Extreme value or trolls on top? The characteristics of the most-litigated patents

University of Pennsylvania Law Review

Volumn 158, Issue 1, 2009, Pages 1-37

  Allison, John R ^a   Lemley, Mark A ^b   Walker, Joshua ^c  

^a UNIVERSITY OF TEXAS AT AUSTIN   (United States)

^b STANFORD LAW SCHOOL   (United States)

^c STANFORD UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 74049103030     PISSN: 00419907     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Review

References (107)

1. On these two views of the world, see DAN L. BURK & MARK A. LEMLEY, THE PATENT CRISIS AND HOW THE COURTS CAN SOLVE IT 3-6 (2009).
2. See, e.g., Patent Trolls: Fact or Fiction: Hearing Before the Subcomm. on Courts, the Internet, and Intell. Prop. of the H. Comm. on the fudiciary, 109th Cong. 14-18 (2006) (statement of Dean Kamen, President, Deka Research & Development Corporation) [hereinafter Kamen Testimony}, available at http://frwebgate.access.gpo.gov/cgibin/getdoc.cgi?dbname=109- house-hearings&docid=f:28201.pdf (questioning whether patent trolls even exist); Jennifer Kahaulelio Gregory, The Troll Next Door, 6 J. MARSHALL REV. INTELL. PROP. L. 292, 305-09 (2007) (arguing that the term "patent troll" obscures real problems in the patent system); Mark A. Lemley & Carl Shapiro, Patent Holdup and Royalty Stacking, 85 TEX. L. REV. 1991, 2009 (2007) (asserting that nonpracticing entities account for thirty to forty percent of suits in the computer and electronics industries); Marc Morgan, Stop Looking Under the Bridge for Imaginary Creatures: A Comment Examining Who Really Deserves the Title Patent Troll, 17 FED. CIR. B.J. 165, 166 (2007) ("In reality only two percent of all patent litigation is linked to so-called trolling." (footnote omitted)).
3. Press Release, Stanford Law Sch., Stanford Law School Launches Intellectual Property Litigation Clearinghouse (Dec. 8, 2008), available at http://www.law.stanford.edu/program/centers/iplc/#press-releases. The website of the Stanford IP Litigation Clearinghouse (IPLC) is located at http://lexmachina.stanford.edu.
4. When we speak of value, we refer to private value, or value to the owner. We also refer only to the fact of value and not to any quantitative measure of value. We defend the litigation-value connection extensively in our prior work. See john R. Allison et al., Valuable Patents, 92 GEO. L.J. 435, 440-43 (2004) (explaining the intuitive litigationvalue connection and its strong empirical support);
5. see also John R. Allison & Thomas W. Sager, Commentary, Valuable Patents Redux: On the Enduring Merit of Using Patent Characteristics to Identify Valuable Patents, 85 TEX. L. REV. 1769, 1794 (2007) (defending the value-litigation connection). The economics literature supports this connection as well.
6. See, e.g., Dietmar Harhoff et al., Citations, Family Size, Opposition and the Value of Patent Rights, 32 RES. POL'Y 1343, 1345 (2003) (finding that "patents which are upheld against opposition ... are particularly valuable");
7. Jean O. Lanjouw & Mark Schankerman, Characteristics of Patent Litigation: A Window on Competition, 32 RAND J. ECON. 129, 140 (2001) [hereinafter Lanjouw & Schankerman, Characteristics] (finding that litigation correlates with patent value and that "the number of claims is another ... indicator of the 'bits of information' contained in a patent, and therefore its value");
8. Stuart J.H. Graham et al., Post-Issue Patent "Quality Control": A Comparative Study of U.S. Patent Re-examinations and European Patent Oppositions 6-22 (Nat'l Bureau of Econ. Research, Working Paper No. 8807, 2002) (comparing USPTO and EPO opposition mechanisms and finding that the most valuable patents were challenged in both systems); Dietmar Harhoff & Markus Reitzig, Determinants of Opposition Against EPO Patent Grants-The Case of Biotechnology and Pharmaceuticals 4 (Ctr. for Econ. Policy Research, Discussion Paper No. 3645, 2002) (confirming that "patents with above-average values are more likely to be attacked");
9. Jean O. Lanjouw & Mark Schankerman, Enforcing Intellectual Property Rights 5 (Nat'l Bureau of Econ. Research, Working Paper No. 8656, 2001) ("[M]ore valuable patents ... are much more likely to be involved in suits.");
10. cf. Jean O. Lanjouw & Josh Lerner, The Enforcement of Intellectual Property Rights: A Survey of the Empirical Literature, 49/50 ANNALES D'ECONOMIE ET DE STATISTIQUE 223 (1998) (surveying the literature on the issue). This Article both strengthens that conclusion and demonstrates for the first time a strong relationship between the number of times a patent is enforced and the determinants of value. It also allows us to refute a hypothesis that we addressed but could not resolve in our prior work: that litigated patents are valuable, but not the most valuable patents. Litigation, on this account, is evidence of weakness in a patent. Perhaps competitors quietly take licenses to the truly valuable patents, and the ones they fight about are akin to an "upper-middle class" of potentially valuable but less-than-perfect patents.
11. Mark A. Lemley, Rational Ignorance at the Patent Office, 95 NW. U. L. REV. 1495, 1507 (2001).
12. See, e.g., Jonathan A. Barney, A Study of Patent Mortality Rates: Using Statistical Survival Analysis to Rate and Value Patent Assets, 30 AIPLA QJ. 317, 329 (2002) ("A relatively large number of patents appear to be worth little or nothing while a relatively small number appear to be worth a great deal.");
13. Jean O. Lanjouw et al., How to Count Patents and Value Intellectual Property: The Uses of Patent Renewal and Application Data, 46 J. INDUS. ECON. 405, 412 & n.3 (1998) (noting a variety of studies finding "that most patents are of very little value");
14. Kimberly A. Moore, Worthless Patents, 20 BERKELEY TECH.. LJ. 1521, 1526 (2005) (documenting the failure of most patentees to pay maintenance fees costing only a few thousand dollars) ;
15. Thomas Ewing, Book Review, 43 SANTA CLARA L. REV. 631, 633 (2003) ("Some of the authors simply recount patent procurement and litigation statistics ad nauseum and do not seem to understand that some patents really do have no value whatsoever since no one would ever practice the disclosed technology, as claimed."). Maintenance fees are due in increasing amounts at 3.5 years, 7.5 years, and 11.5 years after the patent issues. 35 U.S.C. § 41(b) (2006). The fees are $830 at 3.5 years, $1, 900 at 7.5 years, and $2,910 at 11.5 years. Id. Those fees are halved for small entities. Id. § 41 (h) (1).
16. See AM. INTELL. PROP. LAW ASS'N, REPORT OF THE ECONOMIC SURVEY, at 1-91 (2007) (reporting that the median cost of high-stakes patent lawsuits is five million dollars per side in legal fees through trial).
17. Allison et al., supra note 4; see also Allison & Sager, supra note 4, at 1794 (defending the statistical power of the results in the earlier study).
18. Allison et al., supra note 4, at 438.
19. Id. at 468-69.
20. Id. at 460.
21. Due to increased availability of electronic filings in federal court, the ability to identify patents in suit has improved markedly, particularly since 2003. Moreover, electronic access varies by district, potentially making this patent data set underinclusive for certain districts despite manual collection of cases from those districts. Nevertheless, the patents identified represent the best, most representative data set available.
22. For purposes of this analysis, we include declaratory-judgment actions as well as actions filed by the patent owner. Until 2007, the rules for declaratory judgment required a clear threat of suit by the patent owner.
23. See Teva Pharm. USA, Inc. v. Pfizer, Inc., 395 F.3d 1324, 1333 (Fed. Cir. 2005) (requiring reasonable apprehension of imminent suit), abrogated by Medlmmune, Inc. v. Genentech, Inc., 549 U.S. 118 (2007) (permitting a declaratory judgment action so long as there is an actual controversy between the parties). We count only separate lawsuits; many patent lawsuits are filed against multiple defendants in a single proceeding.
24. We do not address the outcomes of those lawsuits in this Article; at is the subject of a companion piece by the authors, tentatively entitled Patent Quality and Risk Aversion Among Repeat Patent Litigants.
25. We did not attempt to create a comprehensive typology of such areas, but for obvious reasons only identified and defined those technology and industry areas we actually encountered in the population of the 106 most-asserted patents and the sample of 106 once-litigated patents. Although the size of our data sets is sufficient for sound statistical analysis, the relatively small numbers of observations necessarily results in our having encountered fewer technology and industry areas than we would have encountered in a much larger patent data set The technology categories are not necessarily mutually exclusive because modern inventions so often involve multiple technologies. Our industry categories are also not all mutually exclusive, reflecting the reality of modern industry crossovers. For example, a software-implemented telecommunications process or product rightly belongs in both a computer and a communications industry category. There are, however, fewer inventions belonging in more than one industry category than there are inventions belonging in more than one technology area because mixes of technologies in inventions are more common than industry crossovers.
26. John R. Allison & Mark A. Lemley, Who's Patenting What? An Empirical Exploration of Patent Prosecution, 53 VAND. L. REV. 2099, 2114 (2000).
27. Although we report descriptive statistics and bivariate comparisons for all of our technology and industry areas, we defined a few technology and industry areas encountered in the data set that we ultimately did not subject to statistical testing because the number of observations for such areas was so small as to render statistical analysis meaningless. We did this to create definitions of categories that could also be used in data sets other than the ones we used in this study.
28. The difficulty in defining a software patent, and the detailed reasoning that supports our definition, can be found in Arti K. Rai, John R. Allison & Bhaven N. Sampat, University Software Ownership and Litigation: A First Examination, 87 N.C. L. REV. 101, 111-15 (2009).
29. U.S. Patent No. 5, 551, 216 (filedJuly 20, 1995) (issued Sept. 3, 1996).
30. Mark Lemley and Nathan Myhrvold are currently working on an article, tentatively entitled The Complex Ecology of Patent Plaintiffs, employing this method.
31. After a diligent search, we could not identify the entity status of a few patent owners. We have classified those entities as 10 (Undetermined) and have excluded them from our entity-status analyses.
32. In social science research, a confidence level of 95% (p-value of 0.05) is typically treated as sufficient to show statistical significance. Thus, a confidence level of 99% (p-value of 0.01) shows a far greater degree of confidence that the differences are not due to random chance.
33. We used accepted statistical techniques to adjust many of the variables to normalize skewed distributions before making bivariate statistical comparisons.
34. For a discussion of patent continuations, see Mark A. Lemley & Kimberly A. Moore, Ending Abuse of Patent Continuations, 84 B.U. L. REV. 63 (2004).
35. This explains the otherwise odd result that the median number of applications filed is 1.5-exactly half of the patents had one application, and half had more than one.
36. A number of studies have used forward citations as evidence of patent value. See, e.g., Bronwyn H. Hall, Adam Jaffe & Manuel Trajtenberg, Market Value and Patent Citations, 36 RANDJ. ECON. 16, 29-30 (2005) (finding that an extra citation per patent increases a firm's market value by three percent);
37. Dietmar Harhoff et al., Citation Frequency and the Value of Patented Inventions, 81 REV. ECON. & STAT. 511, 515 (1999) (concluding from surveys in the United States and Germany that patents renewed to full term are more frequently cited than those that lapse and that citation frequency increases with economic value for full-term patents);
38. Manuel Trajtenberg, A Penny for Your Quotes: Patent Citations and the Value of Innovations, 21 RAND J. ECON. 172, 172 (1990) (demonstrating the use of patent counts weighted by citations as evidence of patent value);
39. cf. Lanjouw & Schankerman, Characteristics, supra note 4, at 130 (finding that citations received predicted litigation when those citations were made by competitors).
40. Hall, Jaffee & Trajtenberg, supra note 26, at 30-31.
41. The method of adjustment to account for the different ages of patents involves placing each patent in the data set into a cohort of other patents in the data set that were issued during the same year. Thus, each cohort is one year, although cohorts of more than one year could be used if necessary even though that would decrease precision somewhat. The number of forward citations received by each patent is divided by the average number of forward citations received by other patents in the same cohort. This gives us the adjusted number of forward citations for that patent in the data set. The process is repeated for every other patent in the same cohort and then repeated for each patent in the other year cohorts. To obtain the adjusted number of forward citations for an entire data set, we then averaged the quantity of adjusted number of forward citations received by all patents in the set The method is from Bronwyn H. Hall, Adam B. Jaffe & Manuel Trajtenberg, The NBER Patent-Citations Data File: Lessons, Insights, and Methodological Tools, in PATENTS, CITATIONS, & INNOVATIONS: A WINDOW ON THE KNOWLEDGE ECONOMY403, 434-37 (Adam B. Jaffe & Manuel Trajtenberg eds., 2002).Because of the unusual skew in forward citations, the means for untransformed forward citations look nearly identical even though the differences in the distributions are both dramatic and highly significant As a result, we also report in Table 2 the logtransformed value for adjusted forward citations, which makes the differences quite clear.
42. Self-citations are references to the patent as prior art in subsequent patents issued to the same inventor or assignee. Self-citations have been identified as an independent indicator of private patent value apart from overall forward citations of which the self-citations are a part. See, e.g., Hall et al., supra note 26, at 31-33 (finding selfcitation to be a highly significant indicator of market value). The apparent reason is that self-citations provide evidence that the patent owner is building a portfolio of patents on related technologies, and a portfolio of patents often has a value that is greater than the sum of its parts.
43. See, e.g., Gideon Parchomovsky & R. Polk Wagner, Patent Portfolios, 154 U. PA. L. REV. 1, 5-6 (2005). Because there can be multiple inventors on a patent, and because ownership of a patent can sometimes change after issuance, there is occasionally difficulty in identifying a particular forward citation as a selfcitation. Thus, we used the following decision rule for identifying self-citations: a forward citation is a self-citation if either (a) the owners of the main patent and the forward citation are the same, or (b) the owners are different, the inventors in the main patent and the forward citation are the same, and there are no co-inventors (i.e., no other inventors). To apply this decision rule, we had to examine the front pages of each of the 3419 patents that constituted forward citations to patents in our two data sets. Economists do not examine individual forward citations and thus use a blunter test to identify self-citations-a forward citation is a self-citation when the assignee (owner) of the patent is the same in the instant patent and the forward citation.
44. See, e.g., Hall et al., supra note 28, at 424. This approach clearly does not capture any nuance, but it has the advantage of being automatizable.
45. Allison et al., supra note 4, at 449 n.58 (noting that the number of claims in a patent can depend on a number of factors correlated to value);
46. see also john R. Allison & Emerson H. Tiller, The Business Method Patent Myth, 18 BERKELEY TECH. L.J. 987, 1052-56 (2003) (reviewing the literature on numbers of claims as an indicator of patent value);
47. Lanjouw & Schankerman, Characteristics, supra note 4, at 140-42 (discussing patent claims as an underutilized indicator of patent value);
48. Kimberly A. Moore, Xenophobia in American Courts, 97 Nw. U. L. REV. 1497, 1544 (2003) (noting that, while the number of claims has been shown to be an indicator of private value, the reason is not that more claims cause the patent to have greater breadth (or scope) as economists have asserted, as it is the generality of claim language that creates breadth);
49. cf. john R. Allison & Mark A. Lemley, The Growing Complexity of the United States Patent System, 82 B.U. L. REV. 77, 104 (2002) [hereinafter Allison & Lemley, Growing Complexity] ("[T]he number of claims could also reflect resource constraints, drafting style, uncertainty about the law or the significance of an invention, or a host of other factors that are not necessarily driven by patent value.").
50. See, e.g., Allison & Tiller, supra note 30, at 1036-39 (arguing that there is a correlation between the number of prior art references and patent value);
51. Allison et al., supra note 4, at 453 (finding a statistically significant relationship between prior art references and litigation);
52. Harhoff et al., supra note 4, at 1360 (finding a relationship between prior art references cited and other measures of patent value). But see Lanjouw & Schankerman, Characteristics, supra note 4, at 41 ("[A] litigated patent is likely to cite fewer prior patents per claim than a randomly selected patent."). The theory behind the relationship of prior art references and value is that the more citations that are considered during prosecution by the examiner, the less likely it is that some prior art exists that will invalidate the patent The more prior art considered, in other words, the more likely a patent is to survive subsequent litigation.
53. See, e.g., John R. Allison & Mark A. Lemley, Empirical Evidence on the Validity of Litigated Patents, 26 AIPLA QJ. 185, 232-34 (1998) (showing that courts are more likely to invalidate patents on the basis of uncited prior art than on prior art cited to the PTO);
54. Moore, supra note 30, at 1538 ("Patents that include more citations or more diverse citations are more likely to be valid."). Because lawyers know this, the value relationship may reflect not only the strength of patents that cite a lot of prior art, but also efforts by applicants to "bulletproof patents they expect to litigate by citing a great deal of art.
55. Allison et al., supra note 4, at 453-55. Because many of the patents in this sample were issued before January 1, 2001, when the PTO began identifying examineradded prior art on the face of the patent, we were unable to determine whether it was the applicant or the examiner that provided most of this art However, other work has shown that virtually all nonpatent prior art (over ninety percent) is provided by applicants, not examiners.
56. See Mark Lemley & Bhaven N. Sampat, Examiner Characteristics and the Patent Grant Rate 10 (John M. Olin Program in Law & Econ., Working Paper No. 369, 2009). Thus, it is quite likely that the disparity in nonpatent prior art citations is a result of applicant submissions, not examiner diligence.
57. See, e.g., Allison & Lemley, supra note 16, at 2113-15, 2148 tbl. 1 (showing the distribution of studied patents among fourteen technology categories).
58. See, e.g., Allison & Lemley, Growing Complexity, supra note 30, at 93 & tbl. 1, 94 (documenting the growing diversity of patented technologies).
59. Allison et al., supra note 4, at 472.
60. A reminder-inventions can involve more than one industry or technology class, so the proportions exceed 100%. We simply compared proportions between the most-litigated and once-litigated data sets.
61. Allison et al., supra note 4, at 472.
62. We found that 44 of the 106 most-litigated patents and 31 of the 106 oncelitigated patents were sold before litigation. The difference was not statistically significant, however, so we rely only on the aggregate numbers.
63. Katz had 20 of the top 106 most-litigated patents issued in his name. In fact, his impact on the most-litigated patents is even greater, because at litigation he owned still other patents acquired after issue.
64. Cf. Gwendolyn G. Ball & Jay P. Kesan, Transaction Costs and Trolls: Strategic Behavior by Individual Inventors, Small Firms, and Entrepreneurs in Patent Litigation 14 (Univ. 111. Law & Econ.Paper Series, Research Paper Series No. LE09-005, 2009), available at http://papers.ssm.com/sol3/papers.cfm?abstract-id=1337166 (finding that small firms are frequent patent litigants).
65. This is consistent with Judge Kimberly Moore's finding that assignments of litigated patents are quite common. See Kimberly A. Moore, Populism and Patents, 82 N.Y.U. L. REV. 69, 92-93 (2007) ;
66. see also Carlos J. Serrano, The Dynamics of the Transfer and Renewal of Patents 3 (Nat'l Bureau of Econ. Research, Working Paper No. 13938, 2008) (finding that the likelihood that a patent will be transferred increases with indicia of value).
67. An alternative explanation, however, is that once a company pays money to buy a patent, it is more likely to enforce that patent since it wants a return on its investment. Similarly, the correlation between value and assignment may reflect the nature of the acquiring entities, at least for a given subset of cases. An entity purchasing thirdparty patents to assert will presumably do precisely that. Moreover, at least some patent holders may create special-purpose entities immediately prior to, and for, litigation. On the other hand, one must assume that an acquirer will evaluate the likely value of patents before buying them and will not knowingly purchase patents that are either of highly dubious validity or of such narrow scope as to be of little use.
68. See Mark A. Lemley, Are Universities Patent Trolls?, 18 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 611, 612 (2008) (arguing that universities, although nonpracticing patent owners, are not necessarily patent trolls).
69. The differences we report are statistically significant at a 95% or greater confidence level for class 1, and at a 99% confidence level for classes 5 and 8.
70. See Stanford IP Litigation Clearinghouse, http://lexmachina. stanford.edu (last visited Oct 15, 2009) (reporting that 13, 793 total patent suits were filed between 2003 and 2007 inclusive). A list of the most-litigated patents and the number of suits in the database is attached as the Appendix.
71. See infra Appendix.
72. This may overstate the role played by NPEs, however, because many of those suits involve more than one patent in our data set.
73. Technology-area regression p = 0.0303; industry-area regression p = 0.0359.
74. Technology-area regression p = 0.0223; industry-area regression p = 0.0061.
75. Technology-area regression p = 0.0282; industry-area regression p = 0.0062.
76. The full results of the logistic regressions are available from the authors on request
77. Many patent characteristics are necessarily correlated rather highly, such as the number of claims, various types of prior art references, number of U.S. applications leading to a particular patent, and number of forward citations. This correlation has a common cause: applicants for certain patents perceive in advance that the patents likely will have value to them, that they will be important, and even that they are more likely to be litigated, and this perception often leads the applicants to make such patents stronger and broader by drafting more claims, finding and citing more prior art, and actively crafting the patent using continuation applications. Their actions, in turn, tend to create private value for the patents. Applicants do not have the same degree of control over the number of forward citations as they do over other patent metrics, except for self-citations in their own later patents (which, as previously explained, is itself an independent indicator of value). However, these are patents that the owners are more likely to assert through litigation or licensing, and they get more attention, leading to more forward citations by others in later patents. As a consequence, the number of forward citations is also correlated with the internal patent characteristics noted above.
78. Allison et al., supra note 4, at 439-43.
79. This is not to suggest, however, that the objective measures of value accurately capture all, or even most, of the value of patents; they are necessarily imperfect indicators of that value. 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) ("[T] he misuse of patent metrics has both fostered dire predictions and created unrealistic expectations about the capacity of patent data to guide policy."). But the fact that they are not. perfect predictors does not render them worthless.
80. See Allison & Sager, supra note 4, at 1794 (concluding that, despite some imperfections, patent characteristics can be useful in identifying valuable patents).
81. But see Lemley & Moore, supra note 24, at 71-83 (arguing that continuations are mostly unnecessary and do more harm than good).
82. On the difference, see Allison et al., supra note 4, at 439-40.
83. Burk and Lemley argue that modern claim construction can systematically disadvantage patentees since even one error in claim drafting or one lost claimconstruction fight may result in a finding of either invalidity or noninfringement Dan L. Burk & Mark A. Lemley, Fence Posts or Sign Posts? Rethinking Patent Claim Construction, 157 U. PA. L. REV. 1743, 1763 (2009). Drafting more claims hedges against this risk by giving the patentee multiple shots at an error-free claim. Continuations also hedge against this risk by allowing patentees to rewrite their claims after the fact
84. See Allison & Lemley, supra note 31, at 231-34 (finding that it is much harder to invalidate patents based on art that was cited to the PTO, leading to the possibility of "bulletproofing" a patent application by including as much prior art as possible). Relatedly, a sea of citations may actually diminish the PTO's ability to analyze invalidity with respect to the most salient prior art included therein, since examiners operate under severe time constraints and are not given more time to examine applications just because they include more prior art
85. We test this hypothesis in an upcoming companion paper that investigates the outcomes of the most-litigated-patent cases. See supra note 14 and accompanying text
86. Lemley, supra note 5, at 1508-10.
87. See, e.g., Joseph Farrell & Carl Shapiro, How Strong Are Weak Patents?, 98 AM. ECON. REV. 1347, 1362 (2008) (noting that "weak" patents of questionable validity are very costly to society and urging a "targeted application of resources" to improve the patent review process);
88. Joseph Farrell & Robert P. Merges, Incentives to Challenge and Defend Patents: Why Litigation Won't Reliably Fix Patent Office Errors and Why Administrative Patent Review Might Help, 19 BERKELEY TECH. L.J. 943, 960-64 (2004) (arguing that litigation does not efficiently correct PTO mistakes and urging better funding and improved processes for patent review).
89. See, e.g., Mark D. Janis, Rethinking Reexamination: Toward a Viable Administrative Revocation System for U.S. Patent Law, 11 HARV. J.L. & TECH. 1, 118-21 (1997) (recommending the adoption of a post-grant opposition mechanism incorporating lessons learned from U.S. and European experience);
90. Robert P. Merges, As Many as Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform, 14 BERKELEY TECH. L.J. 577, 614 (1999) (advocating a "coherent efficient opposition procedure");
91. Craig Allen Nard, Certainty, Fence Building and the Useful Arts, 74 IND. L.J. 759, 764 (1999) (proposing "the implementation of a post-grant opposition proceeding");
92. J.H. Reichman, From Free Riders to Fair Followers: Global Competition Under the TRIPS Agreement, 29 N.Y.U.J. INT'L L. & POL. 11, 31 (1997) ("A national opposition system seems uniformly advisable.");
93. John R. Thomas, Collusion and Collective Action in the Patent System: A Proposal for Patent Bounties, 2001 U. ILL. L. REV. 305, 326-30 ("The general acknowledgment of the wisdom of oppositions is striking.")
94. See Doug Lichtman & Mark A. Lemley, Rethinking Patent Law's Presumption of Validity, 60 STAN. L. REV. 45, 61-63 (2007) (arguing in support of an additional, optional "gold-plated" patent-review process);
95. Mark Lemley, Doug Lichtman & Bhaven Sampat, What to Do About Bad Patents, REG., Winter 2005-2006, at 10, 10-12 (advocating for a twotiered patent review system). While one might view applicant bulletproofing as a form of "gold plating," the latter term refers to a proposal that the PTO devote additional resources to examine certain applications and award a patent that has correspondingly higher deference.
96. See Beth Simone Noveck, "Peer to Patent": Collective Intelligence, Open Review, and Patent Reform, 20 HARV. J.L. & TECH. 123, 143-51 (2006) (supporting an open peerreview framework for patent reviews).
97. See Allison & Sager, supra note 4, at 1787-88 (describing how prospective investors may sift through a set of patents looking for ones of value).
98. See BURK & LEMLEY, supra note 1, at 49-66 (documenting industry differences in innovation and patenting); .
99. see also Josh Lerner, The Litigation of Financial Innovations 2 (Nat'l Bureau of Econ. Research, Working Paper No. 14324, 2008) (finding that financial-services patents are litigated 27 to 39 times more than ordinary patents).
100. See supra note 2 and accompanying text.
101. Colleen Chien studied the percentage of all suits filed by trolls. In defining only a subset of nonpracticing entities as trolls and excluding lawsuits by individuals, she found that while 17% of suits were filed by trolls, those suits accounted for 28% of the defendants in patent cases-a share that has continued to rise over time. Colleen V. Chien, Of Trolls, Davids, Goliaths, and Kings: Narratives and Evidence in the Litigation of High-Tech Patents, 87 N.C. L. REV. 1571, 1608-11 (2009).
102. See Lemley, supra note 43, at 630-31 (arguing that identifying trolls is not as important as preventing abuses of the patent system by any plaintiff).
103. See eBay Inc. v. MercExchange L.L.C., 547 U.S. 388, 393-94 (2006) (requiring proof of irreparable injury before granting injunctions in patent cases).
104. See, e.g., BIC Leisure Prods., Inc. v. Windsurfing Int'l, Inc., 1 F.3d 1214, 1219, 1223 (Fed. Cir. 1993) (reversing an award of lost profits because the patentee and the infringer did not compete);
105. cf. Del Mar Avionics, Inc. v. Quinton Instrument Co., 836 F.2d 1320, 1326 (Fed. Cir. 1987) (describing the "general rule" that patentees producing the patented item are entitled to lost-profits damages);
106. John E. Dubiansky, An Analysis for the Valuation of Venture Capital-Funded Startup Firm Patents, 12 B.U.J. SCI. & TECH. L. 170, 177 (2006) ("In the licensing context, however, the patent owner is not engaged in an enterprise which utilizes the patent. Consequentially, the owner has no profits to have lost, and is only eligible to receive a reasonable royalty.").
107. The count here is based on the number of lawsuits in which the patent could be verified using IPLC data. Because some lawsuits, including some of Katz's lawsuits, involve more than one patent, the actual number of cases filed by the plaintiffs identified here is smaller than simply adding the number of suits for each patent would suggest