficient competitive advertising).
178. See supra text accompanying notes 143-44.
179. See supra note 94.
180. See Mireles, supra note 144, at 168. Obviously, the original patentee can practice the original invention but cannot practice the improvement. See id. The improver can practice neither the original invention nor the improvement without a license from the original patentee. See id.
181. See Oren Bar-Gill & Gideon Parchomovsky, Essay, The Value of Giving Away Secrets, 89 VA. L. REV. 1857, 1860 (2003) (suggesting that patentees do this by publishing without patenting aspects of their inventions).
182. Suppose there is a 50% chance that there will be no market for the improved product and a 50% chance that consumers will be willing to spend enough extra for the improved product to produce an additional $2 million for a single improver. If the underlying invention remained under patent protection, then the patentee would be willing to invest up to $1 million in the improvement.
183. See supra Parts I.A.2, I.B (discussing Duffy's model).
184. See Duffy, supra note 25, at 489-90 (arguing that the patentee has an advantage, though only a limited one, over competitors in the search for improvements). If, however, assuring patentees that they will be able to win the race for improvement will lead to earlier searches for such improvements, Duffy's analysis arguably provides a basis for granting patentees a stronger property right.
185. See Mireles, supra note 144, at 168.
186. See 35 U.S.C. § 103(a) (2000).
187. See Douglas Gary Lichtman, The Economics of Innovation: Protecting Unpatentable Goods, 81 MINN. L. REV. 693, 712-14 (1997) (discussing the possibility that some significant inventions will be ineligible for patent protection).
188. See generally JAFFE & LERNER, supra note 115 (documenting the problem of overly relaxed standards and suggesting potential reforms).
189. See supra note 31 (discussing the cost of patent prosecution).
190. See supra Part I.A.2.
191. Prior to 1995, the patent term was seventeen years from the date of issuance. See Uruguay Round Agreements Act, Pub. L. No. 103-465, § 532(a), 108 Stat. 4809, 4984 (1994) (codified as amended at 35 U.S.C. § 154 (2000)); see also Mark A. Lemley & Kimberly A. Moore, Ending Abuse of Patent Continuations, 84 B.U. L. REV. 63, 80 (2004).
192. See 35 U.S.C. § 154(a) (2).
193. See Lemley & Moore, supra note 177, at 79-80.
194. See 35 U.S.C. §§ 101-103.
195. See supra note 11 and accompanying text.
196. See Oren Bar-Gill & Gideon Parchomovsky, Essay, A Marketplace for Ideas?, 84 TEX. L. REV. 395 (2005). The authors argue that those who conceive of ideas that need development should get intellectual property protection but then be required to auction their ideas to potential idea developers. See id. at 399-402. If the high bidder in the development stage fails "to produce a patent or product within a given time period, say two years," id. at 421, then the idea would be placed in the public domain, id. at 418-19. Because this idea regime is virtually identical to the existing patent regime, see id. at 426-27 (imposing traditional requirements for patentability along with a requirement of "developability"), the proposal is effectively close to one that would combine mandatory auctions with a development requirement. To the extent that nondevelopment is a significant concern for their proposal, a better approach might be to penalize a nondeveloper by reauctioning the intellectual property right and by making auction revenues payable to the government.
197. See supra pp. 1069-70. Thus, a development mandate will only achieve its goal if placing the idea in the public domain successfully deters nondevelopment and did not often actually need to occur.
198. For a history of nineteenth-century private patent extension bills, see Richard M. Cooper, Legislative Patent Extensions, 48 FOOD & DRUG L.J. 59, 59-62 (1993).
199. See id. at 62-73.
200. This program is known as the Drug Price Competition and Patent Term Restoration (Hatch-Waxman) Act of 1984, Pub. L. No. 98-417, 98 Stat. 1585 (codified as amended in scattered sections of 15, 21, 28, & 35 U.S.C). See Cooper, supra note 184, at 62-73; supra notes 132-35 and accompanying text.
201. See supra Part II.A.
202. See Mossinghoff, supra note 133, at 188.
203. See William M. Landes & Richard A. Posner, Indefinitely Renewable Copyright, 70 U. CHI. L. REV. 471 (2003).
204. See John F. Duffy, Intellectual Property Isolationism and the Average Cost Thesis, 83 TEX. L. REV. 1077, 1094 (2005) ("In a world where . . . government-conferred property rights are increasingly auctioned rather than gifted," laws providing intellectual property extensions "look more anomalous, and there is a general and principled theory for resisting them . . . .").
205. In a second-price sealed bid auction, the winner pays the amount of the second-highest bid; this gives each bidder an incentive to bid the bidder's actual valuation. See William Vickrey, Counterspeculation, Auctions, and Competitive Sealed Tenders, 16 J. FIN. 8, 20-23 (1961) (introducing this auction form and discussing its benefits).
206. But cf. Dan L. Burk & Mark A. Lemley, Is Patent Law Technology-Specific?, 17 BERKELEY TECH. L.J. 1155, 1183-85 (2002) (arguing that although the rules of patent law are the same across technology areas, in practice their application varies by industry). 193 See Duffy, supra note 58, at 1 ("The problem of fixing an optimal patent term . . . is generally viewed as achieving a balance between the incentives necessary to encourage innovation and the inefficiencies associated with the exclusive right.").
207. For an analysis of the possible benefits and costs of individualized tailoring of intellectual property protection, see Glynn S. Lunney, Jr., Patent Law, the Federal Circuit, and the Supreme Court: A Quiet Revolution, 11 SUP. CT. ECON. REV. 1, 39-48 (2004).
208. For an article suggesting the possibility of nonuniform patent terms with patentees selecting their own patent terms based on a fee schedule, see Francesca Cornelli & Mark Schankerman, Patent Renewals and R&D Incentives, 30 RAND J. ECON. 197, 197-98 (1999).
209. See supra Part II.B.3.
210. Payment ought to be to the government rather than to the patentee. Otherwise, the patent extension auction is simply a patent extension program in disguise, and the patentee would always be able to keep the patent by entering a very high bid.
211. See Parchomovsky & Siegelman, supra note 158, at 1462.
212. See Michael Abramowicz, Perfecting Patent Prizes, 56 VAND. L. REV. 115, 200-05 (2003) (arguing that the monopoly distortion from patents is generally greater than the distortion associated with taxation).
213. The more common situation in which taxes may promote internalization involves negative externalities. See generally ARTHUR CECIL PIGOU, THE ECONOMICS OF WELFARE (Transaction Publishers 2002) (1920) (providing the classic defense of using taxation to internalize negative externalities). A patent extension achieved through an auction may allow internalization of the positive externalities of patent development.
214. Cf. Abramowicz, supra note 198, at 209-10 (discussing the typical problem of rent seeking in the current patent context).
215. See infra notes 203, 209 and accompanying text.
216. See Abramowicz, supra note 198, at 210.
217. For an overview of how the owner of an intellectual property right might seek to tie the protected product with another to extend the owner's market power, see Troy Paredes, Comment, Copyright Misuse and Tying: Will Courts Stop Misusing Misuse?, 9 HIGH TECH. L.J. 271, 298-302 (1994).
218. A patent extension might be useful for reasons other than patent underdevelopment, such as when the patentee has special expertise in price discriminating and, thus, in improving efficiency. See generally Harold Demsetz, The Private Production of Public Goods, 13 J.L. & ECON. 293, 301-02 (1970) (explaining how price discrimination can improve efficiency in the provision of public goods). Here, however, I have focused solely on whether an auction mechanism might solve the patent underdevelopment problem.
219. A variant would require the patentee to submit a sealed bid at the same time third parties submit sealed bids and then would award the patentee the patent if and only if its sealed bid exceeded the next highest by at least the markup. In this regime, third parties could offer bids as percentages of the patentee's bid up to a specified maximum. This might be useful if third parties are not expected to have good information about the patent's value. A third party would then need to consider only the extent to which the patent would be less valuable to it as a result of the lack of ownership continuity and the transaction costs associated with selling the patent back to the patentee after the expiration of the original term.
220. See supra note 191 and accompanying text (discussing the second-price sealed bid auction approach).
221. Conceivably, auctions could reduce the cost of that failure. For example, the government might use the auction mechanism suggested by Michael Kremer. See Michael Kremer, Patent Buyouts: A Mechanism for Encouraging Innovation, 113 Q.J. ECON. 1137, 1146-48 (1998). Under that mechanism, an auction would be used to value a patent, but would consummate with only some low positive probability. See id. at 1147. With high probability, the price signals from the auction would help derive a value to compensate the patent holder who, in this case, is the winner of the patent extension at the initial auction. See id. There are obvious affinities between my proposal and Kremer's: both seek to use information from bids for patents at auction to overcome problems with the patent system. See id. The ends, however, are opposite. Kremer's proposal is potentially useful when there are efficiency advantages to placing a patent in the public domain, while the approach described here is useful when there are efficiency reasons to prevent a patent from falling into the public domain. They could, however, be combined to account for cases in which the patentee's initial decision turns out to be mistaken.
222. The optimal markup would decline with an increase in the penalty percentage because the penalty percentage would make the option of declining the extension less attractive.
223. See generally Parchomovsky & Wagner, supra note 29 (arguing that patents may be more valuable as part of patent portfolios than alone).
224. See, e.g., Douglas Gary Lichtman, Pricing Prozac: Why the Government Should Subsidize the Purchase of Patented Pharmaceuticals, 11 HARV. J.L. & TECH. 123, 132 n.21 (1997) ("When facing linear demand and zero marginal cost, a monopolist maximizes profit by selling to exactly half the consumers. Geometrically, this means that [consumer surplus] = [deadweight loss] = (0.5 * [producer surplus]).").
225. See Michael Abramowicz, Patent Auctions 43-44 (2005) (unpublished manuscript, on file with author).
226. See Ian Ayres & Paul Klemperer, Limiting Patentees' Market Power Without Reducing Innovation Incentives: The Perverse Benefits of Uncertainty and Non-Injunctive Remedies, 97 MICH. L. REV. 985, 1028-31 (1999).
227. Ayres and Klemperer provide a general argument for lowering the level of patent damages and for limiting injunctive relief, but a jurisdiction seeking to implement their proposal could set the price level without holding an auction. See id.
228. Conceivably, a third party diat won an auction could also seek another patent extension. The third party's victory, however, would ordinarily signal that the benefits of ownership continuity were not as high as the patentee thought they might be.
229. Mark Lemley interprets the theory that intellectual property rights may ensure the efficient management of property to imply that "there seems little reason to terminate that right after a period of years." See Lemley, supra note 75, at 131. Indeed, my analysis suggests that a definite termination may not be optimal. Lemley might overstate the point, however, because those who advance what he refers to as "ex post" justifications for intellectual property concede that intellectual property rights have costs, including deadweight loss. See id. at 129. The challenge, which I seek to meet here, is to find a way to extend patents just long enough to allow for efficient management and development of the property rights up until the point where the cost of further extensions would exceed their benefit.
230. See Landes & Posner, supra note 189.
231. See 17 U.S.C. § 106(2) (2000).
232. Id. §§ 302-304.
233. Landes & Posner, supra note 189, at 484-88; see Michael Abramowicz, A Theory of Copyright's Derivative Right and Related Doctrines, 90 MINN. L. REV. 317 (2005) (arguing that the derivative right is justified in part because it reduces the number of derivative works made).
234. See Landes & Posner, supra note 189, at 487-88 ("Not only would the public rapidly tire of Mickey Mouse, but his image would be blurred, as some authors portrayed him as a Casanova, others as catmeat, others as an animal-rights advocate, still others as the henpecked husband of Minnie.").