To Promote Innovation

  • 4,264 views
Uploaded on

To Promote Innovation: …

To Promote Innovation:
The Proper Balance of
Competition and Patent Law and Policy
A Report by the Federal Trade Commission
FEDERAL TRADE COMMISSION
October 2003

More in: Education
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
No Downloads

Views

Total Views
4,264
On Slideshare
0
From Embeds
0
Number of Embeds
1

Actions

Shares
Downloads
0
Comments
0
Likes
5

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. FEDERAL TRADE COMMISSION To Promote Innovation: The Proper Balance ofCompetition and Patent Law and Policy A Report by the Federal Trade Commission October 2003
  • 2. FEDERAL TRADE COMMISSIONTIMOTHY J. MURIS ChairmanMOZELLE W. THOMPSON CommissionerORSON SWINDLE CommissionerTHOMAS B. LEARY CommissionerPAMELA JONES HARBOUR CommissionerSusan A. Creighton Director, Bureau of CompetitionJ. Howard Beales III Director, Bureau of Consumer ProtectionLuke Froeb Director, Bureau of EconomicsWilliam E. Kovacic General CounselAnna H. Davis Director, Office of Congressional RelationsRosemarie A. Straight Executive DirectorReport ContributorsSusan S. DeSanti, Deputy General Counsel for Policy Studies, Office of the General CounselWilliam E. Cohen, Assistant General Counsel for Policy Studies, Office of the General CounselGail F. Levine, Deputy Assistant General Counsel for Policy Studies, Office of the General CounselHillary J. Greene, Project Director for Intellectual Property, Office of the General CounselMatthew Bye, Attorney, Policy Studies, Office of General CounselMichael S. Wroblewski, Assistant General Counsel for Policy Studies, Office of the General CounselRobin Moore, formerly Attorney, Policy Studies, Office of General CounselMichael Barnett, formerly Attorney, Policy Studies, Office of General CounselNicole Gorham, Legal Assistant, Policy Studies, Office of General CounselCecile Kohrs, Legal Assistant, Policy Studies, Office of General CounselDavid Scheffman, formerly Director, Bureau of EconomicsMark Frankena, Associate Director for Competition Analysis, Bureau of EconomicsRoy Levy, Economist, Economic Policy Analysis, Bureau of EconomicsAlden F. Abbott, Assistant Director for Policy and Evaluation, Bureau of CompetitionSuzanne Michel, Special Counsel for Intellectual Property, Bureau of CompetitionPaige Pidano, formerly Legal Assistant, Bureau of CompetitionKarina Lubell, formerly Legal Assistant, Bureau of CompetitionInquiries concerning this report should be directed to:Susan S. DeSanti, Deputy General Counsel for Policy Studies, Office of the General Counsel(202) 326-3190 or sdesanti@ftc.gov
  • 3. Acknowledgments: The Commission wishes to note the expertise and time contributed by Hearings participants.For all of their contributions, the Commission conveys its thanks. The Commission thanks the Antitrust Division of the Department of Justice and the Patentand Trademark Office for participating in many of the panels at, and for recommending many ofthe participants in, the Hearings. The Commission thanks the Competition Policy Center and the Berkeley Center for Law andTechnology at the University of California at Berkeley for providing facilities to allow some ofthe Hearings to be held on the West Coast.Cover: Clip art on cover obtained under license from Microsoft Corporation. Patents: Front Cover Patent No. 549,160 - Selden Road Engine Patent No. 4,779, 268 - Frame Decoding for Digital Signal Transmission Patents: Back Cover Patent No. 4,302,281 - Method for Producing Pulp Patent No. 4,805,654 - Sun Shield for Automobiles
  • 4. TO PROMOTE INNOVATION: THE PROPER BALANCE OF COMPETITION AND PATENT LAW AND POLICY EXECUTIVE SUMMARY Innovation benefits consumers the FTC and the Antitrust Division of thethrough the development of new and Department of Justice (DOJ) (forthcoming),improved goods, services, and processes. will discuss and make recommendations forAn economy’s capacity for invention and antitrust to maintain a proper balance withinnovation helps drive its economic growth the patent system.and the degree to which standards of livingincrease.1 Technological breakthroughs Competition and Patent Law andsuch as automobiles, airplanes, the personal Policy Promote Innovation andcomputer, the Internet, television, Benefit the Public.telephones, and modern pharmaceuticalsillustrate the power of innovation to increase Competition through free enterpriseprosperity and improve the quality of our and open markets is the organizing principlelives. for most of the U.S. economy. Competition among firms generally works best to achieve Competition and patents stand out optimum prices, quantity, and quality ofamong the federal policies that influence goods and services for consumers. Antitrustinnovation. Both competition and patent law, codified in the Sherman Act, the FTCpolicy can foster innovation, but each Act, and other statutes, seeks “to maximizerequires a proper balance with the other to consumer welfare by encouraging firms todo so. Errors or systematic biases in how behave competitively.”3one policy’s rules are interpreted and appliedcan harm the other policy’s effectiveness. Competition can stimulateThis report by the Federal Trade innovation. Competition among firms canCommission (FTC) discusses and makes spur the invention of new or better productsrecommendations for the patent system to or more efficient processes. Firms may racemaintain a proper balance with competition to be the first to market an innovativelaw and policy.2 A second joint report, by technology. Companies may invent lower- cost manufacturing processes, thereby 1 increasing their profits and enhancing their Federal Reserve Board Vice Chairman Roger ability to compete. Competition can promptW. Ferguson, Jr., Patent Policy in a Broader Context,Remarks at 2003 Financial Markets Conference of the firms to identify consumers’ unmet needsFederal Reserve Bank of Atlanta (April 5, 2003), at and develop new products or services tohttp://www.federalreserve.gov/boarddocs/speeches/2003/20030407/default.htm. 2 3 The Federal Trade Commission issues reports I PHILLIP E. AREEDA & HERBERT HOVENKAMP,pursuant to Section 6(f) of the Federal Trade Commission ANTITRUST LAW : AN ANALYSIS OF ANTITRUST PRINCIPLESAct, 15 U.S.C. § 46(f). AND THEIR APPLICATION ¶100a at 4 (2000). 1
  • 5. satisfy them. and technical information is part of the consideration that the inventor gives the Patent policy also can stimulate public.6innovation. The U.S. Constitutionauthorizes Congress “[t]o promote the Competition and Patents Must WorkProgress of Science and useful Arts, by Together in the Proper Balance.securing for limited Times to . . . Inventorsthe exclusive Right to their respective . . . Competition and patents are notDiscoveries.”4 To obtain a patent, an inherently in conflict. Patent and antitrustinvention (that is, a product, process, law “are actually complementary, as bothmachine, or composition of matter) must be are aimed at encouraging innovation,novel, nonobvious, and useful. Moreover, a industry, and competition.”7 Patent lawpatentee must clearly disclose the invention. plays an important role in the property rightsA patent confers a right to exclude others regime essential to a well-functioningfrom making, using, or selling in the United competitive economy. For example, firmsStates the invention claimed by the patent may compete to obtain the property rightsfor twenty years from the date of filing the that patents convey. Patents do notpatent application. necessarily confer monopoly power on their holders,8 and most business conduct with This property right can enable firms respect to patents does not unreasonablyto increase their expected profits from restrain or serve to monopolize markets.investments in research and development, Even when a patent does confer monopolythus fostering innovation that would not power, that alone does not create an antitrustoccur but for the prospect of a patent. violation. Antitrust law recognizes that aBecause the patent system requires public patent’s creation of monopoly power can bedisclosure, it can promote a dissemination ofscientific and technical information thatwould not occur but for the prospect of a placing in their hands a means through the use of whichpatent. their wants may be supplied.” 1 WILLIAM ROBINSON, THE LAW OF PATENTS FOR USEFUL INVENTIONS § 22 at 305 (1890), cited in ROBERT P. MERGES & JOHN F. DUFFY , Like competition policy, patent PATENT LAW AND POLICY: CASES AND MATERIALS 361 (3dpolicy serves to benefit the public. “The ed. 2002).basic quid pro quo contemplated by the 6 See James E. Rogan, Prepared Remarks ofConstitution and the Congress for granting a James E. Rogan, Under Secretary of Commerce forpatent monopoly is the benefit derived by Intellectual Property and Director of the United Statesthe public from an invention with substantial Patent and Trademark Office (2/6/02) 2, at http://www.ftc.gov/opp/intellect/rogan.htm.utility.”5 The public disclosure of scientific 7 Atari Games Corp. v. Nintendo of Am., 897 F.2d 1572, 1576 (Fed. Cir.1990). 4 U.S. CONST. art. I, § 8. Other sections of thisconstitutional provision authorize copyright law. 8 ROBERT L. HARMON, PATENTS AND THE FEDERAL CIRCUIT § 1.4(b) at 21 (5th ed. 2001) (“Patent 5 Brenner v. Manson, 383 U.S. 519, 534-35 rights are not legal monopolies in the antitrust sense of the(1966). The consideration an inventor gives in return for a word. Not every patent is a monopoly, and not everypatent “is the benefit which he confers upon the public by patent confers market power.”). 2
  • 6. necessary to achieve a greater gain for competition that might have developedconsumers. based on the obvious technology. See Box 1. Conversely, competition policy can Analogously, the SupremeCourt has recognized theimportance of competition to the Bo x 1. An Inv alid Pa tent on an Obv ious Inven tion Can H arm 9 Com petition .patent system. “[F]reecompetition” is “the baseline” on In 1895, George Selden obtained a U.S. patent with a claim so broadwhich “the patent system’s that “it literally encompasse[d] most automobiles ever made.” Yetincentive to creative effort the basic invention covered by that claim – putting a gasoline engine 10depends.” By limiting the on a chassis to make a car – was so obvious that many p eop le worldwide thought of it independently as soon as the most primitiveduration of a patent, “[t]he Patent gasoline engines were developed. The association that licensed theClause itself reflects a balance Selden patent collected hundreds of thousands of dollars in royaltiesbetween the need to encourage – raising costs and reducing the output of automobiles – beforeinnovation and the avoidance of Henry Fo rd and others cha llenged the patent, and the patent claimmonopolies which stifle was judicially narrowed in 1 911 . See M ERGES & D UFFY , P A T EN T L A W A N D P OLICY : C A S ES A N D M ATERIALS at 644-46.competition without anyconcomitant advance in the‘Progress of Science and usefulArts.’”11 The patentability requirements for undermine the innovation that the patentnovelty and nonobviousness “are grounded system promotes if overzealous antitrustin the notion that concepts within the public enforcement restricts the procompetitive usegrasp, or those so obvious that they readily of a valid patent. See Box 2.could be, are the tools of creation availableto all.”12 The FTC/DOJ Hearings Examined the Balance of Competition and A failure to strike the appropriate Patent Law and Policy.balance between competition and patent lawand policy can harm innovation. For To examine the current balance ofexample, if patent law were to allow patents competition and patent law and policy, theon “obvious” inventions, it could thwart FTC and the DOJ held Hearings from February through November 2002. The Hearings took place over 24 days, and 9 See Bonito Boats, Inc. v. Thunder Craft Boats, involved more than 300 panelists, includingInc., 489 U.S. 141, 146 (1989) (federal patent laws embody business representatives from large and“a careful balance between the need to promote innovation small firms, and the independent inventorand the recognition that imitation and refinement throughimitation are both necessary to invention itself and the very community; leading patent and antitrustlifeblood of a competitive economy.”). organizations; leading antitrust and patent practitioners; and leading scholars in 10 Id. at 156. 11 Id. at 146. 12 Id. at 156. 3
  • 7. Bo x 2. Overzealous Antitrust Enforcement Can Commission’s conclusions about and Underm ine the Innovation that Patents Promo te. recommendations for the patent system. In the 1970s, antitrust enforcers viewed CONCLUSIONS AND grantbacks (e.g., when a licensee has improved patented technology, it “grants back” to the RECOMMENDATIONS original patentee access to the improvement) as automatically illegal. Mo re recently, antitrust I. Although Most of the Patent enforcers recognize that “[g]rantbacks can have procompetitive effects,” for example, by System Works Well, Some encouraging a patentee to license its patent in the Modifications Are Needed to first place, thereb y enab ling the licensee’s improvement. Antitrust enforcers now evalua te Maintain A Proper Balance likely pro com petitive and anticompetitive effects of Competition and Patent of grantbacks. Past antitrust rules may have Law and Policy. deterred some p rocom petitive grantbacks, however, thus deterring some innovations using paten ted technolo gy. See U.S. Department of The patent system does, for the most Justice and Federal T rade C omm ission, Antitrust part, achieve a proper balance with Guidelines for the Licensing of Intellectual competition policy. The statutory standards Pro perty § 5.6 (A pr. 6, 1 995 ), reprinted in 4 Trade of patentability appear largely compatible Reg. Rep. (CCH) ¶ 13,132, available at http://www.usdoj.gov/atr/public/guidelines/ipguide with competition; properly interpreted, they .htm. tend to award patents only when necessary to provide incentives for inventions, their commercial development, or their disclosure. Congress has enacted neweconomics and antitrust and patent law.13 In statutes that protect competition by, amongaddition, the FTC received about 100 written other things, facilitating disclosures ofsubmissions. Business representatives were patent applications. The Court of Appealsmostly from high-tech industries: for the Federal Circuit, the sole court forpharmaceuticals, biotechnology, computer most patent law appeals, has broughthardware and software, and the Internet.14 stability and increased predictability toThis report discusses Hearings testimony various elements of patent law. This hasand independent research, and explains the reduced legal uncertainty and facilitated business planning. The Patent and 13 Trademark Office (PTO) has implemented The Commission thanks the DOJ and thePatent and Trademark Office for participating in many of initiatives to deal with new types of patentsthe panels at the Hearings and for recommending many of and has released a Strategic Plan for the 21stthe participants in the Hearings. For providing facilities to Century to improve patent quality (i.e.,allow some of the Hearings to be held on the West Coast,the Commission thanks the Competition Policy Center and reduce errors) and streamline procedures.15the Berkeley Center for Law and Technology at the Hearings participants found much to praiseUniversity of California at Berkeley. in the current patent system. The Commission wishes to note the expertise andtime contributed by Hearings participants. For all of theircontributions, the Commission conveys its thanks. 15 See United States Patent and Trademark Office, The 21st Century Strategic Plan, at 14 See Appendices A and B. www.uspto.gov/web/offices/com/strat21/index.htm. 4
  • 8. A. Questionable Patents Can Deter or Nonetheless, many participants in Raise the Costs of Innovation.and observers of the patent system expressedsignificant concerns that, in some ways, the One firm’s questionable patent maypatent system is out of balance with lead its competitor to forgo R&D in thecompetition policy. Poor patent quality and areas that the patent improperly covers. Forlegal standards and procedures that example, firms in the biotech industryinadvertently may have anticompetitive reported that they avoid infringingeffects can cause unwarranted market power questionable patents and therefore willand can unjustifiably increase costs. Such refrain from entering or continuing with aeffects can hamper competition that particular field of research that such patentsotherwise would stimulate innovation.This report makes several Bo x 3. Bloc king Patentsrecommendations for the legal standards,procedures, and institutions of the patent The patents of others can block a patentee ’s abilitysystem to address such concerns. to exploit its own invention. For example: “[S]uppose that Admiral Motors obtains a patent on anII. Questionable Patents Are a internal combustion engine for use in automobiles. Later, Significant Competitive Betty Beta purchases an automobile marketed by Admiral Mo tors that embodies the patented invention. B eta Concern and Can Harm experiments with her new car and develops a dramatica lly Innovation. improved fuel injector useable only in the patented Admiral Mo tors engine. Even if Beta patents her improved fuel injector, she cannot practice that A poor quality or questionable technology without infringing A lpha’s b asic pa tent. . . .patent is one that is likely invalid or Unless one of the parties licenses the other, Beta mustcontains claims that are likely overly wait until Admiral Mo tors’ patent expires beforebroad. Hearings participants raised practicing her own patented improvement invention.”concerns about the number of ROGER E. SCHECHTE R & JO H N R. T H O M AS, I NTELLECTUAL P ROPERTY : T HE L A W OF C OPYRIGHTS ,questionable patents issued.16 Such P ATENTS A N D TR A D EMARKS § 20.1.1 at 462 (20 03). Ifpatents can block competition, see Box 3, the blocking patent is invalid or overbroad, then noand harm innovation in several ways. public benefits exist to justify its effects on follow-on innovation. appear to cover.17 Such effects deter market entry and follow-on innovation by 16 For example, software firms raised concerns 17about patents that they believed should not have been See, e.g., FTC/DOJ Hearings on Competitiongranted, because the inventions were obvious based on and Intellectual Property Law and Policy in thepreceding work in the area. While praising patents as the Knowledge-Based Economy, David J. Earp Testimony Feb.basis for their industry, biotech firms also raised concerns 26, 2002, at pages 290-91, 238 (hereinafter, citations tothat some overbroad patents may discourage further transcripts of these Hearings state the speaker’s last name,innovation in some biotech areas. See generally Chs. 2 and the date of testimony, and relevant page(s)); Blackburn3. 2/26 at 296; Caulfield 3/19 at 161. 5
  • 9. competitors and increase the potential for In some industries, such as computerthe holder of a questionable patent to hardware and software, firms can requiresuppress competition. access to dozens, hundreds, or even thousands of patents to produce just one If a competitor chooses to pursue commercial product. One industryR&D in the area improperly covered by the representative from a computer hardwarequestionable patent without a license to that firm reported that more than “90,000 patentspatent, it risks expensive and time- generally related to microprocessors are heldconsuming litigation with the patent holder. by more than 10,000 parties.”19 Many ofIf the competitor chooses to negotiate a these patents overlap, with each patentlicense to and pay royalties on the blocking several others. This tends to createquestionable patent, the costs of follow-on a “patent thicket” – that is, a “dense web ofinnovation and commercial development overlapping intellectual property rights thatincrease due to unjustified royalties. a company must hack its way through in order to actually commercialize new Another option is to find a legal technology.”20means to invalidate the patent. PTOprocedures allow only very limited Much of this thicket of overlappingparticipation by third parties, however. A patent rights results from the nature of thelawsuit in federal court may not be an technology; computer hardware andalternative, because a competitor may not software contain an incredibly large numbersue to challenge patent validity unless the of incremental innovations. Moreover, aspatent holder has threatened the competitor more and more patents issue on incrementalwith litigation. If the competitor is not on inventions, firms seek more and morethe verge of marketing an infringing patents to have enough bargaining chips toproduct, the patent holder may have no obtain access to others’ overlappingreason to threaten litigation. In these patents.21 One panelist asserted that the timecircumstances, as one biotech representative and money his software company spends oncomplained, “there are these bad patents that creating and filing these so-called defensivesit out there and you can’t touch them.”18 If patents, which “have no . . . innovative valuelitigation does take place, it typically costs in and of themselves,” could have beenmillions of dollars and takes years to better spent on developing newresolve. This wastes resources.B. In Industries with Incremental 19 Detkin 2/28 at 667-68. Innovation, Questionable Patents Can Increase “Defensive 20 Carl Shapiro, Navigating the Patent Thicket: Patenting” and Licensing Cross Licenses, Patent Pools, and Standard-Setting, in 1 Complications. INNOVATION POLICY AND THE ECONOMY 119, 120 (Adam Jaffe et al. eds., 2001). 21 The forthcoming FTC/DOJ joint report will discuss the proper antitrust evaluation of licensing 18 Blackburn 2/26 at 295-96. techniques used in such situations. 6
  • 10. technologies.22 economic significance, “it is much cheaper for society to make detailed [patent] validity Questionable patents contribute to determinations in those few cases [in whichthe patent thicket. In the context of a patent patents are challenged] than to investthicket, questionable patents can introduce additional resources examining patents thatnew kinds of licensing difficulties, such as will never be heard from again.”24royalties stacked one on top of another, and Accordingly, the FTC’s recommendationscan increase uncertainty about the patent focus first on procedures and presumptionslandscape, thus complicating business used in challenging questionable patents,planning. Questionable patents in patent because such challenges are more likely tothickets can frustrate competition by current involve patents of competitive significance.manufacturers as well as potential entrants.Because a manufacturer needs a license to Recommendation 1:all of the patents that cover its product, firmscan use questionable patents to extract high As the PTO Recommends, Enactroyalties or to threaten litigation.23 For Legislation to Create A Newexample, a questionable patent that claims a Administrative Procedure to Allowsingle routine in a software program may be Post-Grant Review of andasserted to hold up production of the entire Opposition to Patents.software program. This process can deterfollow-on innovation and unjustifiably raise The PTO discusses patentcosts to businesses and, ultimately, to applications only with the patent applicant.consumers. Until recently, third parties could only bring certain relevant documents to the attentionC. Recommendations to Improve of, and, in limited circumstances, file a Patent Quality and Minimize written protest with, an examiner or to Anticompetitive Costs of the request the PTO Director to reexamine a Patent System. patent. To address this situation, Congress passed legislation to establish limited One recent article argues procedures that allow third parties topersuasively that because most patent participate in patent reexaminations. Recentapplications involve claims of little amendments have improved those procedures, but they still contain important restrictions and disincentives for their use. 22 Greenhall 2/27 at 377, 420. Once a questionable patent has issued, the most effective way to challenge it is through 23 “Large and small companies are increasingly litigation. Litigation generally is extremelybeing subjected to litigation (or its threat) on the basis ofquestionable patents.” United States Patent andTrademark Office Fee Modernization Act of 2003:Hearing Before the Subcomm. on Courts, the Internet, andIntellectual Property of the House Comm. on the Judiciary,108th Cong. 2 (2003) (Statement of Michael K. Kirk,Executive Director, American Intellectual Property LawAssociation), available at 24http://www.aipla.org/html/Legislative/108/testimony/FeeLe Mark A. Lemley, Rational Ignorance at theg.htm. Patent Office, 95 NW. L. REV. 1495, 1497 (2001). 7
  • 11. costly and lengthy,25 and is not an option should preside over the proceeding, whichunless the patent owner has threatened the should allow cross-examination andpotential challenger with patent carefully circumscribed discovery, andinfringement litigation. which should be subject to a time limit and the use of appropriate sanctions authority. The existing procedures attempt to Limitations should be established to protectbalance two perspectives. On the one hand, against undue delay in requesting post-grantthird parties in the same field as a patent review and against harassment throughapplicant may have the best information and multiple petitions for review. Theexpertise with which to assist in the authorizing legislation should include aevaluation of a patent application, and delegation of authority permitting the PTO’stherefore might be useful participants in the conclusions of law to receive deference fromprocess of deciding whether to grant a the appellate court. Finally, as is the casepatent. On the other hand, the limited with settlements of patent interferences,involvement of third parties in the issuance settlement agreements resolving post-grantand reexamination of patents reflects proceedings should be filed with the PTOgenuine concern to protect patent applicants and, upon request, made available to otherfrom harassment by competitors. This government agencies.remains an important goal. To continue toprotect against the possibility of competitors Recommendation 2:harrassing patent applicants, any newprocedure should be available only after a Enact Legislation to Specify thatpatent issues. Challenges to the Validity of a Patent Are To Be Determined Because existing means for Based on a “Preponderance of thechallenging questionable patents are Evidence.”inadequate, we recommend anadministrative procedure for post-grant An issued patent is presumed valid.review and opposition that allows for Courts require a firm that challenges a patentmeaningful challenges to patent validity to prove its invalidity by “clear andshort of federal court litigation. To be convincing evidence.” This standardmeaningful, the post-grant review should be appears unjustified. A plethora ofallowed to address important patentability presumptions and procedures tip the scalesissues.26 The review petitioner should be in favor of the ultimate issuance of a patent,required to make a suitable threshold once an application is filed. In addition, asshowing. An administrative patent judge many have noted, the PTO is underfunded, and PTO patent examiners all too often do not have sufficient time to evaluate patent 25 A biotechnology case, for example, can cost applications fully. These circumstancesbetween five and seven million dollars and take two or suggest that an overly strong presumption ofthree years to litigate. See Ch. 3. a patent’s validity is inappropriate. Rather, 26 courts should require only a “preponderance At a minimum, patent challengers should be of the evidence” to rebut the presumption ofable to raise issues of novelty, nonobviousness, writtendescription, enablement, and utility. validity. 8
  • 12. The PTO works under a number of facts, the PTO does not have facilities withdisadvantages that can impede its ability to which to test the accuracy or reliability ofreduce the issuance of questionable patents. such information.Perhaps most important, the courts haveinterpreted the patent statute to require the Moreover, presumptions in PTOPTO to grant a patent application unless the rules tend to favor the issuance of a patent.PTO can establish that the claimed invention For example, “[i]f the examiner does notdoes not meet one or more of the produce a prima facie case [of obviousness],patentability criteria. Once an application is the applicant is under no obligation tofiled, the claimed invention is effectively submit evidence of nonobviousness.”29presumed to warrant a patent unless the PTO Similarly, “[o]ffice personnel . . . must treatcan prove otherwise. as true a statement of fact made by an applicant in relation to [the asserted The PTO’s procedures to evaluate usefulness of the invention], unlesspatent applications seem inadequate to countervailing evidence can be provided thathandle this burden. The patent prosecution shows that one of ordinary skill in the artprocess involves only the applicant and the would have a legitimate basis to doubt thePTO. A patent examiner conducts searches credibility of such a statement.”30 Likewise,of the relevant prior art,27 a focal point of the “[t]here is a strong presumption that anexamination process, with only the adequate written description of the claimedapplicant’s submissions for assistance. The invention is present when the application ispatent applicant has a duty of candor to the filed.”31PTO, but that duty does not require anapplicant to search for prior art beyond that The PTO’s resources also appearabout which the applicant already knows.28 inadequate to allow efficient and accurateIf the patent applicant makes assertions or screening of questionable patentfiles documentary evidence regarding certain applications. Patent applications have doubled in the last twelve years and are increasing at about 10% per year.32 With 27 yearly applications approximating 300,000, “Prior art” consists of materials – oftenpatents and publications, although affidavits and testimonyalso may present prior art – that reflect one or more of the 29features or elements of the claimed invention. An MPEP § 2142.invention is “obvious” if it does not represent a sufficientstep beyond the prior art. 30 United States Patent and Trademark Office, Utility Examination Guidelines, 66 Fed. Reg. 1092, 1098- 28 The PTO’s Manual of Patent Examining 99 (2001).Procedure (MPEP) states that the agency “does notinvestigate” duty of disclosure issues and “does not . . . 31 United States Patent and Trademark Office,reject” applications on that basis. See United States Patent Guidelines for Examination of Patent Applications underand Trademark Office, Manual of Patent Examining the 35 U.S.C. 112 ¶ 1, “Written Description” Requirement,Procedure § 2010 (8th edition 2001) (explaining that such 66 Fed. Reg. 1099, 1105 (2001).PTO determinations “would significantly add to theexpense and time involved in obtaining a patent with little 32or no benefit to the patent owner or any other parties with Lerner 2/20 at 157; James Langenfeld,an interest”), available at Innovation, Competition, and Intellectual Property:http://www.uspto.gov/web/offices/pac/mpep/mpep.htm Providing an Economic Framework (2/20/02) (slides) at 6,(hereinafter MPEP). at http://www.ftc.gov/opp/intellect/langenfeld.pdf. 9
  • 13. they arrive at the rate of about 1,000 each grant of a patent application, combined withworking day.33 A corps of some 3,000 the limited resources available to the PTO,examiners must deal with the flood of counsel against requiring “clear andfilings.34 Hearings participants estimated convincing evidence” to overturn thatthat patent examiners have from 8 to 25 presumption. We believe the “clear andhours to read and understand each convincing evidence” burden can undermineapplication, search for prior art, evaluate the ability of the court system to weed outpatentability, communicate with the questionable patents,36 and therefore weapplicant, work out necessary revisions, and recommend that legislation be enacted toreach and write up conclusions. Many found amend the burden to a “preponderance of thethese time constraints troubling.35 Hearings evidence.”participants unanimously held the view thatthe PTO does not receive sufficient funding Recommendation 3:for its responsibilities. Tighten Certain Legal Standards Finally, the PTO grants patents based Used to Evaluate Whether Aonly on the “preponderance of the Patent Is “Obvious.”evidence.” This standard applies in thecontext of an underlying presumption that Patent law precludes patenting if thethe patent should be granted unless the PTO differences between the claimed inventioncan prove otherwise. It does not seem and the prior art37 are such that “the subjectsensible to treat an issued patent as though it matter as a whole would have been obvioushad met some higher standard of at the time the invention was made to apatentability. person having ordinary skill in the art.”38 “Nonobviousness asks whether a Defenders of the application of the development is a significant enough“clear and convincing” evidence standard technical advance to merit the award of aurged that a finding of patent validity by a patent.”39 A proper application of thisneutral government agency using a statutory requirement is crucial to preventknowledgeable examiner justifies placing a the issuance of questionable patents,heavy burden on those who challenge a including trivial patents and patents onpatent’s validity. We disagree. inventions essentially already in the publicPresumptions and procedures that favor the domain. The courts have developed a variety of tests to evaluate the obviousness of a claimed invention. Two in particular – 33 Chambers 2/8 (Patent Law for AntitrustLawyers) at 86 (hereinafter 2/8 (Patent Session)). 36 34 See T.S. Ellis 7/11 at 119-20. Chambers 2/8 (Patent Session) at 84. 37 35 See supra note 25. See, e.g., Dickinson 2/6 at 64-65 (“Patentexaminers need more time to examine.”); Kirschner 2/26 at 38 35 U.S.C. § 103.242-43 (time available “clearly inadequate” for ameaningful examination of a biotech patent application); 39Kesan 4/10 at 100 (time constraints do not allow adequate See MERGES & DUFFY , PATENT LAW ANDsearch for software prior art). POLICY: CASES AND MATERIALS at 644. 10
  • 14. the “commercial success test” and “the sufficiently searching inquiry when theysuggestion test” – require more thoughtful conclude that commercial successapplication to weed out obvious patents. demonstrates a claimed invention is not obvious. Under current standards, if thea. In applying the “commercial patent holder shows that the claimed success” test, 1) evaluate on a case- features of the patent are coextensive with by-case basis whether commercial those of a successful product, then it is success is a valid indicator that the presumed that the invention – rather than claimed invention is not obvious, other factors – caused the commercial and 2) place the burden on the success. The burden shifts to the challenger patent holder to prove the claimed to present evidence to rebut that invention caused the commercial presumption.40 success. This test fails to ask, first, whether The Supreme Court has advised that, factors other than the invention may havein some circumstances, courts may consider caused the commercial success. By contrast,the commercial success of a claimed the PTO properly requires that commercialinvention to indicate that it was not obvious. success be “directly derived from theFor example, in some cases early in the invention claimed” and not the result oftwentieth century, courts found the “business events extraneous to the merits ofcommercial success of an invention that the claimed invention.”41 Second, thesatisfied a long-felt need that had resisted judicial standard too easily shifts the burdenthe efforts of others to solve the problem to the challenger. The patent holder is thetended to show the claimed invention was best source of information on what hasnot obvious. caused the commercial success of its product and should be required to show that, in fact, Commercial success can result from the claimed invention caused themany factors, however, some of which have commercial success.nothing to do with the claimed invention.For example, marketing, advertising, or an b. In applying the “suggestion” test,incumbent’s unique advantages may cause assume an ability to combine orcommercial success. An undue reliance on modify prior art references that iscommercial success to show nonobviousness consistent with the creativity andcan raise a number of competitive concerns. problem-solving skills that in factCommercially successful inventions may be are characteristic of those havingmore likely than others to occur even ordinary skill in the art.without the prospect of a patent. Patents oncommercially successful products are more If the prior art already would havelikely to confer market power than those on suggested the claimed invention, then theless successful products. 40 Certain patent experts and other See HARMON, PATENTS AND THE FEDERAL CIRCUIT at 169-70.Hearings participants expressed concern thatcourts and juries sometimes fail to use a 41 MPEP § 716.03(b). 11
  • 15. claimed invention is obvious. If not, then It is important to protect against thethe claimed invention is not obvious. The issuance of obvious patents that may confer“suggestion test” thus asks a helpful market power and unjustifiably raise costs.question – that is, to what extent would the Requiring concrete suggestions beyondprior art “have suggested to one of ordinary those actually needed by a person withskill in the art that this process should be ordinary skill in the art,45 and failing to givecarried out and would have a reasonable weight to suggestions implicit from the artlikelihood of success.”42 The Federal Circuit as a whole and from the nature of thejustifiably has sought to protect inventors problem to be solved, is likely to result infrom findings of obviousness based purely patents on obvious inventions and is likelyon hindsight. “Good ideas may well appear to be unnecessarily detrimental to‘obvious’ after they have been disclosed, competition. The Federal Circuit’s mostdespite having been previously recent articulations of the suggestion testunrecognized.”43 The Federal Circuit also seem to signal greater appreciation of thesehas sought to ensure that the PTO provides issues and would better facilitatean administrative record susceptible to implementation of the test in ways sensitivejudicial review. to competitive concerns. Hearings participants expressed Recommendation 4:concern, however, with some recentapplications of the suggestion test. To show Provide Adequate Funding for thethat a claimed invention is obvious, some PTO.cases seem to require the PTO to point toparticular items of prior art that concretely Participants in the Hearingssuggest how to combine all of the features of unanimously expressed the view that thea claimed invention. Such an application of PTO lacks the funding necessary to addressthe suggestion test may have found that the issues of patent quality. Presidential patentclaimed invention of the Selden patent – that review committees have long advocatedis, putting a gasoline engine on a carriage – more funding for the PTO to allow it towas not obvious, because there was no improve patent quality. 46 As recently asdocument that suggested that combination. 2002, the Patent Public Advisory CommitteeThe invention likely was obvious, however; stated that the PTO “faces a crisis in funding“[e]verybody seemed to know that if you gota new engine of any kind, you would put it 45on a carriage.”44 Cf. Barr 10/30 at 53-54 (arguing that current obviousness standards fail to reflect the skill of his company’s engineers, who “every day” independently invent things that have been deemed nonobvious). 42 Brown and Williamson Tobacco Corp. v. 46Philip Morris, 229 F.3d 1120, 1124 (Fed. Cir. 2000) E.g., THE ADVISORY COMMISSION ON PATENT(emphasis added). LAW REFORM, REPORT TO THE SECRETARY OF COMMERCE (Aug. 1992), available at 43 http://world.std.com/obi/USG/Patents/overview; REPORT Arkie Lures, Inc. v. Gene Larew Tackle, Inc., OF THE INDUSTRIAL SUBCOMM. FOR PATENT AND119 F.3d 953, 956 (Fed. Cir. 1997). INFORMATION POLICY OF THE ADVISORY COMM. ON INDUSTRIAL INNOVATION, REPORT ON PATENT POLICY 44 Duffy 7/10 at 132-33. (1979). 12
  • 16. that will seriously impact . . . the quality of . examiner requests them. These statements. . issued patents.”47 The FTC strongly could materially enhance examiners’ abilityrecommends that the PTO receive funds to provide quality patent examinations bysufficient to enable it to ensure quality drawing more fully on the patent applicant’spatent review. knowledge base to identify the most relevant portions of prior art references.Recommendation 5: b. Encourage the use of examiner Modify Certain PTO Rules and inquiries under Rule 105 to obtain Implement Portions of the PTO’s more complete information, and 21st Century Strategic Plan. reformulate Rule 105 to permit reasonable follow-up.a. Amend PTO regulations to require that, upon the request of the PTO Rule 105 permits examiners to examiner, applicants submit request “such information as may be statements of relevance regarding reasonably necessary to properly examine or their prior art references. treat the matter [under examination].”50 The Commission recommends that the PTO Some Hearings participants asserted make a concentrated effort to use examinerthat, far from holding back information, inquiries more often and more extensively.patent applicants tend to provide an As one panelist emphasized, “to get betterexaminer with numerous prior art citations, quality and shrink the amount of work,”resulting in lots of “information,” but little there is a need to seek more knowledge in“knowledge.”48 The 2002 version of the the possession of applicants, who typicallyPTO’s 21st Century Strategic Plan proposed “know more about the technology than therequiring applicants that cited more than 20 examiner does, and [know] where you mightprior art references to provide statements to find something that might be relevant.”51 Toexplain the relevance of references, but the be fully effective, however, Rule 105 shouldPTO has now withdrawn that proposal.49 be amended so that applicants who reply thatThe FTC’s proposal is more modest than the they do not know the answer to thePTO’s original proposal; it would require examiner’s inquiry, or that the necessaryrelevance statements only when the information “is not readily available to the party or parties from which it was requested” are not accepted as a complete 47 PATENT PUBLIC ADVISORY COMMITTEE, reply,52 as they are now, but rather areANNUAL REPORT 6 (Nov. 29, 2002), available at treated as responses on which the examinerhttp://www.uspto.gov/web/offices/com/advisory/acrobat/pp may follow up.acannual12-05-02.pdf. 48 E.g., Kesan 10/25 at 60-61. 50 37 C.F.R. § 1.105. 49 United States Patent and Trademark Office21st Century Strategic Plan, Mandatory Information 51 Kushan 4/11 at 89.Disclosure Statements (IDS), P-09 at 3 (June 3, 2002). SeeThe 21st Century Strategic Plan, available at 52www.uspto.gov/web/offices/com/strat21/index.htm. See 37 C.F.R. § 1.105. 13
  • 17. c. Implement the PTO’s Recommendation 6: recommendation in its 21 st Century Strategic Plan that it expand its Consider Possible Harm to “second-pair-of-eyes” review to Competition – Along with Other selected areas. Possible Benefits and Costs – Before Extending the Scope of Second-pair-of-eyes review allows Patentable Subject Matter.the PTO quickly to flag issues that needfurther attention by the examiner or the Section 101 of the Patent Act states,examiner’s supervisor. The PTO first used “Whoever invents or discovers any new andthis method to improve the quality of useful process, machine, manufacture, orbusiness method patents, and it received composition of matter, or any new andgood reviews from participants in the patent useful improvement thereof, may obtain asystem. The Commission believes that patent.”54 Despite this broad mandate,expanding this program to fields with courts have long held certain types ofsubstantial economic importance, such as inventions unpatentable. Traditionalsemiconductors, software, and common law exceptions include phenomenabiotechnology, as well as other new of nature, abstract intellectual concepts,technologies as they emerge, could help to mental steps, mathematical algorithms withboost patent quality in areas where it will no substantial practical application, printedmake the most difference. matter, and, for many years, business methods.d. Continue to implement the recognition that the PTO “forges a Over the past twenty-five years, balance between the public’s however, the scope of patentable subject interest in intellectual property and matter has expanded significantly. For each customer’s interest in his/her example, the Supreme Court, through two patent and trademark.”53 landmark decisions in 1980, held that both man-made, living organisms and computer The PTO functions as a steward of software constitute patentable subject matterthe public interest, not as a servant of patent pursuant to Section 101. In 1999, theapplicants. The PTO must protect the public Federal Circuit ruled that business methodsagainst the issuance of invalid patents that can be patented. Some Hearings participantsadd unnecessary costs and may confer claimed that patents on computer softwaremarket power, just as it should issue valid and business methods are not necessary topatents to encourage invention, disclosure, spur the invention, commercialand commercial development. development, or public disclosure of 53 United States Patent and Trademark Office,FY2002 Corporate Plan 28 (2001) (describing role of PTOUnder Secretary and Director), athttp://www.uspto.gov/web/offices/com/corpplan/fy2002/in 54dex.html. 35 U.S.C. § 101. 14
  • 18. software or business methods.55 Others In addition to questionable patents,disagreed. Some Hearings participants other portions of the patent system raisecontended that software and business competitive concerns. This section brieflymethod patents can raise significant describes each issue and the Commission’scompetitive concerns and deter innovation, recommendation(s) to address it.especially because so much of theinnovation in those fields builds Recommendation 7:incrementally on preceding work. This mayraise the potential for thickets of patents to Enact Legislation to Requirehinder, rather than accelerate, innovation Publication of All Patentand commercial development. Applications 18 Months After Filing. The constitutional intention thatpatents “promote the Progress of Science Until relatively recently, patents wereand useful Arts” should be taken into published only when issued; patentaccount in interpreting the scope of applications were not published. During thepatentable subject matter under Section 101. time that would pass between the filing of aDecisionmakers should ask whether granting patent application and the issuance of apatents on certain subject matter in fact will patent, an applicant’s competitor could havepromote such progress or instead will hinder invested substantially in designing andcompetition that can effectively spur developing a product and bringing it toinnovation. Such consideration is consistent market, only to learn, once the patent finallywith the historical interpretation of issued, that it was infringing a rival’s patentpatentable subject matter, which implicitly and owed significant royalties. Thisrecognizes that granting patent protection to scenario disrupts business planning, and cancertain things, such as phenomena of nature reduce incentives to innovate and discourageand abstract intellectual concepts, would not competition.advance the progress of science and theuseful arts. For future issues, it will be A relatively new statute requires thathighly desirable to consider possible harms most patent applications – all except thoseto competition that spurs innovation – as filed only in the United States – bewell as other possible benefits and costs – published 18 months after filing. Patentbefore extending the scope of patentable applicants are protected from copying ofsubject matter. their inventions by statutory royalty rights, if the patent ultimately issues. This newIII. Other Patent Laws and procedure appears to have increased Procedures Also Raise business certainty and promoted rational planning, as well as reduced the problem of Competitive Concerns. unanticipated “submarine patents” used to hold up competitors for unanticipated royalties. For these reasons, Hearings 55 See generally Ch. 3. See also Robert M. participants advocated expanding the 18-Hunt, You Can Patent That? Are Patents on Computer month publication requirement to includePrograms and Business Methods Good for the Economy?,Q1 BUSINESS REVIEW 5, 14 (2001). patents filed only domestically, because such 15
  • 19. patents may well have competitive proposed remedy for the opportunisticsignificance. Protection from copying broadening of claims should also protectsimilar to that already available for other such legitimate uses. Creating interveningpublished applications should be extended to or prior use rights would most directlythose filing domestic patent applications as achieve this balance; it would cure potentialwell, and any necessary protections for competitive problems without interferingindependent inventors also should be with legitimate needs for continuations.considered in terms of their likely costs and Such rights should shelter inventors andbenefits. users that infringe a patent only because of claim amendments following a continuationRecommendation 8: or other similar application,56 provided that the sheltered products or processes are Enact Legislation to Create developed or used (or the subject of Intervening or Prior User Rights substantial preparation for use) before the to Protect Parties from amended claims are published. Infringement Allegations That Rely on Certain Patent Claims Recommendation 9: First Introduced in a Continuing or Other Similar Application. Enact Legislation to Require, As a Predicate for Liability for Willful After publication of its patent Infringement, Either Actual,application, an applicant may continue to Written Notice of Infringementamend its claims. Through this claim from the Patentee, or Deliberateamendment process, a patent that states Copying of the Patentee’sbroader claims than those published at 18 Invention, Knowing It to Bemonths can still emerge. If the applicant Patented.uses procedures such as continuingapplications to extend the period of patent A court may award up to three timesprosecution, the potential for the amount of damages for a defendant’santicompetitive hold up increases. Indeed, willful infringement of a patent – that is, theseveral panelists asserted that some defendant knew about and infringed theapplicants keep continuing applications patent without a reasonable basis for doingpending for extended periods, monitor so. Some Hearings participants explaineddevelopments in the relevant market, and that they do not read their competitors’then modify their claims to ensnare patents out of concern for such potentialcompetitors’ products after those treble damage liability. Failure to readcompetitors have sunk significant costs in competitors’ patents can jeopardize plans fortheir products. Patent reform efforts have a noninfringing business or researchlong focused on how to remedy strategy, encourage wasteful duplication ofopportunistic broadening of claims to effort, delay follow-on innovation that couldcapture competitors’ products. Legitimate reasons exist to amend 56 See infra Ch. 4(II)(C)(1) for a description ofclaims and use continuing applications. Any the types of filings that should be covered. 16
  • 20. derive from patent disclosures, and consideration and incorporation of economicdiscourage the development of competition. insights in their decisionmaking. It is troubling that some businesses IV. The FTC Will Pursue Stepsrefrain from reading their competitors’ to Increase Communicationpatents because they fear the imposition of between Antitrust Agenciestreble damages for willful infringement.Nonetheless, infringers must not be allowed and Patent Institutions.to profit from knowingly and deliberatelyusing another’s patented invention due to a Many Hearings participantslow likelihood that the patent holder can expressed concern that the patent andafford to bring suit or obtain substantial competition communities appear to exist indamages. The FTC’s recommendation separate worlds, interacting infrequently atwould permit firms to read patents for their best. Patent practitioners and scholarsdisclosure value and to survey the patent further expressed concern that patentlandscape to assess potential infringement institutions do not always fully understandissues, yet retain a viable willfulness or accommodate economic learning ordoctrine that protects both wronged competition concerns. Increased interactionpatentees and competition. appears desirable to foster better understanding and communication betweenRecommendation 10: the patent and competition communities. Expand Consideration of The FTC wishes to do its part to Economic Learning and improve communication between the Competition Policy Concerns in competition and patent communities. Patent Law Decisionmaking. Accordingly, the FTC will pursue the steps listed below. The Supreme Court has made clearin several decisions that there is room for A. The FTC Will Increase itspolicy-oriented interpretation of the patent Competition Advocacy Rolelaws.57 Indeed, to find the proper balance through Filing Amicus Briefs inbetween patent and competition law, such Appropriate Circumstances.policy-oriented interpretations are essential.Over the past twenty-five years, the The Commission will renew itsincorporation of economic thinking into commitment to the filing of amicus briefs inantitrust has provided significant insights important patent cases that can affectthat have substantially improved the competition, as well as in cases at thedevelopment of antitrust law and intersection of patent and antitrust law.competition policy. The Federal Circuit and When such cases have high stakes for thethe PTO may also benefit from much greater public, the Commission can serve the public interest by filing amicus briefs to present its perspectives regarding the implications of certain issues for consumer welfare. 57 See, e.g., supra notes 10-12; Graham v. JohnDeere Co., 383 U.S. 1 (1966). 17
  • 21. B. In Appropriate Circumstances, the V. Conclusion FTC Will Ask the PTO Director to Reexamine Questionable Patents Both patents and competition make that Raise Competitive Concerns. significant contributions to innovation, consumer welfare, and our nation’s A collective action problem may prosperity. We recognize the importance offrustrate business challenges to questionable the patent system; the recommendations inpatents. Instead of challenging a patent’s this Report are designed to increase thevalidity, many firms may simply license it, likelihood that the valid patents are issuedbecause no single firm has the incentive to and upheld. There is broad consensus on thefinance an expensive legal challenge that significant role that these patents can play towould benefit all of the affected firms, not spur innovation and to encourage thejust the challenger. An enforcement agency, disclosure and commercial development ofhowever, can consider the cost of a inventions.questionable patent to an entire industry andto consumers and can solve this coordination The importance of competition as aproblem. In appropriately narrow spur to innovation also should becircumstances, the FTC will do so. recognized. More patents in more industries and with greater breadth are not always theC. The FTC Will Encourage best ways to maximize consumer welfare. A Increased Communication questionable patent can raise costs and between Patent Institutions and prevent competition and innovation that the Antitrust Agencies. otherwise would benefit consumers. The FTC looks forward to working closely with One means of improving interagency the PTO and other patent organizations tocommunication would be the establishment increase communication and include allof a Liaison Panel between the FTC and the parties in discussion and implementation ofDOJ’s Antitrust Division (collectively, the the FTC’s recommendations.Antitrust Agencies) and the PTO. Such apanel could function as a practical, policy-oriented group designed to permit theexchange of views on important issues asthey arise. Another means would be toestablish an Office of CompetitionAdvocacy within the PTO. Such an officecould, when appropriate, advise PTOpolicymakers about the likely competitiveimpact and economic consequences ofpolicy decisions. A final means would be torequest that Congress amend themembership categories of the Patent PublicAdvisory Committee (“P-PAC”) to includecompetition experts and economists. 18
  • 22. TABLE OF CONTENTSEXECUTIVE SUMMARYCHAPTER 1 INTRODUCTION AND BACKGROUNDI. THE RELATIONSHIP OF COMPETITION AND PATENT LAW AND POLICY A. Each Policy Reflects Fundamental Assumptions about How Best to Organize an Economy and Encourage Innovation B. Competition and Patent Policy Both Promote Consumer Welfare Over Time, and Competition and Patent Policy Generally Work Well Together C. Tension Can Arise Between Competition and Patent Law and Policy in Certain Limited CircumstancesII. VIEWS ON HOW BEST TO BALANCE COMPETITION AND PATENTS TO ACHIEVE CONSUMER WELFARE HAVE VARIED WIDELY OVER TIME A. For Much of the Twentieth Century, Patent and Antitrust Law Have Traded Ascendency with Each Other B. 1980-1990: Congress and the Courts Strengthen Patents, and Antitrust Incorporates an Updated Economic FrameworkIII. COMPETITION AND PATENT POLICY CONTINUE TO SEEK A PROPER BALANCE, AND GROWTH OF THE KNOWLEDGE-BASED ECONOMY ADDS NEW CHALLENGES A. Antitrust and Patent Policy Have Worked to Achieve Better Balance B. The Growth of the Knowledge-Based Economy Creates Ongoing Controversy and Challenges Competition and Patent Policy to Continue Seeking a Better BalanceIV. THE HEARINGS EXAMINED THE CURRENT BALANCE OF COMPETITION AND PATENT LAW AND POLICY IN FOSTERING INNOVATION A. The Hearings Did Not Address Certain Fundamental Questions or Issues with International Ramifications
  • 23. B. The Hearings Examined the Appropriate Balance of Competition and Patent Law and Policy from a Competition and Economic Perspective C. Organization of the ReportCHAPTER 2 THE ROLE OF COMPETITION AND THE PATENT SYSTEM IN SPURRING INNOVATIONIntroductionI. PATENTS’ EFFECTS ON STAND-ALONE INNOVATION A. Patents Can Spur Stand-Alone Innovation B. Costs Of, and Limits To, Patents’ Power to Spur Stand-Alone InnovationII. COMPETITION’S EFFECTS ON INITIAL AND FOLLOW-ON INNOVATION A. Competition Can Spur Innovation, Whether Initial or Follow-On B. Costs Of, and Limits To, Competition’s Power to Spur InnovationIII. PATENTS’ EFFECTS ON FOLLOW-ON INNOVATION A. The Roles of Managed and Independent Follow-On Innovation B. Follow-On Innovation in the Face of a Single Blocking, Initial Patent C. Follow-On Innovation in the Face of Multiple Existing PatentsConclusionCHAPTER 3 BUSINESS TESTIMONY: CURRENT INNOVATION LANDSCAPE IN SELECTED INDUSTRIESI. SUMMARYII. THE PHARMACEUTICAL INDUSTRY A. Introduction B. Industry Description C. The Role of Patents in Spurring Pharmaceutical Innovation
  • 24. D. The Role of Competition in Spurring Pharmaceutical Innovation E. The FTC’s Pharmaceutical Industry Enforcement Actions and Generic Drug Study F. ConclusionIII. THE BIOTECHNOLOGY INDUSTRY A. Introduction B. Industry Description C. The Role of Competition in Spurring Biotechnology Innovation D. The Implications of Patent Protection for Innovation E. Licensing Practices for Biotechnology Research Tools F. ConclusionIV. THE COMPUTER HARDWARE INDUSTRIES, INCLUDING SEMICONDUCTORS A. Introduction B. Industry Description C. The Role of Competition in Spurring Computer Hardware Innovation D. Alternative Means of Fostering Innovation E. The Implications of Patent Protection for Innovation F. Tools to Navigate the Patent Thicket G. ConclusionV. THE SOFTWARE AND INTERNET INDUSTRIES A. Introduction B. Industry Description C. The Role of Competition in Spurring Software and Internet Innovation D. Alternative Means of Fostering Innovation
  • 25. E. The Implications of Patent Protection for Innovation F. Licensing Strategies to Navigate the Patent Thicket G. ConclusionCHAPTER 4 COMPETITION PERSPECTIVES ON SUBSTANTIVE STANDARDS OF PATENTABILITYI. STATUTORY STANDARDS OF PATENTABILITYII. THE INTERPRETATION AND APPLICATION OF THE STANDARDS OF PATENTABILITY A. The Interpretation and Application of the Nonobviousness Requirement B. Enablement, Written Description, and Best Mode C. Other Doctrines that Affect Patent Breadth D. Utility and Research Issues E. Business Method Patents: An Illustration of Transition IssuesIII. CONCLUSIONCHAPTER 5 COMPETITION PERSPECTIVES ON HOW PROCEDURES AND PRESUMPTIONS AFFECT PATENT QUALITYI. IMPACT ON COMPETITIONII. PATENT EXAMINATION A. Data on Overall Performance B. Ex Parte Nature C. AnalysisIII. REEXAMINATION, OPPOSITION, AND REVIEW A. Current Procedures B. Proposals for Reform
  • 26. C. AnalysisIV. PATENT LITIGATION A. General Trends B. Presumption of Validity/Clear and Convincing Evidence Recommendation C. Willfulness/Treble Damages RecommendationV. CONCLUSIONCHAPTER 6 COMPETITION AND PATENT POLICY CAN AND SHOULD WORK TOGETHERIntroductionI. ANTITRUST AND PATENT LAW AND POLICY A. Antitrust Law and Policy Can and Should Take Patent Policy into Account to Promote Consumer Welfare Over Time B. Patent Law and Policy Can and Should Take Competition Policy into Account to Promote Consumer Welfare Over TimeII. THE FEDERAL CIRCUIT: GOALS, JURISDICTION, CHOICE OF LAW, AND CASE LAW TRENDS A. The Federal Circuit and its Intended Effect on the Law B. Jurisdiction and Choice of Law Issues at the Federal Circuit C. Trends in the Law of the Federal CircuitIII. INSTITUTIONAL CONSIDERATIONS FOR THE ANTITRUST ENFORCEMENT AGENCIES AND THE PTO A. Recommendations Relating to the PTO B. The FTC Will Pursue Steps to Increase Communication Between Antitrust Agencies and Patent Institutions
  • 27. APPENDIX A CONTRIBUTORS TO FTC/DOJ HEARINGSAPPENDIX B PUBLIC COMMENTSAPPENDIX C GLOSSARY OF PATENT TERMSAPPENDIX D SELECTED FEDERAL STATUTES
  • 28. CHAPTER 1 INTRODUCTION AND BACKGROUNDI. THE RELATIONSHIP OF COMPETITION AND PATENT LAW AND POLICY . . . . 3 A. Each Policy Reflects Fundamental Assumptions about How Best to Organize an Economy and Encourage Innovation . . . . . . . . . . . . . . . . . . . . . . . . . 3 1. Competition Policy and Antitrust Law . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Patent Policy and Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 B. Competition and Patent Policy Both Promote Consumer Welfare Over Time, and Competition and Patent Policy Generally Work Well Together. . . . . . . . . . . 7 C. Tension Can Arise Between Competition and Patent Law and Policy in Certain Limited Circumstances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1. Grant of a Patent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 a. Is the Patent Warranted? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 b. Does the Grant of the Patent Confer Market Power on the Patentholder or Unnecessarily Increase Transaction Costs? . . . . 11 2. Business Conduct with Respect to a Patent. . . . . . . . . . . . . . . . . . . . . . . 12 a. Is Antitrust Enforcement Warranted? . . . . . . . . . . . . . . . . . . . . . 13 b. Does Antitrust Enforcement Undermine the Incentives Created by the Patent System? . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3. Enhancing Consumer Welfare Requires a Proper Balance of Competition and Patent Law and Policy . . . . . . . . . . . . . . . . . . . . . . . . . 14II. VIEWS ON HOW BEST TO BALANCE COMPETITION AND PATENTS TO ACHIEVE CONSUMER WELFARE HAVE VARIED WIDELY OVER TIME. . . . . . 14 A. For Much of the Twentieth Century, Patent and Antitrust Law Have Traded Ascendency with Each Other. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1. 1890-1930: Patents Receive Little Antitrust Scrutiny . . . . . . . . . . . . . . 15 2. 1930-1980: Antitrust Is Generally Ascendant. . . . . . . . . . . . . . . . . . . . . 15 B. 1980-1990: Congress and the Courts Strengthen Patents, and Antitrust Incorporates an Updated Economic Framework. . . . . . . . . . . . . . . . . . . . . . . . . . 18 1. Congress and the Courts Strengthen Patents . . . . . . . . . . . . . . . . . . . . . . 18 a. Congress Creates the Court of Appeals for the Federal Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 b. The Supreme Court Interprets Patentable Subject Matter Broadly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2. Antitrust Incorporates an Updated Economic Framework. . . . . . . . . . . . 22
  • 29. III. COMPETITION AND PATENT POLICY CONTINUE TO SEEK A PROPER BALANCE, AND GROWTH OF THE KNOWLEDGE-BASED ECONOMY ADDS NEW CHALLENGES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 A. Antitrust and Patent Policy Have Worked to Achieve Better Balance . . . . . . . . 23 1. Antitrust Policy Has Continued to Implement New Economic Learning in Addressing the Intersection of Antitrust and Patents. . . . . . 23 2. Patent Policy Has Implemented Certain Reforms and Rules that Can Lessen Anticompetitive Conduct and Increase Competition . . . . . . 26 a. Congress Enacted the American Inventors Protection Act of 1999 (AIPA). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 b. The Federal Circuit Has Increased Business Certainty and Has Noted Competition Concerns in Certain Contexts . . . . . . . . . . . 28 c. The PTO Has Implemented Certain Reforms that Can Aid Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 B. The Growth of the Knowledge-Based Economy Creates Ongoing Controversy and Challenges Competition and Patent Policy to Continue Seeking a Better Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 1. Follow-On Innovation, Product Commercialization, and Patent Proliferation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2. Procedures that Third Parties Can Use to Challenge Questionable Patents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3. Patent Prosecutions and Examinations within the PTO . . . . . . . . . . . . . . 33 4. Patent Quality and Patentable Subject Matter. . . . . . . . . . . . . . . . . . . . . . 35IV. THE HEARINGS EXAMINED THE CURRENT BALANCE OF COMPETITION AND PATENT LAW AND POLICY IN FOSTERING INNOVATION. . . . . . . . . . . . . 35 A. The Hearings Did Not Address Certain Fundamental Questions or Issues with International Ramifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 B. The Hearings Examined the Appropriate Balance of Competition and Patent Law and Policy from a Competition and Economic Perspective . . . . . . . . . . . . . 36 1. The Legal System Should Provide Efficient Incentives for All Types of Innovation, Including Both Single-Stage and Follow-On Innovation . . . 36 2. Safeguard the Patent System’s Disclosure Function . . . . . . . . . . . . . . . . 37 3. The Patent System Should Avoid Creating or Upholding Unwarranted Patents that Confer Market Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4. The Patent System Should Rely on Substantive Standards and Procedures that Minimize the Sum of Error and Process Costs and the Detrimental Effects of Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 C. Organization of the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
  • 30. CHAPTER 1 INTRODUCTION AND BACKGROUND Innovation benefits consumers agencies contribute to R&D funding,through the development of new and especially in national defense, health, andimproved goods and services, and spurs space.5 Second, the federal government setseconomic growth. An economy’s capacity policies that influence how businesses andfor invention and innovation helps drive its individuals invest many more billions ofeconomic growth and the degree to which dollars in R&D. Tax and environmentalstandards of living increase.1 Indeed, the policies, for example, all can influenceUnited States economy and the economies of which R&D companies undertake and howother countries have enjoyed “huge much they spend.productivity gains from the development andrapid adoption of new information and Competition and patents stand outcommunication technologies.”2 The among the federal policies that influencetechnological breakthroughs that introduced private R&D.6 Competition among firms“automobiles, airplanes, radio, television, prods inventors to be first in the market withspace travel, telephones, internet, modern a new product or service at a price andpharmaceuticals, and the like” illustrate how quality that consumers want. Patent policyinnovation improves the quality of our lives encourages prospective inventors to investin ways that are hard to measure and time and money in inventions, because aunderscore the importance of stimulating patent’s grant of the exclusive right to make,innovation.3 sell, and use the invention for a certain period of time can allow inventors to realize The federal government has aprofound impact on R&D in the U.S. First,the federal government funds certain R&D. Science 1 (2003), at http://www.aaas.org/spp/rd/ca03.pdf.In FY 2003, federal investment in R&D hit a 5 For example, the Department of Defensenew record of $117 billion, a 13.8 percent (DOD) accounts for half the total federal R&D portfolio.increase over FY 2002 and the largest dollar Support for R&D makes up 97 percent of the budget of theincrease in history.4 Many government National Institute of Health (NIH). The National Science Foundation (NSF) accounts for about 20 percent of federal support to academic institutions for basic research. The 1 National Aeronautics and Space Administration (NASA) Federal Reserve Board Vice Chairman Roger spends two-thirds of its budget (excluding the SpaceW. Ferguson, Jr., Patent Policy in a Broader Context, Shuttle program) on R&D. See Koizumi & Turner,Remarks at 2003 Financial Markets Conference of the Congressional Action on Research and Development in theFederal Reserve Bank of Atlanta (April 5, 2003), at FY 2003 Budget at 11-16; NFS websitehttp://www.federalreserve.gov/boarddocs/speeches/2003/20 http://www.nsf.gov/home/programs; AAAS R&D Funding030407/default.htm. Update, FY 2003 Omnibus Bill Complete NIH Doubling Plan; Large Increases for Bioterrorism R&D and Facilities 2 Id. 1, 3 (Feb. 25, 2003), at http://www.aaas.org/spp/rd/nih03f.pdf. 3 American Bar Association Section of AntitrustLaw, The Economics of Innovation: A Survey (Public 6 In the Hearings, panelists focused on patentsComment) 2, at and not other forms of intellectual property. Most of thehttp://www.ftc.gov/opp/intellect/0207salabasrvy.pdf antitrust cases involving intellectual property involve(hereinafter ABA (Economics stmt)). patents in particular. See, e.g., 1 HERBERT HOVENKAMP ET AL., IP AND ANTITRUST: AN ANALYSIS OF ANTITRUST 4 Kei Koizumi & Paul W. Turner, Congressional PRINCIPLES APPLIED TO INTELLECTUAL PROPERTY LAW §Action on Research and Development in the FY 2003 1.3c at 1-14 (2002) (hereinafter HOVENKAMP ET AL., IPBudget, American Association for the Advancement of AND ANTITRUST).
  • 31. returns sufficient to encourage the initial To understand better the currentinvestments.7 relationship between competition and patent law and policy, and whether it strikes the Competition and patent policy are proper balance, the Federal Tradebound together by the economics of Commission (FTC) and the Antitrustinnovation and an intricate web of legal Division of the Department of Justice (DOJ)rules that seek to balance the scope and held a series of Hearings from Februaryeffect of each policy. Errors or systematic through November 2002. The Hearings tookbiases in the interpretation or application of place over 24 days, with more than 300one policy’s rules can harm the other panelists, including experienced businesspolicy’s effectiveness. For example, patent representatives from large and small firms,law precludes the patenting of an “obvious” representatives from the independentinvention. If, however, patent law sets the inventor community, all of the leadingbar for “obviousness” too low, and patent and antitrust organizations, many oferroneously allows patents on “obvious” the leading antitrust and patent practitioners,inventions, then patent law can thwart and scholars in economics and antitrust andcompetition that otherwise might have patent law. Care was taken to solicit alldeveloped based on the obvious technology. points of view, and the transcripts of theConversely, competition policy – as Hearings provide a wide spectrum of well-implemented through antitrust law – considered experience with and perspectivesprohibits only anticompetitive business on patent and competition-related issues. Inconduct. If antitrust enforcement addition, written comments were solicited;erroneously condemns efficient, welfare- the FTC received about 100 writtenenhancing conduct with respect to a valid submissions.patent, then antitrust enforcement canundermine the incentives the patent system The FTC took the lead in examiningcreates to encourage innovation. A the issues addressed in this report, whichchallenge for both policies is to find the discusses what the FTC has learned and, asproper balance of competition and patent appropriate, makes recommendations forprotection.8 changes to patent law and policy to achieve a better balance with competition policy. The DOJ and the FTC worked together 7 See generally infra Ch. 2(I). developing the record for a forthcoming 8 See Bonito Boats, Inc. v. Thunder Craft Boats, joint report that will examine antitrust’sInc., 489 U.S. 141, 146 (1989) (“From their inception, the approach to maintaining the proper balancefederal patent laws have embodied a careful balance with the patent system.between the need to promote innovation and therecognition that imitation and refinement through imitationare both necessary to invention itself and the very lifebloodof a competitive economy.”); Richard Posner, Antitrust inthe New Economy, 68 ANTITRUST L.J. 925, 927 (2001)(“The patent and copyright laws try to strike the output-maximizing balance by giving the creator of intellectualproperty some but not complete protection fromcompetition.”); 1 HOVENKAMP ET AL., IP AND ANTITRUST §1.3b at 1-14 (patents can limit the reach of antitrust law,and antitrust constrains what a patentee can do with its patent). 2
  • 32. I. THE RELATIONSHIP OF in competition will produce more and price COMPETITION AND lower than monopolists. Monopolists not only take money away from consumers by PATENT LAW AND raising prices, but they impose a POLICY ‘deadweight loss’ on society by reducing their output below the level whichA. Each Policy Reflects consumers would be willing to purchase at a Fundamental Assumptions competitive price.”12 Thus, economics about How Best to Organize informs us that effective competition is the an Economy and Encourage best mechanism for achieving the optimum mix of products and services in terms of Innovation price, quality, and consumer choice. Moreover, economic learning focuses on the1. Competition Policy and Antitrust importance of competition in enhancing Law consumer welfare not only with respect to existing products, but also the development Competition through free enterprise of new and improved products andand open markets is the organizing principle services.13 Monopolists can have fewerfor most of the U.S. economy.9 The United incentives to innovate than do competitiveStates generally has chosen antitrust law firms.14(rather than regulation) to provide thegoverning rules for competition. For the last Antitrust law protects competitiontwenty years, antitrust law has recognized and the competitive process “by preventingenhancing consumer welfare as the single certain types of conduct that threaten a freeunifying goal of competition policy.10 To market.”15 Antitrust evaluates agreementsserve that objective, competition policy and among firms to determine whether theyantitrust enforcement use a framework basedon sound economics.11 Economics affirms that“[c]ompetition is good for a variety of 12 1 HOVENKAMP ET AL., IP AND ANTITRUST §1.2reasons. Basic economics teaches that firms at 1-5 through 1-6. See also William M. Landes & Richard A. Posner, Market Power in Antitrust Cases, 94 HARV. L. REV. 937, 991 (1981). 9 See, e.g., 1 HOVENKAMP ET AL., IP AND 13ANTITRUST § 1.3a at 1-10 (“[A] fundamental principle of See generally infra Ch. 2(II)(A).our economic system is the proposition that free marketcompetition will best ensure an efficient allocation of 14 1 HOVENKAMP ET AL., IP AND ANTITRUST § 1.2resources in the absence of market failure.”). at 1-6. Others emphasize that, depending on the circumstances, monopolists also can have greater 10 See, e.g., IIA PHILLIP E. AREEDA & HERBERT incentives to innovate. See infra at Ch. 2(II)(A)(3).HOVENKAMP, ANTITRUST LAW : AN ANALYSIS OFANTITRUST PRINCIPLES AND THEIR APPLICATION ¶ 401 (2d. 15 1 HOVENKAMP ET AL., IP AND ANTITRUST § 1.2ed. 2002). at 1-5. See also Northern Pacific Railway v. United States, 356 U.S. 1, 4 (1958) (“The Sherman Act was designed to 11 See, e.g., HERBERT HOVENKAMP, FEDERAL be a comprehensive charter of economic liberty aimed atANTITRUST POLICY: THE LAW OF COMPETITION AND ITS preserving free and unfettered competition as the rule ofPRACTICE § 2.3a (2d ed. 1999). trade.”). 3
  • 33. “unreasonably restrain trade.”16 For the Supreme Court has held that the purposeexample, antitrust prohibits naked of the antitrust laws is to protectagreements among competitors on the price competition, not competitors.20 Thus,they will charge or which customers each antitrust enforcement has ceased protectingwill serve. For most other agreements, individual firms in favor of protectingantitrust evaluates likely procompetitive and consumer welfare, because protectinganticompetitive effects.17 Antitrust law also individual firms often served to harmconstrains the creation of market power consumers by protecting firms fromthrough mergers,18 and prohibits competition.21 Antitrust’s focus onmonopolization and attempts and consumer welfare also reveals thatconspiracies to monopolize.19 governmental impediments to, or exemptions from, competition can be as In recognizing consumer welfare as harmful to consumers as private businessits proper goal, antitrust law has relinquished restraints.22earlier doctrines that sought to protectcompetitors rather than competition. Indeed, 2. Patent Policy and Law The U.S. economy also reflects the 16 Sherman Act of 1890 § 1, 15 U.S.C. § 1; belief that limited exclusive rights inStandard Oil Co. v. United States, 221 U.S. 1, 59-69 intellectual property – as distinguished from(1911); see also Federal Trade Commission Act of 1914§ 5, 15 U.S.C. § 45. tangible property – can encourage innovation, which also benefits consumers.23 17 Broadcast Music, Inc. v. Columbia Article I, Section 8 of the ConstitutionBroadcasting System, Inc., 441 U.S. 1, 8 (1979). Seegenerally Federal Trade Commission and U.S. Department authorizes Congress “[t]o promote theof Justice, Antitrust Guidelines for Collaborations Among Progress of Science and useful Arts, byCompetitors § 2 (April 2000), available at securing for limited Times to Authors andhttp://www.ftc.gov/os/2000/04/ftcdojguidelines.pdf. 18 Clayton Act of 1950 § 7, 15 U.S.C. § 18; 15 20U.S.C. § 45. Market power arises when the “defendant (1) Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc.,can profitably set prices well above its costs and (2) enjoys 429 U.S. 477, 488 (1977) (“The antitrust laws, however,some protection against a rival’s entry or expansion that were enacted for the protection of competition, notwould erode such supracompetitive prices and profits.” IIA competitors” (internal citations omitted)).AREEDA & HOVENKAMP, ANTITRUST LAW: AN ANALYSIS 21OF ANTITRUST PRINCIPLES AND THEIR APPLICATION ¶ 501 at See generally I PHILLIP E. AREEDA & HERBERT90. See also United States v. E.I. duPont de Nemours & HOVENKAMP, ANTITRUST LAW: AN ANALYSIS OFCo., 351 U.S. 377 (1956); Federal Trade Commission and ANTITRUST PRINCIPLES AND THEIR APPLICATION ¶100 at 3-U.S. Department of Justice, Horizontal Merger Guidelines 7 (2d ed. 2000).§ 1.1 (1992), available at 22http://www.ftc.gov/bc/docs/horizmer.htm; FTC/DOJ See, e.g., HOVENKAMP, FEDERAL ANTITRUSTHearings on Competition and Intellectual Property Law POLICY: THE LAW OF COMPETITION AND ITS PRACTICEand Policy in the Knowledge-Based Economy, William E. § 18.1a at 680.Kovacic Testimony Feb. 8, 2002 (Antitrust Law for PatentLawyers), at page 33 (hereinafter, citations to transcripts of 23 ROBERT L. HARMON, PATENTS AND THEthese Hearings state the speaker’s last name, the date of FEDERAL CIRCUIT § 1.2 at 11 (5th ed. 2001) (noting that thetestimony, and relevant page(s)). Antitrust does not exclusive right granted by a patent “was for the nationalconstrain all exercises of market power, however. See infra purpose of advancing the useful arts – the process todayCh. 1(I)(B). called technological innovation[,]” and serves “the public interest in technological advancement.” (Footnotes 19 15 U.S.C. § 2. omitted)). 4
  • 34. Inventors the exclusive Right to Bo x 1-1 . Two of the Basics of the Patent Documenttheir respective Writings andDiscoveries.”24 The patent A patent contains a great deal of information. Among thestatute25 confers a right to most important are the patent’s “specification” and “claims.” Theexclude others from making, specification must provide a “written description of the invention, and of the manner and process of making and using it,” and mustusing, or selling in the United disclose the “best mode” know n to the inventor of carrying out theStates the invention claimed by invention. 35 U.S.C. §§ 111, 112.the patent for twenty years fromthe date of filing the patent The patent’s “claims” are “the portion of the patentapplication.26 document that defines the patentee’s rights.” Markman v. Westview Instruments, Inc ., 517 U.S. 370, 372 (1992). Since the claims essentially articulate the “metes and bounds” of the patentee ’s To obtain a patent, an intellectual property, they are one of the most important parts of theinvention (that is, a product, mod ern patent document. See generally R OBERT M ERGES & J O H Nprocess, machine, or D UFFY , P ATENT L A W A N D P OLICY : C A S ES A N D M ATE RIALS 25-26composition of matter) must be (3d ed. 2002).novel, nonobvious, and useful,and must meet certainrequirements for the description of the describe what the inventor claims so thatinvention.27 A patentee must disclose the others can easily discern the boundaries ofinvention clearly enough so that one skilled the patent;29 and must tell the public thein that art can make and use it without inventor’s “best mode” – most effectiveundertaking a great deal of method – for practicing the invention.30 Seeexperimentation;28 must highlight or also Box 1-1. 24 Patent law reflects certain differences U.S. CONST. art. I, § 8 also authorizesCongress to establish the copyright system. between intellectual property and tangible property. Problems of copying by third 25 The first U.S. patent statute was passed by the parties make it generally more difficult forfirst U.S. Congress; it has been substantially revised from holders of intellectual property to excludetime to time. See generally ROBERT PATRICK MERGES &JOHN FITZGERALD DUFFY , PATENT LAW AND POLICY: others from its use than it is for holders ofCASES AND MATERIALS 1-13 (3d ed. 2002) (reviewinghistory of patent law); ROGER E. SCHECHTER & JOHN R.THOMAS, INTELLECTUAL PROPERTY: THE LAW OFCOPYRIGHTS, PATENTS, AND TRADEMARKS § 13.2 at 283-87 Patent and Trademark Office (2/6/02) 2 (disclosure for(2003) (reviewing history of patent law). right to exclude is a “remarkable trade-off”), at http://www.ftc.gov/opp/intellect/rogan.htm (hereinafter 26 35 U.S.C. § 154(a)(2). Rogan (stmt)). 27 29 35 U.S.C. §§ 101-103, 112. 35 U.S.C. § 112. HARMON, PATENTS AND THE FEDERAL CIRCUIT § 5.4 at 218 (“The inquiry under 28 See 35 U.S.C. § 112; J.E.M. Ag Supply, Inc. v. § 112¶ 2 focuses on whether the claims, as interpreted inPioneer Hi-Bred Int’l, Inc., 534 U.S.124, 142 (2001) (“The view of the written description, adequately perform theirdisclosure required by the Patent Act is the ‘quid pro quo function of notifying the public of the patentee’s right toof the right to exclude’” (internal citations omitted)); exclude.”); MERGES & DUFFY , PATENT LAW AND POLICY:MERGES & DUFFY , PATENT LAW AND POLICY: CASES AND CASES AND MATERIALS at 262.MATERIALS at 262; James E. Rogan, Prepared Remarks of 30James E. Rogan, Under Secretary of Commerce for 35 U.S.C. § 112; MERGES & DUFFY , PATENTIntellectual Property and Director of the United States LAW AND POLICY: CASES AND MATERIALS at 263. 5
  • 35. tangible property to do so.31 Once third create.”36 Patent policy serves consumerparties have learned about an invention, they interests in innovation through other meansmay copy and use it.32 Intellectual property as well.37 By requiring disclosure of theis also “non-rivalrous” – that is, many patented invention in an issued patent,38 thepeople may use innovative technology, and patent system can encourage furtherthey all may use it without diminishing innovations if inventors forego keeping theirothers’ ability to use it.33 Many people may inventions as trade secrets and insteademploy an innovation without depletion, and disclose their inventions.39 The patentit is hard to identify and prevent those who system also can encourage furtherwill not pay for its use from using it.34 In innovation by facilitating investment in thesuch circumstances, inventors are unlikely to research, development, and marketinghave sufficient incentives to pursue and necessary to commercialize a product.40 Oneproduce their inventions.35 To preserve incentives to invent, 36 1 HOVENKAMP ET AL., IP AND ANTITRUST §patent policy protects inventors from such 1.1 at 1-2.misappropriation. “The principal basis for 37 See generally infra Ch. 2(I)(A)(2), (I)(A)(3).intellectual property protection in the UnitedStates is the utilitarian or economic 38 See 35 U.S.C. § 112.incentive framework. That is, intellectual 39property in the United States is See MERGES & DUFFY , PATENT LAW AND POLICY: CASES AND MATERIALS at 259 (explaining that byfundamentally about incentives to invent and the late eighteenth century, many viewed the primary benefit of the patent system as “the technological know- how behind the inventor’s patent. . . . This was a major 31 change in the economic role of patents, for it shifted the See 1 HOVENKAMP ET AL., IP AND ANTITRUST emphasis from the introduction of finished products into§ 1.1 at 1-2. commerce to the introduction of new and useful 32 information to the technical arts[.]” (emphasis in original)); 1 HOVENKAMP ET AL., IP AND ANTITRUST § 1.1 R. Levin 2/6 at 100 (research has shown the disclosureat 1-3 through 1-4; see also Thomas 2/8 (Patent Session) at requirement is “quite procompetitive”); SCHECHTER &14-15. THOMAS, INTELLECTUAL PROPERTY: THE LAW OF COPYRIGHTS, PATENTS AND TRADEMARKS § 13.4 .1 at 288 33 HAL VARIAN, MICROECONOMIC ANALYSIS 414- (noting that “[t]rade secrets do not enrich the collective415 (3d ed. 1992); 1 HOVENKAMP ET AL., IP AND knowledge of society, . . ., nor do they discourage othersANTITRUST § 1.1 at 1-2; see also DonPaul Olshove, from engaging in duplicative research.”); Donald S.Comments Regarding Competition & Intellectual Property Chisum, Comment: Anticipation, Obviousness,(Public Comment) 3, at Enablement: An Eternal Golden Braid, 15 AIPLA Q.J. 57http://www.ftc.gov/os/comments/intelpropertycomments/ol (1987) (explaining that primary purpose of disclosureshovedonpaul.htm. requirement is to “put[] the invention in full possession of the public so the invention may be freely made and used 34 1 HOVENKAMP ET AL., IP AND ANTITRUST § after expiration of the patent”).1.1 at 1-3 through 1-4. See generally, Thomas 2/8 (PatentSession) at 14-15. But see infra Ch. 2(II)(A)(2) (firms sometimes favor trade secrecy over patents as an appropriation 35 SCHECHTER & THOMAS, INTELLECTUAL mechanism) and Ch. 3(IV)(D) (firms sometimes obtainPROPERTY: THE LAW OF COPYRIGHTS, PATENTS, AND patents only when they view trade secrecy as impossible).TRADEMARK § 13.4.1 at 288 (noting that, if inventions can 40easily be duplicated or exploited by free riders, “[t]he HARMON, PATENTS AND THE FEDERAL CIRCUITresulting inability of inventors to capitalize on their § 1.2 at 11; SCHECHTER & THOMAS, INTELLECTUALinventions would lead to an environment where too few PROPERTY: THE LAW OF COPYRIGHTS, PATENTS ANDinventions are made.” (Footnote omitted.)). TRADEMARKS § 13.4.1 at 289. See infra Ch. 2 (I)(A). 6
  • 36. patent scholar has described the public Secretary of Commerce for Intellectualpurposes of the patent grant as “an incentive Property and Director of the United Statesto invention, investment, and disclosure.”41 Patent and Trademark Office James Rogan stated that “patent law and competition law .B. Competition and Patent . . are highly compatible and serve many Policy Both Promote similar ends.”45 Others have also observed that antitrust and patent law “are Consumer Welfare Over complementary efforts to promote an Time, and Competition and efficient marketplace and long-run, dynamic Patent Policy Generally competition through innovation.”46 Both Work Well Together doctrines can function to promote consumer welfare.47 Patent and antitrust law “are actuallycomplementary, as both are aimed at http://www.usdoj.gov/atr/public/speeches/10162.htmencouraging innovation, industry, and (hereinafter Opening Day Comments).competition.”42 In introducing these 45hearings, FTC Chairman Muris emphasized Rogan (stmt) 3.that “properly understood, IP law and 46 1 HOVENKAMP, JANIS, & LEMLEY, IP ANDantitrust law both seek to promote ANTITRUST § 1.3 at 1-12 through 1-13. See also M.innovation and enhance consumer Thompson 2/25 at 7; American Bar Association Section of Antitrust Law, Statement 6-7, atwelfare.”43 Then-Assistant Attorney General http://www.ftc.gov/opp/intellect/020628busey.pdfCharles James similarly noted that (hereinafter ABA Antitrust Section (stmt)); Sheila F.“intellectual property and antitrust law share Anthony, Antitrust and Intellectual Property Law: From Adversaries to Partners, 28 AIPLA Q.J. 1 at 1, 7 (2000), atthe common purpose of promoting dynamic http://www.ftc.gov/speeches/other/aipla.htm.competition and thereby enhancingconsumer welfare.”44 Likewise, Under Antitrust and patent law show similarities and differences in each’s consideration of short and long run effects on consumer welfare. “Patent law and the incipiency elements of antitrust law are similar in that they 41 HARMON, PATENTS AND THE FEDERAL CIRCUIT both are ultimately based on inherently uncertain§ 1.4(b) at 22. predictions of what is going to happen in the future. The difference is that in the antitrust regime, we sometimes are 42 Atari Games Corp. v. Nintendo of Am., 897 concerned about conduct that in the short term may beF.2d 1572, 1576 (Fed. Cir.1990). See also R. Hewitt Pate, benign or even helpful to consumers, but that may beRefusals to Deal and Intellectual Property Rights, 10 GEO. harmful in the long run, whereas in the patent regime weMASON L. REV. 429, 429 (2002) (“Intellectual property and are willing to tolerate immediate consumer harm [e.g.,antitrust laws share a common objective – to encourage monopoly pricing] in the expectation that in the long run itinnovation, industry, and competition.”). will benefit consumers by encouraging innovation.” Thomas B. Leary, The Patent-Antitrust Interface, Remarks 43 Timothy J. Muris, Competition and before the ABA Section of Antitrust Law Program,Intellectual Property Policy: The Way Ahead, Before Philadelphia, Pennsylvania 3-4 (May 3, 2001), atAmerican Bar Association Antitrust Section, Fall Forum 2 http://www.ftc.gov/speeches/leary/ipspeech.htm.(Nov. 15, 2001), athttp://www.ftc.gov/speeches/muris/intellectual.htm 47 WARD S. BOWMAN , JR., PATENT AND(hereinafter The Way Ahead). ANTITRUST LAW: A LEGAL AND ECONOMIC APPRAISAL, 2-3 (1973); 1 HOVENKAMP, JANIS, & LEMLEY, IP AND 44 Charles A. James, Opening Day Comments, ANTITRUST § 1.3 at 1-11 (“[W]hen one departs from theJoint DOJ-FTC Hearings on Competition and Intellectual static view of markets and takes a longer-run approach, it isProperty Law and Policy in the Knowledge-Based even plausible that intellectual property and the antitrustEconomy 1 (Feb. 6, 2002), at laws share a common goal”). 7
  • 37. In most cases, competition and limiting the duration of a patent, “[t]hepatent policy work in tandem toward this Patent Clause itself reflects a balancegoal.48 Competition advocates understand between the need to encourage innovationthat “an effective legal regime defining and and the avoidance of monopolies whichprotecting property rights is essential to a stifle competition without any concomitantwell-functioning competitive economy[,]” advance in the ‘Progress of Science andand that “[patent] law plays an important useful Arts.’”51 The patentabilityrole in this overall property rights regime.”49 requirements for novelty andThe patent system spurs competition to nonobviousness “are grounded in the notioninnovate, because it can increase the that concepts within the public grasp, orpotential rewards to successful innovators by those so obvious that they readily could be,limiting the competition that may arise from are the tools of creation available to all.”52the innovation. As the Supreme Court has Thus, patent policy recognizes that certainnoted, “free competition” is “the baseline” limits on patents are necessary to avoidon which “the patent system’s incentive to unnecessarily restraining competition.53creative effort depends.”50 Moreover,patents protect intellectual property that Competition and patent policyfirms use as inputs to compete. Thus, as a approach these issues through differentgeneral matter, competition spurs the means to achieve their congruent goals,creation of patents, and patents protect however.54 Antitrust concerns about harm toinputs that firms use in the competitive competition typically flow from the creationprocess. or exercise of monopoly power in a relevant antitrust market.55 “Intellectual property, Analogously, patent policyrecognizes the value of competition. The 51 Bonito Boats, 489 U.S. at 146.Supreme Court has pointed out that, by 52 Bonito Boats, 489 U.S. at 156. 48 53 See, e.g., American Intellectual Property Law See, e.g., HARMON, PATENTS AND THEAssociation (AIPLA), AIPLA Testimony (Public Comment) FEDERAL CIRCUIT § 1.2 at 12 (“It should not be supposed,2-4 (“we view the two sets of laws as fully sharing however, that there are no public costs associated with thecommon, not conflicting, goals and acting together in right to exclude. These include inflated prices (invariablybalance”), at absorbed by the consumer), which frequently accompanyhttp://www.ftc.gov/os/comments/intelpropertycomments/ai exclusive rights, and overinvestment. The patent systempla.pdf (hereinafter AIPLA (stmt)). seeks to maintain an efficient balance between incentives to create and commercialize and the public costs engendered 49 Muris, The Way Ahead at 2; see also by these incentives.” (Footnotes omitted)).American Bar Association Section of Antitrust Law and 54Section of International Law and Practice, Comments and BOWMAN , PATENT AND ANTITRUST LAW: ARecommendations on the Competition Elements of the LEGAL AND ECONOMIC APPRAISAL at 2; 1 HOVENKAMP ETDoha Declaration, Before the United States Trade AL., IP AND ANTITRUST § 1.3b at 1-13.Representative 12 (2003) (noting that a “functional systemfor the definition, protection and exchange of common 55 Although cases under Sections 1 and 2 of theforms of tangible and intangible property (including Sherman Act may distinguish between “monopoly power”intellectual property)” is necessary for a “successful market and “market power,” this report uses the termseconomy” based on competition), at interchangeably, because the distinction is not importanthttp://www.abanet.org/antitrust/comments/doha.doc. for present purposes. The creation or exercise of monopoly power does not always violate the antitrust laws. See infra 50 Bonito Boats, 489 U.S. at 156. Ch. 1(I)(B). 8
  • 38. while it does not generally create a creation of monopoly power can bemonopoly, may in some cases permit or necessary to achieve a greater gain foreven encourage monopoly in order to give consumers.61 Moreover, antitrust law doesincentives for invention.”56 As Judge Pauline not outlaw monopoly in all circumstances.Newman noted, “[p]atents are directed at For example, monopoly achieved solely withinnovation. That’s their purpose, and of “superior skill, foresight, and industry” doescourse they affect competition. That’s how not violate the antitrust laws.62they work. That’s the only way they work,and that’s why we’re here today.”57 The C. Tension Can Arise Betweenexistence of a patent may enable a firm to Competition and Patent Lawcharge monopoly prices or otherwise limit and Policy in Certain Limitedcompetition.58 Circumstances Patents do not always or evenfrequently confer monopoly power on their Nevertheless, there are opportunitiesowners.59 Indeed, most patents do not for tension between competition and patentconfer monopoly power on their holders,60 law and policy. Broadly speaking, thisand most business conduct with respect to tension most typically arises in two settings.patents does not “unreasonably restrain” or The first involves the grant of a patent; theserve to monopolize markets. Even when a second involves business conduct withpatent does confer monopoly power, that respect to a patent. Competition and patentalone does not create an antitrust violation. policymakers may reach differentAntitrust law recognizes that a patent’s conclusions about whether each policy has adequately accommodated the other’s concerns. 56 1 HOVENKAMP ET AL., IP AND ANTITRUST § 1. Grant of a Patent1.3b at 1-13. 57 Newman 2/6 at 38. See also 1 HOVENKAMP ET Competition policy asks twoAL., IP AND ANTITRUST § 1.3a at 1-9 through 1-10 (“Indeed, questions in connection with the grant of ain order for the intellectual property laws to succeed ingiving authors and inventors an incentive to create, the law patent. The first question is whether themust give them some power over price.” (emphasis in patent is warranted. Patent policy, of course,original)). as set through statutes and decisional law, 58 1 HOVENKAMP ET AL., IP AND ANTITRUST § also seeks to ensure that the Patent and1.3a at 1-10. 59 61 HARMON, PATENTS AND THE FEDERAL CIRCUIT BOWMAN , PATENT AND ANTITRUST LAW: A§ 1.4(b) at 21 (“Patent rights are not legal monopolies in LEGAL AND ECONOMIC APPRAISAL at 3, n. 2.the antitrust sense of the word. Not every patents is amonopoly, and not every patent confers market power.” 62 United States v. Aluminum Co. of Am., 148(Footnote omitted.)). See also ABA Antitrust Section F.2d 416, 430 (2d Cir. 1945) (“The successful competitor,(stmt) 11-12; Kovacic 2/8 (Antitrust Law for Patent having been urged to compete, must not be turned uponLawyers) at 32-33 (hereinafter Antitrust Session); Tom 2/8 when he wins.” Id.). See also U.S. v. Grinnell Corp., 384(Antitrust Session) at 50. U.S. 563, 571 (1996) (the offense of monopoly is distinct from "growth or development as a consequence of a 60 AIPLA (stmt) 21; Cohen 2/20 at 63; superior product, business acumen, or historic accident");Dickinson 2/6 at 52-53; Pitofsky 2/6 at 29-30. ABA Antitrust Section (stmt) 12. 9
  • 39. Trademark Office (PTO) does not grant, and Competition policy and economicthe courts do not uphold, invalid patents. perspectives would ask a somewhat differentThe second question is whether the patent question, one that focuses on whether andconveys market power. The patent system how the patent is necessary to encouragedoes not ask this question. We introduce innovation. For example, one could askeach question here and discuss the issues in whether the claimed invention would havemore depth throughout this report. emerged in roughly the same time frame “but for” the prospect of a patent. Judgea. Is the Patent Warranted? Posner articulated this view as follows: The PTO must issue a patent unless [I]f a court thinks an invention forit can establish a prima facie case for which a patent is being sought wouldrejection of the patent application.63 Patent have been made as soon or almost aslaw establishes the standards of patentability soon as it was made even if thereagainst which the PTO measures a patent were no patent laws, it mustapplication. These standards ask whether pronounce the invention obvious andthe claimed invention is patentable subject the patent invalid.69matter64 that is novel,65 nonobvious,66 anduseful,67 and whether the application meets Analogously, one could ask whether otherthe disclosure requirements.68 measures through which patent law can encourage innovation – disclosure or 63 commercial development of an invention70 – In re Warner, 379 F.2d 1011, 1016 (C.C.P.A.1967), cert. denied, 389 U.S. 1057 (1968). would have occurred as soon “but for” the patent.71 64 35 U.S.C. § 101 (basically, processes,machines, manufactures, and compositions of matter). This question asks whether a patent 65 35 U.S.C. § 102. “The invention must . . . not is necessary to achieve one of the meansbe wholly anticipated by the so-called ‘prior art,’ or public through which the patent system encouragesdomain materials such as publications and other patents.” innovation. If not, then, in theory, a patentSCHECHTER & THOMAS, INTELLECTUAL PROPERTY: THELAW OF COPYRIGHTS, PATENTS AND TRADEMARKS § 13.1 at282 (footnote omitted). without undue experimentation. The patent application 66 35 U.S.C. § 103. “The nonobviousness must also contain distinct, definite claims that set out therequirement is met if the invention is beyond the ordinary proprietary interest asserted by the inventor.” SCHECHTERabilities of a skilled artisan knowledgeable in the & THOMAS, INTELLECTUAL PROPERTY: THE LAW OFappropriate field.” SCHECHTER & THOMAS, INTELLECTUAL COPYRIGHT , PATENTS AND TRADEMARKS § 13.1 at 282PROPERTY: THE LAW OF COPYRIGHTS, PATENTS AND (footnotes omitted.).TRADEMARKS § 13.1 at 282 (footnote omitted). 69 Roberts v. Sears, Roebuck & Co., 723 F.2d 67 35 U.S.C. § 101. “An invention is judged as 1324, 1346 (7th Cir. 1983) (Posner, J., dissenting fromuseful if it is minimally operable towards some practical judgment remanding for a new trial rather than finding thepurpose.” SCHECHTER & THOMAS, INTELLECTUAL claimed invention obvious as a matter of law).PROPERTY: THE LAW OF COPYRIGHTS, PATENTS AND 70TRADEMARKS § 13.1 at 282 (footnote omitted). See supra Ch. 1(I)(A)(2) and infra Ch. 2(I)(A) (discussing purposes of the patent law). 68 35 U.S.C. § 112. “Patent applications must 71include a specification that so completely describes the See generally infra Ch. 2(I), (III) (discussinginvention that skilled artisans are enabled to practice it purposes of patent law from economic perspective). 10
  • 40. should not be granted, because patents can for” approach to analyze whether individualimpose costs on the public.72 By disallowing patents should be granted.75 For example,a patent if it is not necessary to elicit an any property rights system must beinvention (or disclosure or commercial administrable; finding the answer to the “butdevelopment of the invention), this “but for” for” question in most individual cases wouldapproach would leave room for competition not be administrable.76 Instead, the morepolicy to spur innovation and provide manageable standards of the patent statuteconsumers with what they want at optimal have evolved to serve as the means by whichprices, quantity, and quality.73 to measure when to grant a patent. Nonetheless, for conceptual purposes, one From a theoretical perspective, the way to assess the alignment between“but for” approach represents the right way competition and patent law and policy, andto assess whether to grant a patent.74 It is to assess policy choices for an appropriatenot usually possible, however, to use a “but blend of competition and patents, is to examine whether the patent system’s standards of patentability ask questions 72 See, e.g., 1 HOVENKAMP ET AL., IP AND likely to produce results similar to thoseANTITRUST § 1.3a at 1-10 (“Because intellectual propertyrights impose costs on the public, the intellectual property obtained by asking the “but for” question.77laws can be justified by the public goods argument only tothe extent that the laws on balance encourage enough b. Does the Grant of the Patent Confercreation and dissemination of new works to offset thosecosts.” (Emphasis added.)); see also HARMON, PATENTS Market Power on the PatentholderAND THE FEDERAL CIRCUIT § 1.2 at 12 (costs to the public or Unnecessarily Increasecan include inflated prices); see infra Ch. 1(IV)(B)(5) Transaction Costs?(discussing process costs and costs of uncertainty tobusinesses). If an unwarranted patent confers 73 The Supreme Court’s decision in Bonito Boats market power on a patentholder, it canprovides an analogy. There, a unanimous Supreme Court deprive consumers of the benefits ofheld a Florida statute offering patent-like protection for aboat hull molding process to be preempted by the competition without compensating value.78Supremacy Clause. “By offering patent-like protection for Moreover, even if an unwarranted patentideas deemed unprotected under the present federalscheme, the Florida statute conflicts with the ‘strongfederal policy favoring free competition in ideas which do 75not merit patent protection.’” Bonito Boats, 489 U.S. at Most concede that the “but for” standard,168, citing Lear, Inc. v. Adkins, 395 U.S. 653, 656 (1969). although conceptually correct, cannot practically be appliedA “but for” approach analogously would protect free in individual cases. See generally infra Ch. 4(II)(A)(2).competition in areas where a patent was not necessary to 76elicit the invention (or its disclosure or commercial In many cases, it is likely unknowable whetherdevelopment). the claimed invention would have emerged in roughly the same time frame absent the prospect of a patent. Even if 74 Many view the perspective that patents should knowable, the costs of examining that question wouldbe granted only if the invention would not have emerged generally far outweigh any benefits from obtaining a more“but for” the patent system as the “defining proposition” precise measure of whether a patent should be granted.for standards of patentability. See, e.g., Merges 2/28 at 77579; Greenhall 2/27 at 421-22; Farrell 2/28 at 596-97; See generally infra Ch. 4(II)(A)(2) (discussingMusacchia 4/9 at 25-26; Scherer 7/10 at 54; Lunney 7/10 at standards of patentability in relation to “but for” question).97-104; Wamsley 7/10 at 139; Gambrell 10/25 at 41; 78Stoner 10/30 at 37; Kitch 10/30 at 50-51 (“but for” inquiry The issuance of invalid patents that do notthe right thing to think about as a matter of “metatheory”); confer market power may also raise societal costs even ifBarr 10/30 at 53. they do not raise competition issues. 11
  • 41. does not confer market power, a In addition, the issuance ofproliferation of trivial patents can harm unwarranted patents can injure competitioncompetition. even if they do not confer market power. A proliferation of trivial patents can be The patent system, quite properly, detrimental. Innovators “must expenddoes not examine whether the grant of a resources both in searching such patents topatent would likely create market power. As avoid infringement and in negotiating patentUnder Secretary of Commerce for licenses to use the technology. For patentsIntellectual Property and Director of the covering significant developments, it may beUSPTO James Rogan pointed out on the assumed that those additional social costsfirst day of the Hearings, “[p]atent are relatively small compared to the socialexamination does not include an analysis of benefits associated with the advance in thethe potential commercial impact of the art. But that assumption becomes lesspatent. It does not determine the relevant plausible if the advance is relatively trivial. .market in which the invention may be . . Thus, in aggregate, the search andmarketed or sold. No patent examiner transaction costs associated with numerousprojects the economies of scale to be trivial patents may outweigh the relativelyachieved through the invention.”79 In other small benefits associated with suchwords, patent examination does not include patents.”83an assessment of the likely competitivesignificance of a patent.80 2. Business Conduct with Respect to a Patent This is as it should be, especiallygiven the early point at which patent Patent policy generally asks twoapplications typically are filed.81 At that questions about antitrust enforcement inpoint, any attempt to assess the likely connection with a patent. The first questioncompetitive significance of a patent would is whether antitrust enforcement isusually devolve into mere speculation. warranted, or erroneously condemnsNonetheless, the likelihood of market power welfare-enhancing conduct that should beproblems may be greater in some areas of permitted. Antitrust enforcers, of course, arethe economy than others, and that increased also very concerned to ensure that antitrustlikelihood may justify closer scrutiny of enforcement properly distinguishes betweenpatent applications in those areas.82 anticompetitive and procompetitive conduct. The second question is whether antitrust enforcement limits the rights conferred by a 79 Rogan (stmt) 2. patent to undermine the incentives that the patent system creates. This is a question that 80 Id. also concerns antitrust enforcers. We 81 Hughes 2/28 at 611-12, 618. introduce each question and discuss both 82 See infra Chs. 3 (III), (IV) and 4 (II)(E)(business method patents; semiconductors (patent thicket)). issued at all or as soon).But see supra Ch. 1(I)(B) (antitrust does not object to 83patent that conveys market power if the patent is necessary MERGES & DUFFY , PATENT LAW AND POLICY:to elicit an invention that otherwise would likely not have CASES AND MATERIALS at 647. 12
  • 42. more thoroughly throughout the report. Licensing of Intellectual Property.88 The Guidelines outline a framework for antitrusta. Is Antitrust Enforcement analysis of licensing practices and identify Warranted? some of the efficiencies that businesses may seek through particular licensing practices.89 Antitrust law can constrain what apatentee can do with its patent, depending b. Does Antitrust Enforcementon the conduct at issue.84 A patentee may Undermine the Incentives Createduse a patent to obtain unwarranted market by the Patent System?power85 or interfere with competition in avariety of ways.86 The question for antitrust Antitrust scrutiny is more likely ifpolicymakers is how best to distinguish business conduct involves a patent thatbetween procompetitive and anticompetitive confers market power on the patentholderconduct with respect to patents. A proper than if the patent does not confer marketanswer depends in part on understanding the power. A patent that confers market power,role of patents in innovation and competition however, can fulfill precisely the goals ofin particular industries. As this report will the patent system: to preserve incentives todiscuss, patents play different roles in innovate.90 Patents thus present andifferent industries.87 Moreover, to avoid additional concern to antitrust enforcers:errors, antitrust enforcement needs to mistaken antitrust enforcement mayunderstand the efficiencies that businesses undermine the incentives the patent systemmay realize through particular types of creates. If patentees find that antitrustpatent-related conduct. The Antitrust enforcement unwarrantedly limits theirAgencies have addressed this in part by conduct with respect to their patents, thenissuing Antitrust Guidelines for the such enforcement may reduce incentives to invent. Thus, patent perspectives emphasize the need for antitrust enforcement to take care in distinguishing anticompetitive from 84 1 HOVENKAMP ET AL., IP AND ANTITRUST § procompetitive conduct, particularly when1.3b at 1-14. See also United States v. Microsoft Corp.,253 F.3d 34, 63 (2001) (The court rejected appellant’s the patent confers market power.assertion that because intellectual property rights have beenlawfully acquired, their subsequent use cannot result inantitrust liability. Id. “That is no more correct than theproposition that use of one’s personal property, such as abaseball bat, cannot give rise to tort liability.” Id.); Tom2/8 (Antitrust Session) at 53-54. 88 U.S. Department of Justice and Federal Trade 85 This description does not capture all of the Commission, Antitrust Guidelines for the Licensing ofpossible anticompetitive conduct, of course. For example, Intellectual Property (Apr. 6, 1995), reprinted in 4 Tradecertain limited conduct with respect to a patent may be Reg. Rep. (CCH) ¶ 13,132, available atsummarily condemned, without an examination of market http://www.usdoj.gov/atr/public/guidelines/ipguide.htmpower, due to its obvious anticompetitive effects. See (hereinafter IP Guidelines).Second Report (forthcoming). 89 Id. at § 3.3. How these Guidelines are 86 1 HOVENKAMP ET AL., IP AND ANTITRUST § working in practice will be discussed in the second,1.3b at 1-13. See Second Report (forthcoming). forthcoming report. 87 90 See generally infra Ch. 2(II)(A)(2) and Ch. 3. See supra Ch. 1(B). 13
  • 43. 3. Enhancing Consumer Welfare II. VIEWS ON HOW BEST TO Requires a Proper Balance of BALANCE COMPETITION Competition and Patent Law and AND PATENTS TO Policy ACHIEVE CONSUMER The Hearings record provides ample WELFARE HAVE VARIEDevidence of both the tensions and the WIDELY OVER TIMEpotential for greater congruence betweencompetition and patent law and policy. On A. For Much of the Twentieththe one hand, panelists noted antitrust’s Century, Patent andincreased appreciation of the role of patentsin fostering innovation and increased Antitrust Law Have Tradedunderstanding of the efficiencies to be Ascendency with Each Othergained through patent licensing and otherpractices.91 On the other hand, economists Despite the common goals of patentalso emphasized that ever greater intellectual and antitrust law, the doctrines historicallyproperty protection is not necessarily have traded ascendancy between eachsocially beneficial.92 Among other things, other.96 Broadly speaking, throughout muchstronger intellectual property protection of the twentieth century, courts and federalcarries the potential for less price agencies considered patents to confercompetition.93 From a broad policy monopoly power and, correspondingly,perspective, policymakers can maximize viewed antitrust as always opposed toconsumer welfare at a level of IP protection monopoly power.97 Some have argued thatcertainly greater than zero, but less than this perceived conflict led courts to believeabsolute.94 Because both competition and that, in any given case, they had to find thatintellectual property protection may foster either patents or antitrust took precedence.98innovation, these policy tools must be In general, when courts were favoringblended to achieve consumer welfare.95 patents, they were usually disfavoring antitrust, and vice versa. A variety of factors appear to have shaped these shifts, including perceptions about the power of big business, 91 See, e.g., James, Opening Day Comments at 1-2; Pitofsky 2/6 at 29-30; Tom 2/8 (Antitrust Session) at 47-50. 96 1 HOVENKAMP ET AL., IP AND ANTITRUST at § 92 See, e.g., Farrell 2/28 at 596-97; Langenfeld 1.3c at 1-15.2/20 at 10-13, 64. 97 See, e.g., Crown Die & Tool Co. v. Nye Tool 93 See, e.g., HARMON, PATENTS AND THE & Mach. Works, 261 U.S. 24, 37 (1923), citing ContinentalFEDERAL CIRCUIT § 1.2 at 12. Paper Bag Co. v. Eastern Paper Bag Co., 210 U.S. 405 (1908) (patents as monopolies); R. Hewitt Pate, Antitrust 94 See generally James Langenfeld, Intellectual and Intellectual Property, Before the American IntellectualProperty and Antitrust: Steps Toward Striking a Balance, Property Association, 2003 Mid-Winter Institute (Jan. 24,52 CASE W. RES. L. REV. 91, 96-98 (2001); William 2003), atLandes & Richard Posner, An Economic Analysis of http://www.usdoj.gov/atr/public/speeches/200701.pdf.Copyright Law, 18 J. LEGAL STUD. 325 (1989). 95 98 See generally infra Ch. 2(II), (III). Anthony, 28 AIPLA Q.J. at 4. 14
  • 44. the competitive significance of various scrutiny, 101 most typically because the “verypatent licensing practices, the nature and object of these [patent] laws is monopoly. . .role of patents, and the best ways to achieve .”102 Courts often seemed “to immunizeeconomic and technological growth. from antitrust scrutiny the conduct of firms holding patents,”103 even including patent1. 1890-1930: Patents Receive Little pools with outright price fixing.104 Some Antitrust Scrutiny contend that patent owners engaged in “what was arguably rather substantial Passage of the Sherman Act in 1890 overreaching” during this time by seeking to– one hundred years after passage of the first impose restrictions beyond the first sale of aPatent Act in 1790 – set the stage for courts patented product.105to begin construing how these two doctrinesshould interact. Although both patent and 2. 1930-1980: Antitrust Is Generallyantitrust have antecedents dating back Ascendantfarther than the enactment of those twostatutes,99 courts did not give significant An antitrust backlash began in 106attention to the intersection of patents and 1917, when the Supreme Court rejectedantitrust until the early 1900s.100 Early court on antitrust and patent misuse groundsopinions generally refrained from subjecting certain licensing restrictions that moviepatent-related conduct to antitrust exhibitors had imposed.107 By the 1930s, a stronger role for antitrust, and a correspondingly weaker role for patents, were emerging. During that time, some saw 99 Robert Merges and John Duffy point out thatAristotle discussed (and rejected) a proposal for a patent- 101 See, e.g., Bement v. National Harrow Co.,like system in the fourth century B.C.; they trace the history 186 U.S. 70 (1902); Heaton-Peninsular Button-Fastnerof the core concepts of patent law from that time through Co. v. Eureka Specialty Co., 77 F. 288 (6th Cir. 1896);the present. See MERGES & DUFFY , PATENT LAW AND Strait v. National Harrow Co., 51 F. 819 (N.D.N.Y. 1892).POLICY: CASES AND MATERIALS at 1-13. One can also find‘abuse of patent’ cases in England going back to 1600. 1 102 Bement v. National Harrow Co., 186 U.S. atHOVENKAMP ET AL., IP AND ANTITRUST § 1.3c at 1-14 & n. 91. See also Strait v. National Harrow Co., 51 F. 81910, citing Lewis Edmonds, THE LAW AND PRACTICE OF (N.D.N.Y. 1892); Tom 2/8 (Antitrust Session) at 38.LETTERS PATENT FOR INVENTIONS 7-8 (1890) (relating acase in which two people were stripped of patents and 103imprisoned for abusing their patent in the seventeenth Anthony, 28 AIPLA Q.J. at 5.century). Analogously, English courts wrestled with 104competition law early on, and, for example, rejected a Anthony, 28 AIPLA Q.J. at 5 (citing Bementmonopoly granted by Elizabeth I. The Case of v. National Harrow Co., 186 U.S. 70 (1902)).Monopolies, 11 Co. Rep. 84b, 77 Eng. Rep. 1260 (1603). 105Other competition law issues, such as restraint of trade 1 HOVENKAMP ET AL, IP AND ANTITRUST §cases, with parties demonstrating cartel behavior, were 1.3c at 1-15.brought as contract cases. Courts in England and the 106United States refused to uphold such contracts, long before Id.the Sherman Act was written. See generally John E.Lopatka, The Case for Legal Enforcement of Price Fixing 107 Motion Picture Patents Co. v. Universal FilmAgreements, 38 EMORY L.J. 1 (1989). Mfg. Co., 243 U.S. 502 (1917). Movie exhibitors mandated that their patented film projection equipment 100 1 HOVENKAMP ET AL., IP AND ANTITRUST § only be resold at a specified price, and that the projectors1.3c at 1-14 through 1-15. only be used with the licensor’s patented film. 15
  • 45. patent procedures as favoring “the powerful monopoly.”112 As this quotation and otherand the unscrupulous,” noting, among other cases made clear, courts generally continuedthings, the potential for “dragnet” patent to view patents as automatic sources ofapplications, through which firms could monopoly power.113amend their patent claims during lengthyprocedures at the patent office and thereby Around the same time, courts alsocapture competitors’ most recent weakened patent rights, “most notably bydevelopments.108 Others attacked patents imposing a high standard of ‘invention’ as amore broadly.109 Although not all condition of patentability.”114 For example,commentary was anti-patent,110 an “anti- in Cuno Engineering Corp. v. Automaticbusiness” tenor of the times apparently Devices Corp., the Supreme Court reversedcontributed to antitrust’s more active role in a lower court’s judgment that respondentconstraining patent-related conduct.111 had a valid patent that was infringed, reasoning that “the new device, however Another significant factor was the useful it may be, must reveal the flash ofstate of economic learning. Courts limited creative genius, not merely the skill of thethe scope of any exemption from antitrust calling. If it fails, it has not established itsfor patent-related conduct. The Supreme right to a private grant on the publicCourt ruled there was no exemption from domain.”115 The anti-patent posture of theantitrust “beyond the limits of the patent Supreme Court at that time led one dissenting U.S. Supreme Court Justice to observe that “the only patent that is valid is one which this Court has not been able to get 108 Alfred E. Kahn, Fundamental Deficiencies of 112American Patent Law, 30 AM. ECON. REV. 475, 485-86 United States v. Line Material Co., 333 U.S.(1940). See infra Ch. 4(II)(C)(1), for a discussion of the 287, 308 (1948). See also Morton Salt Co. v. G. S.current competitive significance of continuation procedures Suppiger Co., 314 U.S. 488, 492 (1942).at the PTO. See also FRITZ MACHLUP, AN ECONOMICREVIEW OF THE PATENT SYSTEM, SUBCOMM. ON PATENTS, 113 See, e.g., MERGES & DUFFY , PATENT LAWTRADEMARKS , & COPYRIGHTS OF THE SENATE COMM. ON AND POLICY: CASES AND MATERIALS at 1349 (“During theTHE JUDICIARY, 85TH CONG., 2D SESS. 40-42 (Comm. Print middle part of the twentieth century, the courts tended to1957) (discussing charges that “the patent system operates associate patents with monopolies, and hence to view themin favor of economic concentration and bigness”) as narrow exceptions to the nation’s antitrust laws. This(hereinafter MACHLUP, AN ECONOMIC REVIEW OF THE view [was] especially prominent in the Supreme CourtPATENT SYSTEM). cases from the 1930s until the 1960s, . . . .”). 109 See MACHLUP, AN ECONOMIC REVIEW OF THE 114 1 HOVENKAMP ET AL., IP AND ANTITRUST §PATENT SYSTEM at 28-29 (quoting from various articles 1.3c at 1-16; see also SCHECHTER & THOMAS,attacking patents to one degree or another). Machlup’s INTELLECTUAL PROPERTY: THE LAW OF COPYRIGHTS,report also outlines many arguments in favor of patents. PATENTS AND TRADEMARKS § 17.3.1 at 376 (“The anti- monopoly sentiments that arose during the Depression era 110 See, e.g., JOHN BATES CLARK, ESSENTIALS OF did not bode well for the patent system. Courts began toECONOMIC THEORY 360 (1927) (describing why inventors apply an increasingly stringent ‘invention’ standard thatneed patents to maintain incentives to innovate), cited in found most patents wanting.”).MACHLUP, AN ECONOMIC REVIEW OF THE PATENT SYSTEMat 37. 115 314 U.S. 84, 91 (1941) (emphasis added); see also Gerald Sobel, Patent Scope and Competition: Is the 111 See MERGES & DUFFY , PATENT LAW AND Federal Circuit’s Approach Correct?, 7 VA. J. OF LAW &POLICY: CASES AND MATERIALS at 10. TECH. 3, 16-17 (2002). 16
  • 46. Bo x 1-2 . R EPORT OF THE P R E S ID E N T ’ S C OMM ISSION ON THE P ATENT S Y ST EM (1966) “T O P ROMOTE T H E P R O G R E S S O F . . . U SEFUL A RTS ” In 1965, President Johnson established a Commission on the Patent System, which examined the patent system in light of six objectives: raising the quality and reliability of U.S. patents; shortening the period of patent pendency; accelerating the public disclosure of technological advances; reducing the expense of obtaining and litigating a patent; harmonizing the U.S. patent system with that of other major countries; and preparing the patent system to cope with future techno logical developm ents. In 1966, the Comm ission issued its report containing 35 recommendations that addressed a wide range of subject areas. Two areas of recom mendation are particularly noteworthy as they underscore society’s ongoing efforts to increase the value of patent disclosures and to decrease the possibility the system could be gamed so as to underm ine the value of those d isclosures. Publication. The Com mission concluded that early publication of patent applications “could prevent needless duplication of the disclosed work, promote additional technological advances based on the information disclosed, and apprise entrepreneurs of their potential liability.” They recommend ed publication of all pending app lications “eighteen to twenty-four m onths after its earliest effective filing date. . . .” Continuations. Continuations are one means by which claims can be broadened after publication. The difficulty continuations pose is that “unclaimed disclosures in a published application . . . might be protected by broader claims in [a] subsequently issued patent.” The Co mmission believed an absolute bar on continuations was no t feasible b ut, instead , recommended the impo sition of certain limits on an applicant’s ab ility to file continuations.its hands on.”116 each might function to achieve the purposes of the patent system.119 A 1966 Congress responded to these judicial Presidential Commission on the Patenttrends by passing the Patent Act of 1952, System endorsed patents as offering awhich limited the doctrine of patent misuse “unique service,” although the Commissionand strengthened the patent system.117 also recommended 35 changes to the patentCongress issued a lengthy study in 1957 on system. See Box 1-2 (1966 Presidential“The Patent System and the Modern Commission). Also in 1966, the SupremeEconomy,”118 consisting of reports prepared Court articulated an objective test forby a variety of experts. Most references to nonobviousness, based on Section 103 of thethe study note only the conclusion of Patent Act of 1952, which, over time, haseconomist Fritz Machlup that insufficient replaced the more subjective “invention”empirical economic evidence exists tojustify either abolishing or creating a patentsystem, but Machlup also provided useful 119 See MACHLUP, AN ECONOMIC REVIEW OF THEinsights about how patents and competition PATENT SYSTEM at 80 (“If we did not have a patent system, it would be irresponsible, on the basis of our present knowledge of its economic consequences, to recommend 116 instituting one. But since we have had a patent system for Jungerson v. Ostby & Barton Co., 335 U.S. a long time, it would be irresponsible, on the basis of our560, 572 (1949) (Jackson, J., dissenting). present knowledge, to recommend abolishing it.”); id. at 76-79 (discussing rationales for patent system, such as 117 35 U.S.C. §§ 1 et seq. providing incentives to invent, disclose, or invest, and comparing theories suggesting that competition alone could 118 STUDY OF THE SUBCOMM. ON PATENTS, serve those purposes with theories that patents areTRADEMARKS , AND COPYRIGHTS OF THE SENATE COMM. ON necessary in some cases to achieve those purposes); MergesTHE JUDICIARY, 84TH CONG., 2D SESS. (1957). 2/28 at 577-79. 17
  • 47. approach the Court had used earlier.120 lacked both a sound economic foundation and a sufficient appreciation of the Overall, however, antitrust incentives for innovation that patents anddominated and patents were disfavored patent licensing can provide.126during the 1960s and 70s.121 “Most litigatedpatents were held invalid during this B. 1980-1990: Congress and theperiod.”122 Courts of appeal “diverged Courts Strengthen Patents,widely both as to doctrine and basic atittudes and Antitrust Incorporatestoward patents.”123 Some contend that, forthat reason, “industry downplayed the an Updated Economicsignificance of patents.”124 Overzealous Frameworkantitrust enforcement culminated in theDepartment of Justice’s “Nine No-Nos,” a By the late 1970s, two factors werelist of nine licensing practices that the converging to reverse the cycle of antitrust’sJustice Department generally viewed as dominance and patents’ weakness. First,automatically illegal.125 Most now believe general concern about industrial stagnationthat antitrust’s ascendency during this period and a lack of significant technological innovation spurred reassessments of the patent system. Second, scholars, many 120 associated with the “Chicago School of See SCHECHTER & THOMAS, THE LAW OFCOPYRIGHTS, PATENTS AND TRADEMARKS , § 17.3.2 at 377- Economics,” spurred a general rethinking of80 (discussing Graham v. John Deere Co., 383 U.S. 1 antitrust, including its approach to patents.(1966) in relation to 35 U.S.C. § 103 (1952)); HARMON, Although these two factors were not the onlyPATENTS AND THE FEDERAL CIRCUIT § 4.2(a) at 137-39(same). ones that influenced trends in the 1980s, each played a central role. 121 See, e.g., Pate, Antitrust and IntellectualProperty at 6. 1. Congress and the Courts 122 1 HOVENKAMP ET AL., IP AND ANTITRUST Strengthen Patents§ 1.3c at 1-16. In 1971, the Second Circuit reported thatappeals courts found more than 80% of the patentsreviewed to be invalid. Carter-Wallace, Inc. v. Davis- a. Congress Creates the Court ofEdwards Pharmacal Corp., 443 F.2d. 867, 872 (2d Cir. Appeals for the Federal Circuit1971), cert. denied, 412 U.S. 929 (1973). See alsoMERGES & DUFFY , PATENT LAW AND POLICY: CASES ANDMATERIALS at 10 (during 1960s and early 1970s, “[i]t was In 1978, President Carter appointeddifficult to get a patent upheld in many federal courts”). an Advisory Committee to perform a domestic review of industrial innovation. 123 MERGES & DUFFY , PATENT LAW AND POLICY: See Box 1-3 (1979 Commission report onCASES AND MATERIALS at 10-11. See infra Ch. 6(II)(A). patents). Government officials and 124 MERGES & DUFFY , PATENT LAW AND POLICY: policymakers had grown concerned with anCASES AND MATERIALS at 11. overall weakening of R&D and other signs 125 See Bruce B. Wilson, Deputy AssistantAttorney General, Antitrust Division, Remarks before the 126Michigan State Bar Antitrust Law Section (September 21, See, e.g., Muris, The Way Ahead at 1; Pate,1972), reprinted in 5 CCH Trade Reg. Rep. 50,146 Antitrust and Intellectual Property at 7; Pitofsky 2/6 at 29-(transfer binder) (DOJ officials speech articulating what 30 (“‘Nine No-Nos . . .’ a far, far cry from where we arecame to be called the "Nine No-Nos"). today”). 18
  • 48. Bo x 1-3 . R EPORT OF THE I NDUSTRIAL S U B C O M M I T TE E F O R P A T E N T A N D I N F O R M A T IO N P O L I C Y OF T H E A DVISORY C O M M I T T E E O N I NDUSTRIAL I N N O V A T I O N: R E P O R T O N P ATENT P OLICY (1979) In 1978, President Carter appointed an Advisory Committee to perform a domestic review of industrial innovation. In 1979, the Patent and Information Policy Subcommittee (Subcomm ittee) of the Advisory Com mittee issued its R E P O RT O N P ATENT P OLICY . The report had five key proposals, four of which (at pages 153 -55) a re relev ant to this re port: Pro posal I. Upgrade the Patent and Trademark Office. The Subcom mittee observed that the quality of the patent system depends on the quality of the underlying PTO examination, which in turn depends upon “the search of the prior art by the examine r.” The Subco mmittee reco mmended the PT O b e “given the funds and resources to improve its examination procedure. . . .” Propo sal II. Pro vide fo r Ree xam ination of P atents. The Subcom mittee recognized the need for a “fast, inexpensive method for increasing the certainty as to the enforceability and scope of patents over prior art not considered by the PT O.” Und er the p roposal, any party co uld req uest reexamination d uring the life of the patent. The proposed system would provide for the submission of written arguments by interested parties and expedited review by the P TO to pro vide a “simple, inexpensive metho d” to impro ve the q uality of patents with dem onstrated co mmercial value. Propo sal III. Provide a Specialized Appellate Court for Patent Cases. The Subcommittee adv ocated, “a centralized national court with exclusive appellate jurisdiction (sub ject to S upreme C ourt review) o ver patent- related cases as a vehicle for insuring more uniform interpretation of the patent laws.” Propo sal IV. Reduce Cost of Patent Litigation. The Subcom mittee concluded that decreasing “the cost and time involved in resolving patent infringement and validity disputes through litigation” would improve the patent system. The Subcommittee urged, among other things, that Federal courts “exercise a high degree of control over the conduct of patent litigation, with particular concern for the time and expense of discovery.”of economic trouble. One question for the The Report on Patent Policy thatAdvisory Committee was whether, and to emerged recommended “a centralizedwhat extent, patent policies contributed to national court with exclusive appellatethese circumstances. Judge Newman, a jurisdiction (subject to Supreme Courtmember of the Advisory Committee, review) over patent-related cases as arecalled the “low point” at which they found vehicle for insuring more uniformthe U.S. economy: “Investment in basic interpretation of the patent laws.”129science in applied research had disappeared . Committee members concluded that. . . Our production in the United States was increased uniformity and reliability in patentno longer competitive. Old technologies decisions would “contribute meaningfully towere stagnant. New [technologies] were decisions to file patent applications and todormant. . . .”127 Among other problems, commercialize invention, thereby improvingCommittee members attributed these industrial innovation[.]”130 During theconditions to “a diminished patent 1970s, others also discussed the problem ofincentive” in the U.S..128 129 REPORT OF THE INDUSTRIAL SUBCOMM. FOR PATENT AND INFORMATION POLICY OF THE ADVISORY COMM. ON INDUSTRIAL INNOVATION, REPORT ON PATENT 127 Newman 2/6 at 39-42. POLICY 155 (1979). 128 130 Id. Id. 19
  • 49. Bo x 1-4 . Congressional Activity. Congress has actively legislated in the field of patent law since 19 80. C ongress passed amendme nts to the Patent Act in 1980 that required the payment of maintenance fees (35 U.S.C. § 154), made provision for third parties to cite prior art to the PTO (35 U.S.C. § 30 1) and created reexamination procedures (35 U .S.C. § 302). In 1982, Congress created the Court of Appeals for the Federal Circuit (35 U.S.C. § 141, 28 U.S.C. § 1295). Two years later, Congress passed the Hatch Waxman Amendments which, among other things, permitted an extension of patent term to compensate for delay in securing FDA approval to sell new drugs for humans (35 U.S.C. § 156) and consolidated the Boards of Patent App eals and Interferences into a single Board of Patent App eals and Interferences (35 U.S.C. § 14 1). In 1988, Congress passed the Patent Misuse Reform Act, which governs the application of certain patent misuse defenses to patent infringement claims (35 U.S.C. § 271(d)). In 1994, Congress passed the Uruguay Round Agreements Act which, among other reforms, changed the patent term to 20 years from the earliest date on which the application was filed (35 U.S.C. § 154) and allowed the filing of provisional applications (35 U.S.C. § 111). Congress passed the American Inventors Protection Act in 1999, which made multiple changes to patent law, such as requiring that patent applications be published after 18 months where equivalent applications are published abroad (35 U .S.C. § 122), creating an inter partes reexamination procedure (35 U.S.C. §§ 311-313), and granting a prior user right for business method patents (35 U.S.C. § 273). Congress reformed certain features of the inter partes reexa minatio n pro cedure in 2002 .significant inconsistency in patent decisions creation of the Federal Circuit was aand the idea that a centralized appellate watershed event in the history of the U.S.court for patent matters might ease that patent system.”134 Most commentatorsproblem.131 find that, as a general matter, the Federal Circuit strengthened patent rights In 1982, Congress created the Court significantly,135 upholding patent validityof Appeals for the Federal Circuit.132 TheFederal Circuit has exclusive jurisdiction ofall appeals from final district court decisions of one of the well-pleaded claims,” such that “plaintiff’s right to relief necessarily depends on resolution of ain civil actions “arising under any Act of substantial question of federal patent law” (generallyCongress relating to patents.”133 “[T]he referred to as “substantial question” jurisdiction). Christianson v. Colt, 486 U.S. 800 (1988). In Holmes Group, Inc. v. Vornado Air Circulation Systems, Inc., 535 U.S. 826 (2002), the Supreme Court concluded that 131 See, e.g., MERGES & DUFFY , PATENT LAW “arising under” jurisdiction does not give jurisdiction to theAND POLICY: CASES AND MATERIALS at 11 (during 1970s, Federal Circuit, when only a patent counterclaim exists,idea of a single, unified court of appeals for patent cases and the complaint does not assert any claim arising underdiscussed as “way to return patents to a more central federal patent law. See generally infra Ch. 6(II)(B)(1)(a).position in the commercial world”). 134 MERGES & DUFFY , PATENT LAW AND POLICY: 132 28 U.S.C. § 1295. The United States Court of CASES AND MATERIALS at 11.Appeals for the Federal Circuit was created through themerging of two specialized courts of limited subject matter 135 “Since the creation of the Federal Circuit, . . .but nationwide jurisdiction – the U.S. Court of Claims and [i]t is also much easier to get an injunction against anthe U.S. Court of Customs and Patent Appeals. infringer. And money damages have soared too, both on average and in the highest-visibility cases.” MERGES & 133 28 U.S.C. § 1338(a). The Supreme Court has DUFFY , PATENT LAW AND POLICY: CASES AND MATERIALSinterpreted the “arising under” clause to require a at 11. Cf. ROBERT L. HARMON, PATENTS AND THEdetermination of whether the patent allegation forms part of FEDERAL CIRCUIT 147, 161 (Supp. 2002) (originalthe “well-pleaded complaint,” in that patent law either (1) assessment of Federal Circuit as pro-patentee; more“creates the cause of action” (generally referred to as recently, court is moving toward “a more neutral“arising under” jurisdiction), or (2) is a “necessary element position”). 20
  • 50. “more frequently than in the anti-patent era characteristics from any found in nature andof the 30s to the 70s.”136 one having the potential for significant utility.”141 “His discovery is not nature’sb. The Supreme Court Interprets handiwork, but his own; accordingly it is Patentable Subject Matter Broadly patentable subject matter under § 101.”142 The description of patentable subject matter In Diamond v. Chakrabarty,137 the as “anything under the sun that is made bySupreme Court held that a live, human-made man” conveyed a broad sense of themicroorganism was patentable subject potential scope of patents and, in particular,matter under 35 U.S.C. § 101.138 In reaching provided a significant boost to the biotechthis decision, the Court noted that the industry. Indeed, Hearings participants fromCommittee Reports accompanying the 1952 the biotech industry generally credited thePatent Act “inform us that Congress Court’s decision in Chakrabarty as theintended statutory subject matter to ‘include beginning of their industry, without whichanything under the sun that is made by genetic engineering would not have mademan.’”139 The Court distinguished “laws of nearly as much progress.143nature, physical phenomena, and abstractideas,” which have been held not In the 1981 case, Diamond v.patentable,140 from the patentee’s “new Diehr,144 the Supreme Court held that abacterium with markedly different process claim that included use of a computer program was patentable subject matter, concluding that “a claim drawn to 136 “Since the creation of the Federal Circuit, subject matter otherwise statutory does notpatents have been held valid more frequently than in theanti-patent era of the 30s to the 70s.” MERGES & DUFFY , become nonstatutory simply because it usesPATENT LAW AND POLICY: CASES AND MATERIALS at 11 a mathematical formula, computer program,(footnotes omitted). See also Robert P. Merges,Commercial Success and Patent Standards: EconomicPerspectives on Innovation, 76 CAL. L. REV. 803, 820-21(comparing pre- and post-Federal Circuit era statistics);John R. Allison & Mark A. Lemley, Empirical Evidence on 141 447 U.S. at 310.the Validity of Litigated Patents, 26 AIPLA Q. J. 185(1998) (providing a statistical survey of validity decisions 142 Id. The Court’s decision was 5-4, withfrom 1989-1996). See generally infra Ch. 6(II)(C). Justice Brennan writing the dissent, which described the patent laws as “attempt[ing] to reconcile this Nation’s 137 447 U.S. 303 (1980). deep-seated antipathy to monopolies with the need to encourage progress[,]” and argued that both congressional 138 Section 101 states that “[w]hoever invents or intent and public policy militated against finding that livingdiscovers any new or useful process, machine, organisms could be patentable subject matter. Id. at 318-manufacture, or composition of matter, or any new or 19.useful improvement thereof, may obtain a patent therefor,subject to the conditions and requirements of this title.” 35 143 Kirschner 2/26 at 239; Beier 2/26 at 322;U.S.C. § 101 (emphasis added). Chen 2/28 at 628; American Bar Association Section of Intellectual Property Law, Statement of Robert P. Taylor 139 447 U.S. at 309, citing S. REP. NO. 82-1979, on Behalf of Section of Intellectual Property Law Americanat 5 (1952); H.R. REP. NO. 82-1923, at 6 (1952). Bar Association on Competition and Intellectual Property Law and Policy In the Knowledge-Based Economy 140 447 U.S. at 309. “[A]n application of a law (7/11/02) 7, atof nature or mathematical formula to a known structure or http://www.ftc.gov/opp/intellect/020711robertptaylor.pdf.process may well be deserving of patent protection[,]” 144however. Diamond v. Diehr, 450 U.S. 175, 187 (1981). 450 U.S. 175 (1981). 21
  • 51. or digital computer.”145 The Court “view[ed well as possible anticompetitive effects,the patentees’] claims as nothing more than associated with particular business conduct.a process for molding rubber products and Over time, the courts and agencies havenot as an attempt to patent a mathematical largely adopted this updated economicformula.”146 The Court repeated the framework.149observation that patentable subject mattercan “include anything under the sun that is New economic learning led to amade by man,”147 and once again conveyed a more complex and pro-patent understandingbroad sense of the potential scope of patents. of how antitrust should view conduct with respect to patents. In 1981, Antitrust2. Antitrust Incorporates an Updated Division Deputy Assistant Attorney General Economic Framework Abbott B. Lipsky, Jr., renounced the Nine No-Nos as “contain[ing] more error than During the 1970s, change was accuracy,” and reviewed in some detail thebrewing in antitrust as well. Debates took possible efficiency justifications for eachplace between those who focused primarily licensing practice that the Nine No-Noson market structure and market power, and previously had condemned automatically. 150those who, broadly speaking, saw more The then-Chief of the Intellectual Propertyefficiencies than market power in the U.S. Section of the Antitrust Division, Rogereconomy.148 The new economic learning, Andewelt described how patents can benefitassociated with many of those who were competition:called “Chicago School” economists andlawyers, brought an updated economic The availability of exclusive patentframework to antitrust that, among other rights increases the possible rewardthings, emphasized the importance of for R&D. It thereby results in theseeking to understand the efficiencies, as development of some inventions that otherwise would not have been discovered or developed at all or, at 145 Id. at 187. Once again, the decision was 5-4, least, not nearly as early as theywith the dissent by Justice Stevens noting, among otherthings, that it was not at all clear that patent protection was were. For such inventions it isessential for the growth of the software industry. Id. at illogical to talk in terms of the patent217. For discussion of the ongoing controversy about therole of patents in the computer industry, see infra Ch.3(III), (IV). 149 See generally William E. Kovacic & Carl 146 450 U.S. at 191. The dissenting justices Shapiro, Antitrust Policy: A Century of Economic andviewed the case as having greater implications for the Legal Thinking, 14 J. ECON. PERSPECTIVES 43, 54-55patentability of computer programs generally and would (2000). One of the first decisions to use the new learninghave adopted a much narrower rule for when a program- was Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S.related invention could constitute patentable subject matter. 36 (1977), which upheld a supplier’s restriction on the Id. at 219. geographic area in which its distributor could sell. The Supreme Court found that consumers would benefit from a 147 Id. at 182. restriction on competition that prevented competitors from “free riding” on a firm’s promotional efforts. Id. at 54-55. 148 See generally Harold Demsetz, Two Systems 150of Belief about Monopoly, in INDUSTRIAL CONCENTRATION: Abbott B. Lipsky, Jr., Current AntitrustTHE NEW LEARNING 164-84 (Harvey J. Goldschmid et al. Division Views on Patent Licensing Practices, 50eds. 1974). ANTITRUST L.J. 515, 517-24 (1981). 22
  • 52. grant conflicting with a competitive Thus, by the end of the 1980s, economic system. If there were no congressional and court-driven changes had patent grant these inventions would significantly strengthened patents. not have reached the marketplace; Antitrust’s incorporation of updated therefore, the availability of a patent economic thinking led to a generally more served only to benefit competition – favorable view of how conduct with respect to make additional or less expensive to patents influences competition. This choices available to consumers.151 incorporation of economics held the potential for both competition and patentIn 1985, Deputy Assistant Attorney General policy to develop a greater integration andfor Antitrust Charles F. Rule commented on balance.prior failures of the courts and theDepartment of Justice “to recognize some III. COMPETITION ANDfundamental facts about the nature of PATENT POLICYintellectual property and the beneficial role CONTINUE TO SEEK Athat technology licensing plays in a healthy,competitive economy.”152 The 1988 DOJ PROPER BALANCE, ANDAntitrust Enforcement Guidelines for GROWTH OF THEInternational Operations elaborated on these KNOWLEDGE-BASEDearlier policy statements with a section on ECONOMY ADDS NEWintellectual property licensing arrangements CHALLENGESthat outlined consumer benefits fromintellectual property licensing153 andspecifically adopted a rule of reason A. Antitrust and Patent Policyapproach to intellectual property licensing Have Worked to Achieveissues, absent sham. Better Balance 1. Antitrust Policy Has Continued to 151 Roger B. Andewelt, Basic Principles to Apply Implement New Economicat the Patent-Antitrust Interface, Remarks to the Houston Learning in Addressing thePatent Law Association 4-5 (Dec. 3, 1981). Intersection of Antitrust and 152 Charles F. Rule, Technology Licensing and Patentsthe Second American Revolution: Storming the Rampartsof Antitrust and Misuse, Before the John Marshall LawSchool 5 (Feb. 22, 1985). Rule emphasized the role of Antitrust policymakers and enforcerspatents in preventing free riding: “Unless the ‘free rider’ continue to apply the new economic learningproblem is somehow addressed, those who might otherwise that gained precedence in the 1980s. Inundertake risky and expensive R&D will not do so. Fewertechnologies will be developed and consumers will face 1995, the Antitrust Division of thehigher prices and fewer choices.” Id. at 6. See also Department of Justice and the Federal TradeCharles F. Rule, The Antitrust Implications of International Commission jointly issued AntitrustLicensing: After the Nine No-No’s, Remarks before theLegal Conference sponsored by the World Trade Guidelines for the Licensing of IntellectualAssociation and the Cincinnati Patent Law Association Property (IP Guidelines). Like the 1988(Oct. 21, 1986), reprinted in 4 Trade Reg. Rep. (CCH) Guidelines, the 1995 IP Guidelines identify¶ 13,131. 3, 1981. and discuss potential efficiencies associated 153 1988 International Guidelines at §§ 3.6, 3.61. with many licensing practices and 23
  • 53. emphasize that the vast majority of licensing Second, “the Agencies do notpractices are analyzed under the rule of presume that intellectual property createsreason.154 market power in the antitrust context[.]”159 This observation eliminates the automatic The IP Guidelines “embody three conflict between patents and antitrust thatgeneral principles[.]”155 The first is that “for courts perceived by assuming that patentsthe purpose of antitrust analysis, the always create monopoly power in the handsAgencies regard intellectual property as of the patent holder. As noted earlier,being essentially comparable to any other patents may enable the holder to exerciseform of property[.]”156 Some have expressed market power,160 but the Antitrust Agenciesconcern that this statement may mean that do not assume that is necessarily the case.antitrust sees no difference betweenintellectual property and other types of Third, “the Agencies recognize thatproperty. 157 The IP Guidelines themselves intellectual property licensing allows firmsbelie this characterization, explaining that: to combine complementary factors of production and is generally [i]ntellectual property has important procompetitive.”161 The IP Guidelines characteristics, such as ease of explicitly recognize the efficiencies that misappropriation, that distinguish it firms can gain through intellectual property from many other forms of property. licensing, which can “benefit[] consumers These characteristics can be taken through the reduction of costs and the into account by standard antitrust introduction of new products.”162 Further, analysis, however, and do not require the IP Guidelines note that, “[b]y potentially the application of fundamentally increasing the expected returns from different principles. [footnote intellectual property, licensing also can omitted]158 increase the incentive for its creation and thus promote greater investment in research 154 and development.”163 Similarly, the IP The 1995 IP Guidelines superceded the 1988International Guidelines. The 1988 International Guidelines note that “various forms ofGuidelines specified that “[b]ecause they hold significantprocompetitive potential, unless the underlying transfer oftechnology is a sham, the Department analyzes restrictionsin intellectual property licensing arrangements under a ruleof reason [footnote omitted].” § 3.62. The 1995 Guidelinesprovide for a slightly greater possibility of per se treatment, property may vary substantially, and that “[t]he greater orsee IP Guidelines § 3.4, but still make clear that the lesser legal power of an owner to exclude others is alsoAgencies use the rule of reason “[i]n the vast majority of taken into account by standard antitrust analysis.” IPcases.” IP Guidelines § 3.4. Guidelines § 2.1, n. 9. 159 155 IP Guidelines § 2.0. IP Guidelines § 2.0. 160 156 Id. See supra Ch. 1(I)(C)(1) . 161 157 See, e.g., Langenfeld 2/20 at 6-8. IP Guidelines § 2.1. 162 158 IP Guidelines § 2.1. See Tom 2/8 (Antitrust Id. § 2.3.Session) at 50-52. The IP Guidelines further note that the 163power to exclude others from the use of intellectual Id. 24
  • 54. exclusivity”164 can give a licensee the the dissemination of technology.167 In eachincentive to invest in commercializing and case, based on the descriptions of the patentdistributing products that embody the pools the parties provided, the Antitrustintellectual property by “protecting the Division declined to initiate enforcementlicensee against free-riding on the licensee’s action.168investments by other licensees or by thelicensor.”165 Overall, the 1995 IP The FTC challenged one patent pool;Guidelines, like the 1988 International the allegations were resolved through aGuidelines, signal an approach that is far consent order that bars continuation of themore positive toward patent licensing than pooling arrangement.169 The FTC’searlier antitrust perspectives. challenge elicited controversy, especially with regard to what the facts actually In the same vein, since 1997, the showed, but the FTC’s complaint provides aAntitrust Division of the Department of useful comparison to the types ofJustice has issued four Business Review arrangements reviewed by the AntitrustLetters that analyze the antitrust issues Division to reveal which types of patentraised by the proposed patent pools and pools are more likely to pose significantdiscuss the features that reduce competitive antitrust concerns.170 Once again, theconcerns about those pools.166 Each letter Antitrust Agencies have viewed patent poolsexplicitly recognizes that patent pools can that offer legitimate efficiencies far moreprovide competitive benefits by promoting favorably than in the past. 164 The guidelines give as examples field-of-useand territorial restrictions. IP Guidelines § 2.3. 165 Id. 167 See MPEG Pool Letter at 5; Phillips DVD 166 See Letter from Joel I. Klein, Acting Assistant Pool Letter at 5; Hitachi DVD Pool Letter at 5.Attorney General, Antitrust Division, Department of 168Justice, Letter to Garrard R. Beeney, Esq. (June 26, 1997), See MPEG Pool Letter at 9-10; Phillips DVDat http://www.usdoj.gov/atr/public/busreview/1170.htm Pool Letter at 9; Hitachi DVD Pool Letter at 10.(hereinafter MPEG Pool Letter); Letter from Joel I. Klein, 169Assistant Attorney General, Antitrust Division, Department See In re Summit Tech., Inc. & VISX, Inc.,of Justice, to Garrard R. Beeney, Esq. (Dec. 16, 1998), at No. 9286 (FTC Mar. 24, 1998) (complaint), available athttp://www.usdoj.gov/atr/public/busreview/2121.htm http://www.ftc.gov/os/1998/03/summit.cmp.htm (Summit(hereinafter Phillips DVD Pool Letter); Letter from Joel I. Complaint); In re Summit Tech., Inc. & VISX, Inc., No.Klein, Assistant Attorney General, Department of Justice, 9286 (FTC Aug. 21, 1998) (Agreement Containingto Carey R. Ramos, Esq., counsel to Hitachi, Ltd. (June 10, Consent Order To Cease And Desist As To Summit Tech.,1999), at Inc.), available athttp://www.usdoj.gov/atr/public/busreview/2485.htm http://www.ftc.gov/os/1998/08/d09286suagr.htm; In re(hereinafter Hitachi DVD Pool Letter); Letter from Charles Summit Tech., Inc. & VISX, Inc., No. 9286 (FTC Aug. 21,A. James, Assistant Attorney General, Antitrust Division, 1998) (Agreement Containing Consent Order To CeaseDepartment of Justice, to Ky P. Ewing, Esq. (Nov. 12, And Desist As To VISX, Inc.), available at2002), at http://www.ftc.gov/os/1998/08/d09286viagr.htm.http://www.usdoj.gov/atr/public/busreview/200455.htm. 170See also James, Opening Day Comments at 2. See Anthony, 28 AIPLA Q.J. at 18-19. 25
  • 55. 2. Patent Policy Has Implemented created by submarine patents, and partly to Certain Reforms and Rules that conform U.S. practice to international Can Lessen Anticompetitive practice, the AIPA now requires publication Conduct and Increase Competition of a patent application eighteen months after filing, unless the applicant certifies that thea. Congress Enacted the American invention will not be the subject of any Inventors Protection Act of 1999 foreign or international application in (AIPA) jurisdictions that provide for eighteen-month publication.173 The PTO reports that roughly(i). Disclosure of Most Patent 90 percent of all pending patent applications Applications after Eighteen Months are published at eighteen months.174 This Can Reduce Opportunistic Hold Ups new publication requirement can assist through Submarine Patents inventors and businesses to some extent in avoiding hold up and making more informed Over the years, companies have decisions about where (and where not) tocomplained about problems caused by spend R&D resources.175 As several“submarine patents.” The basic scenario isthat a patent applicant allows its applicationto languish in the PTO while watching 173 35 U.S.C. § 122. See John Love 2/28 at 647another company make substantial (18-month publication part of AIPA is response to probleminvestments in a technology or product that of submarine patents); MERGES & DUFFY , PATENT LAW AND POLICY: CASES AND MATERIALS at 62-63 (AIPAwill infringe the yet-to-be-issued patent. brings United States into conformity with many otherOnce the other company’s sunk costs are countries that require publication of application eighteen months after filing).large, the patent applicant obtains the patent,asserts infringement, and “holds up” the 174 John Love 2/28 at 647.other company, demanding supra- 175competitive royalties for a license to the Gable 3/20 at 118-19 (describing how AIPA can “go a long way” to prevent plight of individual small“submarine patent.”171 The company must inventors who “put a lot of money and a lot of effort intoagree to supra-competitive royalties or this process and two or three years down the line, typically,forego its production or innovation. As a in the course of the prosecution of their own patent they found out another patent has issued that covers theirresult, consumers will either pay higher invention and they’re barred from using it.”); John Loveprices for the company’s goods, or will 2/28 at 647 (AIPA gives businesses “an idea of what patentnever get the benefit of the innovation that applications are pending and . . . an indication of where the technology is going also.”); Ronald Myrick, FTC/DOJthe company had to abandon.172 Hearings on IP and Antitrust: Testimony of Ronald Myrick (3/19/02) 20 (eliminating the right to opt-out of application Partly in response to problems publication at eighteen months would “partially eliminate the potential for ‘surprise’ or ‘hold up’ about which some have expressed concern.”), at http://www.ftc.gov/opp/intellect/020319ronmyrickprepared testimony.pdf. But see Thomas 4/10 at 192-93 (asserting 171 See, e.g., Shapiro 11/6 at 15-16, 175-76; see that the AIPA provides no benefit, because it requires onlyalso infra Ch. 2(B)(3)(b)(1) (discussing hold up in the the publication of patent applications that already arecontext of patent thickets) and Second Report published in Europe at the same time; may "save[] a(forthcoming). translation fee on occasion"); Katsh 4/10 at 193 (noting that 18-month publication does not give notice of what 172 See infra at Ch. 2(B)(3)(b)(1) (discussing this additional claims will be sought in continuation practice;scenario in more detail). 18-month publication does not give complete notice and 26
  • 56. panelists noted, patent disclosures may participation by third parties, however;181 forstimulate competition to design around a the most part, it is conducted ex parte in thepatent.176 same manner as the initial patent examination. It has not become a substitute(ii). Patent Quality: Reexaminations of for patent litigation, and some argue that it Questionable Patents has been used as frequently by patentees to strengthen their patents as by challengers to Patent prosecutions – that is, the weed out invalid patents.182administrative procedures through which apatent application becomes a patent – take To afford a greater opportunity forplace ex parte. At the PTO, only the patent third-party participation in theapplicant and the patent examiner(s) discuss reexamination process, Congress enacted anthe patent application; no third parties are inter partes reexamination system as part ofinvolved in that discussion.177 In 1980, the AIPA.183 See also Box 1-5 (1992Congress established an ex parte Advisory Commission on Patent Lawreexamination procedure178 intended to Reform). Due to certain limitations,“strengthen[] investor confidence in the however, third parties had used thecertainty of patent rights by creating a procedure only four times since itssystem of administrative reexamination of enactment, as of the date of the Hearings.184doubtful patents.”179 Congress hoped this In the fall of 2002, Congress revised thereexamination procedure would allow an procedure in hopes that third parties wouldefficient resolution of questions of patent use it more frequently.185 Some remainvalidity and thus would obviate to some skeptical that the revisions were sufficient toextent the need for lengthy and costly patent encourage greater use of the procedurelitigation.180 This ex parte reexamination because one important disincentive to its useprocedure affords little opportunity for remains.186 Nevertheless, the availability of 181 35 U.S.C. §§ 304, 307 (if a third party hasleaves uncertainty). requested the reexamination, it has a right to reply to the patentee’s opening statement on the reexamination issue, 176 See, e.g., Banner 10/30 at 70-71 (discussing but no right to participate beyond that).design-around competition that disclosure of issued patents 182spurs); see generally infra at Ch. 3(III)(D)(1)(b) and Ch. See Mowery 2/27 at 408; Hall 2/28 at 760-61;5(II)(C)(4) for discussion of limitations on the role of Merrill 10/25 at 123; See MERGES & DUFFY , PATENT LAWpatent application disclosures at eighteen months and AND POLICY: CASES AND MATERIALS at 1210-11 (“thedisclosures of issued patents in facilitating business confirmation of a patent in reexamination is accorded aplanning and encouraging design-around competition. great deal of respect by courts, and hence a reexamination can bolster the ‘strength’ of a patent.”). 177 See infra Ch. 5(II)(B). 183 35 U.S.C. §§ 311-18. 178 35 U.S.C. §§ 302-07. 184 See Kunin 7/10 at 70. 179 H.R. REP. NO. 96-1307, at 3, reprinted in 1851980 U.S.C.C.A.N. 6460, 6462. 35 U.S.C. § 315(b) (as amended Nov. 2, 2002). 180 H.R. REP. NO. 96-1307, at 3, reprinted in 1861980 U.S.C.C.A.N. at 6463. See infra Ch. 5(III)(A), (B)(1). 27
  • 57. Bo x 1-5 . T HE A DVISORY C O M M I S S IO N ON P ATENT L A W R E F O R M : A R EPORT TO THE SE C R E T A R Y OF C OMM ERCE (1992) In 1990, Secretary of Commerce R obert Mo sbacher established an Advisory Commission regarding the state of the patent system and the need for any reform. In 1992, the Commission issued its report containing recommendations in three areas: (1) Harmonization-Related Issues. The Commission recom mended publication o f all patent applications within 24 m onths fro m the earliest priority date claimed by the applicant, along w ith certain protections for pa tentees. (2) Patent Enforcement-Related Issues. Litigation. The Commission recognized that “an essential relationship [exists] between the value of patent rights [ ] and the cost of patent litigation.” More specifically, the Commission sought “to ensure that transactional co sts do not pre judice the rights of patentees or the rights of the p ublic through the process of patent enforcement.” The Commission also noted that “[i]ncreased quality of examination will strengthen the p resum ption of validity, which in turn will dec rease the num ber o f unwarranted challenges to patent validity. This will also increase the confidence of the courts in applying the statutory presumption of validity.” Alternatives to Litigation. The Commission advocated modifications to the reexamination system “to provide third parties with a greater opportunity to participate.” For example, it recommended that the basis and scope of reexamination include all aspects of 35 U.S.C. § 112 (i.e., written description, enablement, claim definiteness), except best mod e. (3) Unique Issues Facing the Patent System. This series of recommendations addressed issues ranging from the protection of computer-related inventions to PT O funding through user fees. Recomm endations relevant to the FT C/D OJ Hearings are addressed else where in this report.inter partes reexamination adds a new important implications. Consistency in themechanism through which to address application of the law can reduce the costscompetition concerns about the validity of of business uncertainty and can facilitatepatents associated with market power.187 business planning about how best to compete.b. The Federal Circuit Has Increased Business Certainty and Has Noted In addition, in different contexts, the Competition Concerns in Certain Federal Circuit has interpreted the statute in Contexts ways that support the “notice function” of patents189 – that is, the requirement that a Many panelists at the Hearingsagreed that the Federal Circuit has increasedconsistency in the application of many law); Armbrecht 3/19 at 54-55 (informal poll of membersaspects of patent law.188 This trend has of Industrial Research Institute found general agreement that the Federal Circuit has brought greater stability and predictability to the patent process across industries); Weil 7/11 at 150-51 (Federal Circuit has brought consistency to 187 See also infra Ch. 5(III)(B), (C) for many areas of patent law). But see generally Ch. 6(II)(C)discussion of pros and cons of reexamination and post- (discussion of areas in which some argue that the Federalgrant opposition proceedings as means to address Circuit has increased uncertainty or other costs of thequestionable patents. litigation process). 188 189 See, e.g., Mossinghoff 2/6 at 76-78 (adds See, e.g., Athletic Alternatives, Inc. v. Princecertainty and consistency); Myrick 3/19 at 17 (credits Manufacturing, Inc., 73 F.3d 1573, 1581 (Fed. Cir. 1996)Federal Circuit with uniformity and certainty of patent (“Where there is an equal choice between a broader and 28
  • 58. patent’s “specification shall conclude with general attentiveness to the role of notice inone or more claims particularly pointing out ensuring that competitors know what aand distinctly claiming the subject matter patent does and does not protect serves towhich the applicant regards as his encourage and protect competition outsideinvention.”190 The notice function serves an the scope of a valid patent.important purpose in the framework ofcompetition: “The object of the patent law c. The PTO Has Implemented Certainin requiring the patentee [to distinctly claim Reforms that Can Aid Competitionhis invention] is not only to secure to him allto which he is entitled, but to apprise the (i). Utility Guidelinespublic of what is still open to them.”191Accurate notice of the scope of a patent’s A claimed invention must beclaims can encourage competition in the area “useful” to receive a patent. From time tonot covered by the patent. Although the time, some have raised concerns aboutSupreme Court in one context has found that whether patents have been granted forthe interest in ensuring appropriate research “too close to the laboratory bench”incentives for innovation can override the – that is, basic research not yet “useful”notice function,192 the Federal Circuit’s enough to deserve a patent.193 During the 1990s, some raised this concern with regard to biotech patents in particular.194 The PTOnarrower meaning of a claim, and there is an enablingdisclosure that indicates that the applicant is at least responded by issuing and then revising a setentitled to a claim having the narrower meaning, we of Utility Examination Guidelines,195 which,consider the notice function of the claim to be best served ultimately, have been generally well-by adopting the narrower meaning.”). 190 35 U.S.C. § 112. 191 McClain v. Ortmayer, 141 U.S. 419, 424 concerns with the doctrine of equivalents . . . are not new.(1891) (emphasis added). See also General Electric Co. v. Each time the Court has considered the doctrine, it hasWabash Appliance Corp., 304 U.S. 364, 369 (1938) acknowledged this uncertainty as the price of ensuring the(primary purpose of notice is “to guard against appropriate incentives for innovation, and it has affirmedunreasonable advantages to the patentee and disadvantages the doctrine over dissents that urged a more certain rule.”to others arising from uncertainty as to their [respective] Id. See also Festo v. Shoketsu Kinzoku Kogyo Kabushiki,rights.”). No. 95-1066, 2003 U.S. App. LEXIS 19867 (Fed. Cir. Sept. 26, 2003). 192 In Festo v. Shoketsu Kinzoku Kogyo 193Kabushiki, 535 U.S. 722 (2002), the Court reviewed a case Thomas 2/8 (Patent Session) at 42.involving the doctrine of equivalents, under which “[t]he 194scope of a patent is not limited to its literal terms but See, e.g., Arti K. Rai, Regulating Scientificinstead embraces all equivalents to the claims described.” Research: Intellectual Property Rights and the Norms ofFesto, 535 U.S. at 732. In that case, the Court conceded Science, 94 NW. U. L. REV. 77, 106-07, 138 (1999);that “the doctrine of equivalents renders the scope of Rebecca S. Eisenberg & Robert P. Merges, Opinion Letterpatents less certain [and that it] may be difficult to as to the Patentability of Certain Investions Associateddetermine what is, or is not, an equivalent to a particular with the Identification of Partial CDNA Sequences, 23element of an invention. If competitors cannot be certain AIPLA Q. J. 1, 4-20 (1995).about a patents extent, they may be deterred from engaging 195in legitimate manufactures outside its limits, or they may United States Patent and Trademark Office,invest by mistake in competing products that the patent Utility Examination Guidelines, 66 Fed. Reg. 1092 (2001),secures. In addition the uncertainty may lead to wasteful revising interim guidelines published at 64 Fed. Reg.litigation between competitors, suits that a rule of literalism 71440 (1999), which in tern superseded an earlier version,might avoid.” Id. The Court noted, however, that “[t]hese 60 Fed. Reg. 36263 (1995). 29
  • 59. received.196 Well-considered PTO one of the requirements of the patent statute.guidelines can prevent invalid patents thatcapture basic ideas and research and thus The PTO responded to concernsthwart competition in emerging fields.197 about the possible issuance of many questionable business method patents by(ii). Business Methods Patent Initiatives undertaking the “Business Method Initiative.”202 The primary goals of this In State Street Bank & Trust v. initiative were to identify sources of non-Signature Financial Group,198 the Federal patent prior art and to create mandatoryCircuit ruled that business methods can be fields of search for examiners.203 Inpatented. This decision has generated a fair addition, the PTO adopted another level ofamount of controversy, as has the PTO’s review for business method patents; thissubsequent issuance of hundreds of business level of review involves a “second pair ofmethod patents.199 From a competition eyes” – that is, a more senior examiner or anstandpoint, one could ask whether and, if so, examination panel takes a look at eachwhen a business method should be patented; business method patent application.204 Sincefor example, a patented business method the PTO introduced this program, themay stand in the way of Internet competition allowance rate for business method patentsin some circumstances.200 In addition, it is has decreased, and the PTO believes thatoften very difficult to locate and identify all this decreased allowance rate indicatesrelevant prior art with respect to a claimed improved PTO searches for prior art.205business method invention, because much of Such PTO action can prevent the issuance ofthe relevant prior art does not exist in the invalid patents that may contribute to marketpatent literature, the traditional source of power and restrain competition inrelevant prior art.201 Prior art is the primary unwarranted ways.way that patent examiners determinewhether a claimed invention is nonobvious, 196 See infra Ch. 4(II)(D)(1). 197 See generally Ch. 4(II)(D)(1) for furtherdiscussion of the PTO’s Utility Guidelines. 198 149 F.3d 1368 (Fed. Cir. 1998), cert. denied,525 U.S. 1093 (1999). 199 See infra Ch. 4(II)(E). 202 200 See, e.g., Musacchia 4/9 at 24-25; Young John Love 2/27 at 467-68.4/11 at 61, 63-64; Thomas 4/11 at 59-60; Richard C. Levin, 203Testimony (2/6/02) 2, at Id.http://www.ftc.gov/os/comments/intelpropertycomments/le 204vinrichardc.htm; Langenfeld 2/20 at 18; Kushan 4/11 at Id. at 470.114. 205 Id. at 470-71. See generally infra at Ch. 201 See Thomas 4/11 at 111; see infra Ch. 4(II)(E) for discussion of issues surrounding business4(II)(E)(2). method patents. 30
  • 60. B. The Growth of the technologies that were previously Knowledge-Based Economy unpatented or unpatentable (such as business methods); and increasingly complex Creates Ongoing Controversy technologies (such as biotechnology) must and Challenges Competition be evaluated. and Patent Policy to Continue Seeking a Better Balance Many panelists at the Hearings raised concerns that the patent system is not As discussed above, both antitrust keeping up with these challenges. Theyand the patent system have responded to the asserted that dubious patents, and costlychallenges posed by the knowledge-based patent procedures and litigation to determineeconomy and sought to improve the balance whether such patents are valid or infringed,between competition and patent policy. are stifling competition unnecessarily.Nonetheless, the Hearings revealed that Panelists observed that issues of patentmuch controversy remains about whether quality seem to arise more frequently than iscompetition and patent policy have yet desirable.207 In recent testimony before theresponded adequately to the knowledge- House Judiciary Committees Subcommitteebased economy or found a proper balance. on Courts, the Internet, and IntellectualThe joint FTC/DOJ report (forthcoming) Property, the AIPLA stated that "[l]arge andwill address the issues related to the balance small companies are increasingly beingbetween antitrust law and policy and patents subjected to litigation (or its threat) on thethat were raised at the Hearings. This report basis of questionable patents."208 As noteddiscusses issues related to the balance earlier, invalid patents that confer marketbetween patent law and policy and power unnecessarily thwart competition.competition that were raised at the Hearings. See also Box 1-6 (blocking patents). The growth of the knowledge-based Panelists pointed out that a numbereconomy presents several challenges to the of factors determine patent quality.patent system. One is the sheer number of Substantive standards of patent lawpatents sought and received. As Under determine whether the PTO and the courtsSecretary of Commerce for Intellectual evaluate the validity and scope of patentsProperty and Director of the USPTO James under proper standards. Procedures andRogan stated at the outset of these Hearings, presumptions at the PTO and in the courtsthere is an “unprecedented explosion ofpatent applications” today.206 Other aspectsof the knowledge-based economy also 207 See, e.g., infra Chs. 3(II), (IV) and Ch. 5(I).render the PTO’s mission more difficult.For example, pendency may assume 208 United States Patent and Trademark Officeparticular importance in fast-moving Fee Modernization Act of 2003: Hearing Before the Subcomm. on Courts, the Internet, and Intellectualtechnologies (such as software); prior art Property of the House Comm. on the Judiciary, 108th Cong.may be more difficult to locate for 2 (2003) (Statement of Michael K. Kirk, Executive Director, American Intellectual Property Law Association), available at http://www.aipla.org/html/Legislative/108/testimony/FeeLe 206 Rogan (stmt) 3. g.htm. 31
  • 61. Bo x 1-6 . Block ing Pa tents The patents of others can block a p atentee’s ability to exploit its own invention without a license to the o thers’ patents. Schechter and Thomas provide an example: “[S]upp ose that Admiral Motors obtains a patent on an internal combustion engine for use in automo biles. Later, Betty Beta purchases an automobile marketed by Admiral Mo tors that embodies the patented invention. Beta expe riments with her new car and develops a dramatically improved fuel injector useable only in the patented Admiral Motors engine. Even if Beta patents her improved fuel injector, she cannot practice that technolo gy without infringing Alpha’s basic patent. . . . Unless one of the parties licenses the other, Beta must wait until Admiral Mo tors’ patent expires before practicing her own patented improvement invention.” S CHECHTER & T HOMAS , T HE L AW OF C OPYRIGHTS , P ATENTS & TRADEMARKS § 20.1.1 at 462. If the alleged blocking patent is questionable, business costs, which ultimately may be passed on to consumers, can increase unjustifiably, with the owner of an improvement patent forced to choose between paying royalties on a questionab le patent or engaging in expensive patent litigation. See generally infra Ch. 5(III).further affect the ability to weed out of patent claims.unwarranted patents either before or afterthey are granted. Panelists raised several The real world adds otherissues concerning patent quality and how it complexities as well. In a simple economicaffects the proper balance between model of innovation and patents, eachcompetition and patent policy. We identify invention requires access to only one or aa few. few patents to commercialize the patented product. Certain industries, such as1. Follow-On Innovation, Product pharmaceuticals, have tended to follow this Commercialization, and Patent model.210 Some suggest, however, that more Proliferation and more industries are moving toward the model in which, for commercialization, a The simplest economic model of the product requires access to many patents –patent system assumes that innovation is a dozens, hundreds, or even thousands.211“one-time” event.209 Of course, in the real Reports indicate that this phenomenon canworld, innovation is an ongoing process, increase transactions costs substantially andwith one invention frequently providing a lead to additional problems such as royaltybuilding block for the next. The ongoingnature of innovation poses difficultquestions about how best to preserveadequate incentives for an initial innovatorand maintain adequate incentives forcompetition to become the next innovator. 210 See, e.g., Browder 3/19 at 174; Gregory J.These questions implicate substantive Glover, Competition in the Pharmaceutical Marketplacestandards for determining the proper breadth (3/19/02) 8, at http://www.ftc.gov/opp/intellect/020319gregoryjglover.pdf. 211 See, e.g., R. Levin 2/6 at 98-99; Cohen 2/20 209 See generally infra Ch. 2(I). at 29. 32
  • 62. stacking and patent thickets.212 These issues competition policy, are discussed in depth inhave implications for both patent and this report.215competition policy. See also Box 1-7(difficulties in claimdrafting and interpretation). Bo x 1-7 . Complexities Added by the Difficulty of Drafting and Interp reting Claim s2. Procedures that W hen a firm determines whether it needs access to one or more Third Parties Can patents held by others, it evaluates its planned business activities in relation Use to Challenge to the rights estab lished in others’ patents. Each patentee’s exclusive rights Questionable are based upon the invention, as recited in the claims of the patent. Each Patents claim consists of one sentence that verbally portrays a method, product or process; a patent may contain one or many claims. Sometimes, a patent may contain claims that overlap other claims in that patent, or that overlap Procedures that third claims in other patents.parties can use to challenge The inquiring firm reviews the claims se t forth in patents itquestionable patents may be believes it might infringe without a license. A firm’s activities may infringeinsufficient.213 Third parties only one, many, or all of the claims of the patent. In some cases, a review of the claims in others’ patents may yield uncertain answers. Althoughrarely use inter partes drafting claims sounds straightforward, experience has shown that it is oftenreexamination a very difficult task. As a corollary, issues can arise with some frequencyprocedures;214 moreover, regarding how claims sho uld be interpreted. See generally S CHECTER &participants in the patent T HOMAS , T HE L AW OF C OPYRIGHTS , P ATENTS & T RADEMARKS § § 18.2 and 20.2 at 404-20, 474 -75. Claim interpretation issues can add to thesystem generally view complexity that firms may confront in determining whether their plannedpatent litigation as too activities would infringe absent licenses to use others’ patents.costly and time consuming.Substantial concerns aboutpatent quality, however, have led to calls forimproving existing or developing new 3. Patent Prosecutions andprocedures through which third parties can Examinations within the PTOchallenge questionable patents. Theseissues, including their relationship to A variety of pressures that arise from the nature of recent technological change and innovation confront the PTO. 212 Royalty stacking describes the phenomenon Sometimes these pressures may conflict; forwhereby disparate owners of complementary technologiesdemand higher aggregate royalties than they would if they example, pressure to reduce the pendency ofacted as a group. See infra Ch. 2(III)(C)(3). A patent patent applications may conflict withthicket is a “dense web of overlapping intellectual property pressure to provide additional time forrights that a company must hack its way through in order toactually commercialize new technology.” Carl Shapiro, examinations of particularly complex patentNavigating the Patent Thicket: Cross Licenses, Patent applications. Indeed, patent applicants inPools, and Standard-Setting, in 1 INNOVATION POLICY AND different industries may take different viewsTHE ECONOMY 119, 120 (Adam Jaffe et al. eds., 2001). Seeinfra Ch. 2(III)(C) (describing economics of patentthickets); see generally Ch. 3(II)(D)(4) (describinginstances of royalty stacking and patent thickets). 213 See generally infra Ch. 5(I), (III). 214 215 See generally infra Ch. 5(III). See generally infra Ch. 5(I), (III). 33
  • 63. of which of these issues is most important.216 Opportunities to Broaden Claims. Increasing Complexity and Limited Some believe that an opaque process forTime for Patent Examiners. Throughout the patent prosecution at the PTO can allowHearings, panelists lamented the PTO’s firms unfairly to disadvantage theirinability to provide examiners with competitors. For instance, some assert thatsufficient time to undertake their review.217 applicants can anticompetitively game patentThe increasing complexity of patents continuations to capture subject mattercompounds this challenge. One panelist already developed by a competitor.221 Thisnoted, for example, that typically new raises significant issues for both patent andexaminers have 25 hours, and more competition policy.experienced examiners have 20 hours, toexamine a biotechnology patent. He felt Patent Pendency. Faster technologythese time constraints were “clearly evolution and shorter product life cyclesinadequate given the complexity and have increased the pressure on the PTO todifficulty of biotechnology patents. . . .”218 reduce pendency times.222 As the U.S.This panelist recommended not only that the House of Representatives Committee onPTO double the time allocated for such Science/Subcommittee on Technologypatent examinations, but also that the PTO recognized: “In a growing number ofprovide examiners with more training.219 industries - such as computer hardware andExpertise comes not only from education but software . . . - the pace of advancement hasalso from experience.220 begun to challenge the ability of the patent office to process applications in a time frame that is functionally useful to the inventor.”223 216 See generally infra Ch. 3(III)(D)(2), (V)(B) In fast moving fields, such as electronics,(compare biotech representatives expressing views thatmore thorough examinations are more important thanreducing pendency times with software representativesexpressing concern that patents emerge only after they no Impossible Patents Before Breakfast: Property Rights forlonger have any commercial value). Business Concepts and Patent System Reform, 14 BERKELEY TECH. L.J. 577, 607-8 (1999). See also Gable 217 See, e.g., Dickinson 2/6 at 64; Gable 3/20 at 3/20 at 121 (“There is a very significantly high turnover in121. the examiners particularly . . . in the biotech area as well as the software, method of doing business area.”). 218 Kirschner 2/26 at 243. 221 See generally infra Ch. 4(II)(C)(1). 219 Kirschner 2/26 at 244. 222 Cf. FTC Staff Report, Anticipating the 21st 220 The PTO has sometimes suffered from a Century: Competition Policy in the New High-Tech,“crippling attrition rate,” due to more experienced Global Marketplace, Ch. 6 at 15 (May 1996)examiners going to higher paying private sector jobs; more (“Competition to be first on the market has resulted inrecently, the attrition rate at the PTO has been falling. See shortening product life cycles, at least in high-techHearing Before the Subcomm. on Courts, the Internet and industries.”), available atIntellectual Property of the House Comm. on the Judiciary, http://www.ftc.gov/opp/global/report/gc_v1.pdf.107th Cong. 2 (2002) (Statement of James E. Rogan, Under 223Secretary of Commerce for Intellectual Property and The Patent System and Modern TechnologyDirector, United States Patent and Trademark Office), Needs: Meeting the Challenges of the 21st Century,available at Hearing Charter Before the Subcomm. on Technology ofhttp://www.uspto.gov/web/offices/com/speeches/househrg2 the House Comm. on Science, 104th Cong. (1996),002.htm. See also Robert P. Merges, As Many as Six available at http://www.house.gov/science/patchrt.htm. 34
  • 64. semiconductor, and telecommunications, IV. THE HEARINGSpatents granted years after filing may be of EXAMINED THE“little value.”224 CURRENT BALANCE OF4. Patent Quality and Patentable COMPETITION AND Subject Matter PATENT LAW AND POLICY IN FOSTERING Many at the Hearings noted the INNOVATIONcontinuing expansion of what can constitutepatentable subject matter.225 The transition As noted earlier, the growth of theof subject matter from a status of “generally knowledge-based economy means thatopen to free competition” to a status in increasingly complex questions confrontwhich an inventor may obtain a patent on it antitrust enforcers, and increasinglycan raise questions for competition policy.226 numerous and challenging patentIn addition, panelists explained that the applications and patent issues confront theexpansion of patentable subject matter can patent system. Some claim that thesecause difficult transition periods for patent challenges have led to problems in the patentpolicy. The courts and the PTO must system that cause unnecessary harm todetermine how best to apply existing patent competition and may even require antitrustdoctrines to the newly patentable subject solutions. Others assert that thesematter. challenges have confounded antitrust and require even greater deference to patents. The FTC and the Antitrust Division of the Department of Justice convened these Hearings to learn more about these and other questions. A. The Hearings Did Not Address Certain Fundamental Questions or 224 Michael Kirk, AIPLA/FICPI Colloquium on Issues with InternationalPendency Reduction 9-10 (2001), athttp://www.aipla.org/html/ficpi/2001/ficpi1119.pdf. RamificationsThough not addressing the pendency issue explicitly, onepanelist discussed the consequences of increasing The Hearings did not address certaintechnological change and the value of intellectual propertyprotection as a practical matter. Burk 3/20 at 141 (If a fundamental questions. For example, theproduct has a “very, very short life,” then “some Hearings did not ask whether there shouldintellectual property protections, as they now exist, just are be a patent system. Some panelists noted anot terribly helpful in your business plan.” Instead, suchcompanies sell the product “for six months until our correlation between a strengthened patentcompetitors copy it” and then sell something else.). system during the 1980s and subsequent 225 robust performance of the U.S. economy; See generally infra Ch. 4(II)(E)(3). they suggested a causal link between those 226 See generally infra Ch. 4(II)(E) (discussion ofbusiness method patents). 35
  • 65. events.227 Regardless of whether and to policy. The joint FTC/DOJ report willwhat extent such a link exists, there is no address the appropriate balance of antitrustgainsaying the innovation that businesses law and policy with patents. This reportreport that the patent system has spurred.228 applies a competition and economic perspective to identify the following policy The Hearings also did not ask goals for a proper balancing of patent lawwhether the duration of a patent is optimal; and policy with competition concerns.Congress and international organizationshave recently spoken on the legal length of 1. The Legal System Should Providepatents.229 Similarly, the Hearings did not Efficient Incentives for All Typesaddress various questions – such as whether of Innovation, Including Bothto use a first-to-file or first-to-invent Single-Stage and Follow-Onstandard – that are in discussion among the InnovationUnited States and other countries ininternational fora.230 Single-stage Innovation. Efficient incentives for innovation begin withB. The Hearings Examined the assuring adequate appropriability for single- Appropriate Balance of stage innovation. By conferring a right to exclude, the patent system can enhance Competition and Patent Law appropriability and increase incentives to and Policy from a innovate.231 Patents also may be important Competition and Economic bases for attracting financial support, Perspective particularly for small, new firms without tangible assets and reliable cash flow.232 The Hearings addressed questions Patents can thereby facilitate entry andabout the appropriate balance of innovation.233 The relative importance ofcompetition, antitrust, and patent law and patents for appropriability, however, varies 227 See, e.g., Newman 2/6 at 40-41, 49; see alsoPate, Antitrust and Intellectual Property at 16. 231 See, e.g., Thomas 2/8 (Patent Session) at 13- 228 See generally infra Ch. 3. 15; Langenfeld 2/20 at 7-8; Stoner 2/26 at 108; Taylor 2/27 at 489-90; Duffy 7/10 at 107; Chambers 10/25 at 30; ABA 229 To comply with the Agreement on Trade- (Economics stmt) 17-18; Intellectual Property OwnersRelated Aspects of Intellectual Property Rights (the TRIPS Association, Comments on the Joint Hearings of theAgreement), the United States in 1995 enacted the Uruguay Federal Trade Commission and the Department of JusticeRound Agreement Act, providing, among other things, a Regarding Competition and Intellectual Property Law andpatent term of twenty years from the patent application’s Policy in the Knowledge-Based Economy (Publicfiling date. See 35 U.S.C. § 154(a)(2); see also Comment) 4, atSCHECHTER AND THOMAS, INTELLECTUAL PROPERTY: THE http://www.ftc.gov/os/comments/intelpropertycomments/ipLAW OF COPYRIGHTS, PATENTS AND TRADEMARKS § 13.3.3 o.pdf (hereinafter IPO (stmt)).at 288. 232 See, e.g., Merges 2/28 at 577-78; Scherer 230 See, e.g., Adam I. Hasson, Domestic 7/10 at 53; Hoerner 7/11 at 54; Barton 2/26 at 212.Implementation of International Obligations: The Quest 233for World Patent Law Harmonization, 25 B.C. INTL & See, e.g., Lerner 2/20 at 186; Hall 2/26 atCOMP. L. REV. 373 (2002). 179, 183, 191; Ziedonis 3/20 at 17-18, 87-88. 36
  • 66. from industry to industry.234 infringing uses, and, after the patent expires, the invention becomes part of the public Follow-on Innovation. Innovation domain.often is a cumulative process, with eachstage building on its predecessors. To the 3. The Patent System Should Avoidextent that follow-on innovation flows from Creating or Upholdingsources independent of the initial innovator, Unwarranted Patents that Conferit is vital that efficient incentives to innovate Market Powerexist for the original and for follow-oninnovators.235 “We should be wary of creating unwarranted market power by granting2. Safeguard the Patent System’s unwarranted patents.”238 Unwarranted Disclosure Function market power can produce supracompetitive pricing, deter competition to spur In exchange for receiving a patent, a innovation, and cause other harms topatentee must disclose the nature of the consumers.239 From a patent perspective,invention; disclosure is the basic quid pro an unwarranted patent is one that does notquo of the system.236 Disclosure can provide meet the statutory standards forthe public with knowledge that otherwise patentability. From an economicmight have been kept a trade secret.237 The perspective, however, unwarranted marketpublic may apply that knowledge in non- power can arise from unwarranted patents – that is, patents for inventions that would have emerged in roughly the same time 234 See infra Ch. 2(II)(A)(2). Testimony frame, and for which disclosure andindicated that patents are likely to have greatest commercial development would havesignificance as appropriability mechanisms when R&D occurred, even without the prospect of acosts are high relative to the size of the market, andimitation is quick and easy. See id. patent.240 235 See Scotchmer 2/26 at 128-29. See generallyinfra Ch. 2(III) for a discussion of different theories abouthow best to address this issue. Design-around innovation.Some stress that the patent system directs R&D away from 238imitative and toward innovative efforts by forcing R. Levin 2/6 at 102. Recognition of potentialcompetitors to design around patents. Others respond that market power effects was a theme echoed by many otherdesign-around may be technically impossible or participants. See, e.g., ABA (Economics stmt) 11economically impractical and may entail costly efforts (describing the exercise of market power as a possible costessentially to duplicate the patentee’s invention. See infra of patent protection); Langenfeld 2/20 at 10-13; StonerCh. 2(III)(B)(1) for a discussion of design-around 2/26 at 108-09; Hall 2/26 at 181, 184; Farrell 2/28 at 596;innovation. Katsh 4/10 at 25-26; Gambrell 10/25 at 38-39. 236 239 See, e.g., Rogan 2/6 at 21; Cohen 2/20 at 35; See, e.g., infra Ch. 2(I)(B).Myrick 3/19 at 18. See generally Stoner 2/26 at 109-10.Indeed, some viewed disclosure as the system’s central 240 As noted earlier, many view this perspective –feature. See Myrick 10/30 at 25 (describing focus on that patents should be granted only if the invention woulddisclosure as “really what the patent system is all about”). not have emerged “but for” the patent system – as the “defining proposition” for standards of patentability. See 237 But see infra Ch. 3 (Hearings record mixed on Merges 2/28 at 579. Most concede, however, that the “butwhether businesses use patents when they can keep for” standard cannot practically be applied in individualinventions as trade secrets instead). cases. See generally infra Ch. 4(II)(A). 37
  • 67. 4. The Patent System Should Rely on business investments.245 Trade-offs may be Substantive Standards and necessary among the accuracy, transparency, Procedures that Minimize the Sum and manageability of substantive of Error and Process Costs and the standards246 and the error rates and process Detrimental Effects of Uncertainty and uncertainty costs of different approaches toward quality control.247 The goal is to All legal regimes should consider the minimize the sum of error and process costsextent to which they are subject to error – and the detrimental effects of uncertainty.that is, false negatives and false positives.241In the antitrust context, this translates into C. Organization of the Reportunder-enforcement (failing to challengeanticompetitive conduct) versus over- We begin with what economics canenforcement (erroneously condemning teach us about the relationship ofefficient, welfare-enhancing conduct).242 In competition and patent policy to innovationthe patent context, this translates into and then review business testimony aboutdenying a patent that should have been specific industries. We next examine patentgranted versus granting an unwarranted approaches that may ameliorate perceivedpatent.243 Legal systems also shouldconsider the extent to which they create orminimize costs or business uncertainty 245 See, e.g., Teece 2/26 at 202-04 (“the greaterthrough the use of specific procedures and the ambiguity around intellectual property rights the lesspresumptions.244 Among other problems, likely that the market will be able to work”); Friedman 2/27uncertainty can thwart effective business at 411-12 (patent uncertainties undermine R&D planning, add to risks, and frustrate innovation incentives); Quillenplanning and increase costs of capital for 3/19 at 29 (patent uncertainty raises innovation capital costs); IPO (stmt) 3 (uncertainty adds costs and impairs business planning). 246 See, e.g., Dreyfuss 7/10 at 142-43; Pooley 10/30 at 55-57; see generally Louis Kaplow, Rules Versus 241 Standards: An Economic Analysis, 42 DUKE L. J. 557 See generally Isaac Ehrlich & Richard A. (1992).Posner, An Economic Analysis of Legal Rulemaking, 3 J.LEGAL STUD. 257 (1974). 247 See, e.g., Taylor 10/25 at 51-52; F. Scott 242 Kieff, Summary of Proposed Testimony (Public Comment), See, e.g., CHARLES J. GOETZ & FRED S. atMCCHESNEY, ANTITRUST LAW: INTERPRETATION AND http://www.ftc.gov/os/comments/intelpropertycomments/haIMPLEMENTATION 67-69 (2nd ed. 2002) (discussing rvardlaw.pdf; see generally Richard A. Posner, Anapproach of antitrust law to Type I (false positive) and Economic Approach to Legal Procedure and JudicialType II (false negative) error); Frank H. Easterbrook, The Administration, 2 J. LEG. STUDIES 399 (1973). IssuesLimits of Antitrust, 63 TEX. L. REV. 1, 15-16 (1984). could include whether a patent challenger should need “clear and convincing evidence” to rebut a patent’s 243 See Erik S. Maurer, An Economic presumption of validity (the presumption of validity canJustification for a Broad Interpretation of Patentable save process costs, but may erroneously protect invalidSubject Matter, 95 NW. U. L. REV. 1057, 1094-96 (2001) patents) or whether cases should be decided by only one(arguing that analysis of Type I and Type II errors supports specialized court or by numerous regional courts of appealbroader scope for patentable subject matter). (use of one specialized court may save process costs and may contribute to stability in the law, but may lead to more 244 See generally Isaac Ehrlich & Richard A. errors in the development of the case law, which has notPosner, An Economic Analysis of Legal Rulemaking, 3 J. had the benefit of as many different perspectives on andLEGAL STUD. 257 (1974). insights into the issues at hand). 38
  • 68. problems. We conclude with a discussion of patent terms (App. C), and a list of selectedrecommendations for antitrust and patent federal statutes (App. D).institutions. The following chapters discuss theseissues:Chapter 2: What can we learn from theoretical and empirical economics about the general relationship between competition policy, patents, and innovation?Chapter 3: What can we learn from the examination of individual industries about areas in which the balance between competition and patents seems to be working well or, conversely, might be off- kilter?Chapter 4: What suggestions for substantive patent law reform might address problematic issues raised at the Hearings?Chapter 5: What suggestions for procedural patent law reform might address problematic issues raised at the Hearings?Chapter 6: What suggestions might facilitate greater interaction between antitrust and patent institutions about the issues discussed in this report?In four appendices, we also provide a list ofcontributors to the Hearings (App. A), a listof public comments (App. B), a glossary of 39
  • 69. CHAPTER 2 THE ROLE OF COMPETITION AND THE PATENT SYSTEM IN SPURRING INNOVATIONIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1I. PATENTS’ EFFECTS ON STAND-ALONE INNOVATION . . . . . . . . . . . . . . . . . . . . . 3 A. Patents Can Spur Stand-Alone Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1. Internalize Externalities and Protect Against Free Riding . . . . . . . . . . . . . 4 2. Facilitate Commercialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3. Encourage Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 B. Costs Of, and Limits To, Patents’ Power to Spur Stand-Alone Innovation . . . . . 7II. COMPETITION’S EFFECTS ON INITIAL AND FOLLOW-ON INNOVATION . . . . . 8 A. Competition Can Spur Innovation, Whether Initial or Follow-On . . . . . . . . . . . . 9 1. Cannibalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2. Races to Innovate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3. Schumpeterian Hypothesis and its Critics . . . . . . . . . . . . . . . . . . . . . . . . 12 4. Diversity of R&D Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 B. Costs Of, and Limits To, Competition’s Power to Spur Innovation . . . . . . . . . . 16 1. Appropriability Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2. Duplication of Effort. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16III. PATENTS’ EFFECTS ON FOLLOW-ON INNOVATION . . . . . . . . . . . . . . . . . . . . . . 17 A. The Roles of Managed and Independent Follow-On Innovation . . . . . . . . . . . . . 18 1. Follow-On Innovation Organized by the Initial Innovator . . . . . . . . . . . 18 2. Follow-On Activities by Independent Follow-On Innovators . . . . . . . . . 20 B. Follow-On Innovation in the Face of a Single Blocking, Initial Patent . . . . . . . . 21 1. Design-Around Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2. Division of Rewards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3. Licensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 C. Follow-On Innovation in the Face of Multiple Existing Patents . . . . . . . . . . . . . 25 1. High Transaction Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 a. Stemming from Number of Patents . . . . . . . . . . . . . . . . . . . . . . . 27 b. Stemming from Lack of Benchmarks. . . . . . . . . . . . . . . . . . . . . . 27 2. Hold Up in the Patent Thicket . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 a. Hold Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 b. Strategies to Mitigate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
  • 70. 3. Royalty Stacking and Cournot’s Complements Problem . . . . . . . . . . . . . 32 4. Oligopoly/Group Boycott . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 5. Patent Fences and Patent Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 6. Patent Flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
  • 71. CHAPTER 2 THE ROLE OF COMPETITION AND THE PATENT SYSTEM IN SPURRING INNOVATION Introduction. Competition and award of patent rights can spur stand-alonepatent policy play complementary roles in innovations by limiting free riding,enhancing economic welfare over time.1 facilitating commercialization ofThis chapter explores the economic learning innovations, and encouraging disclosure of– based on economic theory and empirical new ideas.4 Pharmaceutical companies, foreconomic evidence – about the effects that example, rely on patents to prevent freepatent policy and competition can have on riding, recoup their R&D investments, andinnovation and economic welfare.2 learn about new technological breakthroughs, according to many panelists.5 Patents and Stand-Alone Innovation. Biotechnology start-ups rely on their abilityIt is easy to see how patent awards affect to patent their innovations to attractstand-alone innovations, and the discussion investment and continue innovating, somebelow begins with that clear case.3 The panelists stated.6 Awarding patent rights, however, is 1 See supra Ch. 1(I)(B). See also JEAN TIROLE, not costless. An innovator whose patentTHE THEORY OF INDUSTRIAL ORGANIZATION 390-92, 400(1988) (discussing different incentives for innovation). confers market power can raise prices or depress output7 (and, as developed below, 2 The scope of the inquiry here is limited and broad initial patent rights can sometimesomits some of the complexities of different types ofinnovation and regulation. For example, the discussion interfere with follow-on innovation8). Thesedoes not distinguish between “process” and “product” effects may be the price of progress, if theinnovation. (The former term refers to changing the promise of a patent grant is necessary toproduction process to reduce the costs of making a product,and the latter involves improving the quality of the product elicit an invention, its disclosure, oritself. See, e.g., American Bar Association Section of investment in it. If invention, disclosure, orAntitrust Law, The Economics of Innovation: A Survey investment would have occurred even(Public Comment) 4 (reporting that more than three-fourthsof R&D expenditure in the United States is on product without the promise of a patent award,innovation), at however, these costs hurt consumershttp://www.ftc.gov/opp/intellect/0207salabasrvy.pdf(hereinafter ABA (Economics stmt)). Similarly, thisdiscussion does not elaborate upon the important point that infra Ch. 2(III).– in addition to granting intellectual property rights – theU.S. government takes other steps to increase innovation. 4See supra Ch. 1. Nor does this chapter discuss the optimal See infra Ch. 2(I)(A) (discussing how patentslength of patent protection. See, e.g., ABA (Economics can spur stand-alone innovation).stmt) 14-15 (summarizing literature on the optimal patent 5length). (The current patent term of twenty years from the See infra Ch. 3(II)(C). More information aboutfiling of the patent application, see supra Ch. 1(I)(A)(2), this and other real-world illustrations of the economicderives from statutorily implemented international phenomena described in this chapter follow, in Chapter 3.obligations.) 6 See infra Ch. 3(III)(D). 3 See infra Ch. 2(I) (discussing patents’ effects 7on stand-alone innovation). Patent policies can also affect See infra Ch. 2(I)(B) (discussing costs offollow-on innovation, to be sure. For ease of exposition, patents).however, this chapter focuses first on the simpler case of 8stand-alone innovation. For a discussion of other effects See infra Ch. 2(III) (discussing patents’ effectsthat patent policies have on follow-on innovations, see on follow-on innovation).
  • 72. unjustifiedly.9 innovation races can generate duplicative research, which some deem wasteful.13 Competition and Initial and Follow-On Innovation. Like patent policy, Patents and Follow-On Innovation.competition also affects innovation. On the The analysis concludes with a discussion ofone hand, competition can spur innovation the effects of patent grants on follow-onin a wide variety of ways. As an initial innovation.14 Admittedly, the categories ofmatter, competition to win a patent right initial and follow-on innovation are hardlymay drive a race to innovate. Indeed, firms hermetically sealed. The progression ofcompeting to innovate may approach innovation is often continuous. Today’sresearch problems differently, increasing the follow-on innovation often becomes thechances of successful innovation. foundation for a future advance.15 In keepingMoreover, in some circumstances, an with much of the scholarly analysis and forinnovator may reap the benefits of its work ease of exposition, however, this chaptersimply by exploiting its head start on its analyzes initial and follow-on innovationcompetitors. For example, empirical studies separately and discusses the various issues inhave demonstrated that in the semiconductor the context in which they have the greatestand communications equipment industries, significance.patents are less important than other meansof exploiting innovation, means such as Some at the Hearings argued thatmaintaining secrecy, taking advantage of broad initial patent grants facilitate follow-lead time, investing in complementary on innovation by allowing the patentee tomanufacturing processes, and offering organize research flowing from itscomplementary sales and services.10 This innovation.16 By contrast, others contendedchapter explores these and other ways in that broad initial patent rights can sometimeswhich competition can drive innovation.11 impede follow-on innovation that would otherwise emerge from entities independent On the other hand, competition alone of the patentee. A patentee’s refusal tois not a perfect engine of innovation. As license an initial patent on technologynoted above, competition, standing alone, needed for follow-on research can hinderdoes little to limit free riding on others’innovations,12 and competition-driven 13 See infra Ch. 2(II)(B)(2) (discussing costs of duplication of efforts). 9 See infra Ch. 2(I)(B) (discussing costs andlimits of patents’ power to spur stand-alone innovation); 14 See infra Ch. 2(III) (analyzing patents’ effectsCh. 2(III) (discussing patents’ effects on follow-on on follow-on innovation).innovation). 15 See, e.g., FTC/DOJ Hearings on Competition 10 See infra Ch. 2(II)(A)(2) (discussing these and Intellectual Property Law and Policy in thestudies). Knowledge-Based Economy, Suzanne Andersen Scotchmer Testimony Feb. 26, 2002, at page 170 (hereinafter, citations 11 See infra Ch. 2(II)(A) (discussing to transcripts of these Hearings state the speaker’s lastcompetition’s power to spur innovation). name, the date of testimony, and relevant page(s)). 12 16 See infra Ch. 2(II)(B)(1) (discussing See infra Ch. 2(III)(A)(1) (discussing follow-appropriability problems). on innovation organized by the initial innovator). 2
  • 73. follow-on innovation, according to some. hold up, noting that the owner of any one ofOthers, however, stressed the potential the multitude of patented technologiesbenefits of design-around activities and the constituting a software program can hold upavailability of licenses.17 For example, the production of innovative new software.24fact that the ulcer-treating drug Tagamet waspatented forced others to design around it, In short, panelists noted that bothleading to the development of another competition and patent grants can spursuccessful product, Zantac, according to innovation, but both can have adversesome Hearing testimony.18 effects on innovation as well. This chapter aims to outline the costs and benefits of each Some panelists expressed concern approach to enhancing economic welfare.that researchers who require access not justto a single patent but to multiple patents may I. PATENTS’ EFFECTS ONfind their work impeded by high transaction STAND-ALONEcosts,19 royalty stacking, 20 hold up in patent INNOVATIONthickets,21 and oligopolists seeking to barnew entry.22 Panelists made clear that theseare not merely hypothetical concerns. For A. Patents Can Spur Stand-example, some panelists noted that the Alone Innovationplethora of patents in the computer hardwareindustry makes it “virtually impossible to As noted in Chapter 1, intellectualsearch all potentially relevant patents, property is particularly susceptible toreview the claims,” and evaluate the misappropriation, also known as “freeinfringement risk;23 and panelists from the riding.” Patents can limit free riding andsoftware industry complained of the risk of also facilitate commercialization of the intellectual property the patent protects. This chapter addresses each of these 17 scenarios below. It also explores how patent See infra Ch. 2(III)(B)(1) (discussing design-around innovation); Ch. 2(III)(B)(3) (discussing licenses). policy encourages disclosure, and how that disclosure can stimulate further innovation.25 18 See infra id. (discussing examples of design-around innovation). 24 19 See infra Ch. 3(V)(E). See infra Ch. 2(III)(C)(1) (discussingtransaction costs). 25 Patents, like other property rights, can also 20 serve as an underpinning of competition and thereby spur See infra Ch. 2(III)(C)(3) (discussing royalty innovation. For example, firms may use patent rights, likestacking and the Cournot complements problem). other property rights, to compete with each other on 21 innovation. Cf. U.S. Department of Justice and Federal See infra Ch. 2(III)(C)(2) (discussing hold up Trade Commission, Antitrust Guidelines for the Licensingin the patent thicket). of Intellectual Property § 1.0 (Apr. 6, 1995), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,132 (“The intellectual 22 See infra Ch. 2(III)(C)(4) (discussing property laws provide incentives for innovation and itsoligopoly and group boycotts). dissemination and commercialization by establishing enforceable property rights for the creators of new and 23 Robert Barr, Statement (2/28/02) 1, at useful products [and of] more efficient processes.”),http://www.ftc.gov/opp/intellect/barrrobert.doc (hereinafter available atBarr (stmt)). http://www.usdoj.gov/atr/public/guidelines/ipguide.htm. 3
  • 74. 1. Internalize Externalities and innovations, enhancing appropriability. 29 Protect Against Free Riding Economic theory suggests that by conferring such rights to exclude, the patent system Economists recognize that without increases incentives to innovate.30patent protection, “innovators [that produceintellectual property] cannot appropriate the 29 ABA (Economics stmt) 10-12 (discussingfull benefits of their innovation; some of the “invention motivation” rationale for patent protection); seebenefits go to ‘free riders’ without also Stoner 2/26 at 108-09; Thomas 2/8 (Patent Session) atpayment.”26 If innovators know that they 15, Intellectual Property Owners Association, Comments on the Joint Hearings of the Federal Trade Commissioncannot exclude imitators and appropriate the and the Department of Justice Regarding Competition andfruits of their R&D efforts, then they may Intellectual Property Law and Policy in the Knowledge-lack sufficient incentives to undertake the Based Economy (Public Comment) 4, at http://www.ftc.gov/os/comments/intelpropertycomments/ipinnovation in the first place.27 The problem o.pdf.is especially acute when the original Panelists discussed the degree to which such protectioninnovator’s efforts entail substantial fixed from free riding helps entrants. Compare Greenstein 2/20 at 143-47 (discussing entrant’s ability to use patent tocosts, and the imitators can copy the prevent imitation by incumbent); Hall 2/26 at 179, 183,innovation cheaply.28 Patent rights mitigate 191 (patents may facilitate entry by helping with securingthis problem by granting exclusive rights in financing and by allowing firm to exploit its innovation), Hall 2/26 at 190-91 (patents facilitate vertical disintegration and entry by firms with only intangible assets); Arora 2/25 at 72 (patents permit small firms to compete in R&D without having extensive downstream 26 ABA (Economics stmt) 10-12 (discussing assets); Merges 2/28 at 578 (in the raising of capital, the“invention motivation” rationale for patent protection); see marginal importance of patent grows as size of businessalso Stoner 2/26 at 108. Even with a patent, patent holders declines); Nydegger 2/27 at 525-26 (smaller firms acquiremay be unable to appropriate the full benefits of their patents to protect innovative technologies and "hopefullyinnovation because patent protection is limited. For put them on a somewhat level playing field with largerinstance, others can learn of the invention and make use of competitors"); Scherer 7/10 at 53 (patents important tothe knowledge as long as they do not infringe the patent small, new firms without reliable internal cash flow);claims. Taylor 2/27 at 490 (reward essential to attract capital); Hoerner 7/11 at 54 (patents particularly important for start- 27 See, e.g., Alstadt 3/19 at 39 (noting that his ups needing financing) with Cohen 10/30 at 78 (withclient will not pursue concept for new alloy unless patent imperfect capital markets for investment in legal resources,protection is available). Langenfeld 2/20 at 8 (“[i]f you small firms and entrants may have less ability to enforcehave an idea and you can’t protect it adequately, other their patents); Barton 2/26 at 213 (small firms often cannotpeople will steal it and use it and that, obviously, deters afford to litigate). Cf. Liebowitz 2/20 at 233-34your incentive to develop those ideas yourself.”); Duffy (contrasting this traditional goal of patent ownership with7/10 at 107 (discussing inventors’ disincentives to innovate other goals).absent assurances that they can recover R&D costs);Chambers 10/25 at 30 (noting that his clients have foregone 30 It is unlikely that there is too much innovationpursuing “areas or . . . products” because of lack of from the viewpoint of economic welfare. Innovation oftenassurance that “they were going to have a clear ownership generates “large positive spillovers” that the inventorright”). cannot appropriate; as a result, there is a general “underinvestment in innovative activities.” Thomas M. 28 See, e.g., Scherer 7/10 at 52 (stating that Jorde & David J. Teece, Rule of Reason Analysis ofpatents are most likely to be important when R&D costs are Horizontal Arrangements: Agreements Designed to“high relative to the size of the potential market but Advance Innovation and Commercialize Technology, 61imitation can be quick and easy, that is, with imitator R&D ANTITRUST L.J. 579, 584 (1993); see also id. at 583-88costs much lower than those incurred by the innovator”); (summarizing empirical evidence showing that “the socialTaylor 2/27 at 489-90 (patent system is “absolutely returns to innovation are markedly greater than the privateessential” for industries in which firms must expend “high returns”); Dennis Carlton, Antitrust Policy Towardfront-end costs” and in which “their products are easily Mergers When Firms Innovate: Should Antitrustcopied and attract[] free riders”). Recognize the Doctrine of Innovation Markets?, Testimony 4
  • 75. This view of the role of patents appropriability in many industries.34 In otherassumes that invention is “a one-time industries, however, these alternatives maystationary phenomenon, not a cumulative be less readily available.process whereby inventions build on eachother.”31 When innovation is not 2. Facilitate Commercializationcumulative, enhancing appropriability raisesfew concerns about any “offsetting Some inventions lack commercialretardation of innovation that could come capability at first. Only substantialfrom the increased risk of infringement by development can turn them intofollowers in the cumulative chain.”32 When commercially viable products. Economicinnovation is cumulative, however, allowing theory posits that patent rights make it easierthe initial innovator to appropriate more of for inventors to develop relationships withthe rewards from its invention may hinder others to invest in that development.35independent follow-on innovation. Patents can make information a tradeableIndependent firms seeking to build on the commodity by reducing transaction costsinitial innovation would have to bear the risk and enabling licensing negotiations.36of infringement or the cost of negotiating Without patent rights, inventors might haveand paying for licenses. Thus, the granting to rely on secrecy to prevent free-riding onof strong patent rights may carry costs.33 their innovation; by shielding inventors from such free-riding, patents allow them to Appropriability mechanisms other discuss their work with other firms that canthan patents – such as trade secrecy, first- help commercialize the invention.37 If firmsmover advantages, and learning-curve had to rely on trade secrets to protect theiradvantages – may also protect the innovator inventions, it would be “very difficult to . . .from free riding. Indeed, a number of studies efficiently transfer information from thehave shown that such measures typically aremore important than patents for protecting 34 See infra Ch. 2(II)(A)(2) (discussing races to innovate); see also Cohen 2/20 at 25-26; Scherer 7/10 at 51-52. In particular, Prof. Teece has noted that problems in the patent system are sometimes the reason that firms use non-patent means of appropriating value from theirBefore the FTC Hearings on Global and Innovation-Based innovations. See Teece 2/26 at 206.Competition 6 (Oct. 25, 1995) (noting that the “social rateof return [on innovation] exceeds the private one, 35 See generally ABA (Economics stmt) 12.suggesting that more R&D would be desirable”), athttp://www.ftc.gov/opp/global/carlton.htm; Richard J. 36 See, e.g., Bronwyn H. Hall, Patents andGilbert & Steven C. Sunshine, Incorporating Dynamic Innovation (2/26/02) (slides) at 8 (patents allow trade inEfficiency Concerns in Merger Analysis: The Use of knowledge), atInnovation Markets, 63 ANTITRUST L.J. 569, 593-94 n. 60 http://www.ftc.gov/opp/intellect/020226bronwynhhall.pdf.(1995) (noting evidence that “the private return to R&D is Other kinds of intellectual property, such as trade secrets,much less than the social return”). can likewise facilitate trade in information. See, e.g., ROGER M. MILGRIM, 2 MILGRIM ON TRADE SECRETS 31 Stoner 2/26 at 108-109. § 9.01[4], 9-13–9-24 (2003) (noting that trade secrets may be licensed). 32 Id. 37 See, e.g., Kitch 2/20 at 84 (patents enable 33 See infra Ch. 2(III) (discussing patents’ effects contracting to transfer information); Arora 2/25 at 72on follow-on innovation). (patents “enhance the efficiency of knowledge transfer”). 5
  • 76. inventor or even the investor to . . . the entity which they receive patents.42 Thisthat [is] best able to exploit and develop disclosure obligation is a quid pro quo forit.”38 As one panelist put it, without patent obtaining the right to exclude others fromrights, making, using, offering for sale, or selling an invention.43 The purpose of the disclosure [y]ou can imagine the basic problem. obligation is to foster further innovation by An independent inventor goes to a enabling a person skilled in the particular art large firm [and says,] ‘Hey, Ive got a to learn from another’s invention.44 Thus, an great invention.’ And the large firm issued patent “communicates a considerable says, ‘Well, what is it?’ Well, amount of information that can help other without a property right the would-be inventors, including rival firms.”45 conversation might stop.39 Although some questioned whether the Rendering innovation a tradeablecommodity also helps foster specialization. 42 See infra Ch. 4(II)(B) (describing statutoryA small firm that has invented something requirements). See also Rogan 2/6 at 21 (the quid pro quoneed not do alone all the things necessary – for receiving patent rights is disclosure); Myrick 3/19 at 18-19 (stating that “[p]atenting . . . serves the publicfrom the advertising and warranties to sales interest by encouraging still more innovation, which in turnand service – to bring the invention to must be publicly disclosed to be entitled to patentmarket.40 Instead, it can license or sell its protection”). Since 1999, patent law has also required the publication of certain patent applications 18 months afterinvention to another firm, which can then do they are filed, see infra Ch. 4(II)(C)(1); however, throughwhatever tasks are needed to develop and the use of continuations, a patent may issue that containsmarket the invention.41 In these ways, the broader claims than publication initially revealed, see id.patent system facilitates the 43 See Rogan 2/6 at 21; J.E.M. Ag Supply, Inc. v.commercialization of inventions. Pioneer Hi-Bred Int’l, Inc., 534 U.S.124, 142 (2001) (“The disclosure required by the Patent Act is the ‘quid pro quo of the right to exclude’” (internal citations omitted));3. Encourage Disclosure Merges & Duffy, Patent Law and Policy: Cases and Materials at 262. The patent system also promotes 44 See, e.g., R. Levin 2/6 at 100 (stating thatinnovation, some panelists noted, by disclosure function is important and pro-competitive);demanding disclosure. Patent law requires Cohen 2/20 at 23, 34-35 (noting that patent policy aimsapplicants to disclose the inventions for through disclosure to promote innovation); Kushan 10/25 at 131 (stating that disclosure promotes innovation); Thomas 2/8 (Patent Session) at 15; Merges 2/28 at 577; 38 Frankel 4/10 at 6; Scotchmer 4/10 at 65 (noting that Kobak 7/11 at 60. disclosure obviates need for reverse engineering); Chambers 10/25 at 177 (arguing that patents permit 39 Thomas 2/8 (Patent Session) at 17 (noting inventor to talk more freely about invention); Chambers 2/8further that contracts and nondisclosure agreements are (Patent Session) at 83-84 (patents encourage less tradeimperfect). secrecy); cf. Dreyfuss 7/10 at 197 (if society makes it really hard to get patents, there will be more trade secrecy). 40 See, e.g., Thomas 2/8 (Patent Session) at 18;Teece 2/26 at 201 (patents allow small firms to specialize 45 Kenneth W. Dam, The Economicin invention). Underpinnings of Patent Law, 23 J. LEGAL STUDIES 247, 267 (1994); see also Newman 2/6 at 39 (describing patents 41 See, e.g., Thomas 2/8 (Patent Session) at 18- as “the major if not the only source of technical19 (discussing how patent system reduces need for vertical information” in “virtually all fields of technology”);integration). Armbrecht 3/19 at 51-52. 6
  • 77. disclosures that patent law demands are B. Costs Of, and Limits To,adequate,46 others noted that their adequacy Patents’ Power to Spurmight vary by industry.47 In Japan, patents Stand-Alone Innovationare reportedly a more significant source ofnew technical information than in the United Most patents do not confer marketStates.48 power on their holders,49 but when they do, they carry costs. For example, an innovator 46 Arora 2/25 at 73 (concern over whether whose patent confers market power candisclosures are adequate). cause prices of goods and services to be 47 above (and quantities to be below) See, e.g., Kahin 10/25 at 133 (arguing thatpatents induce meaningful disclosure in pharmaceuticals competitive levels. The creation of a patentindustry but not in software industry); Friedman 2/27 at monopoly can “lead[] to restriction of354-55 (contending that patent disclosures are too slow to production, a supracompetitive price, andbe of use in software industry); Thomas 10/30 at 184-85 (inmany post-industrial fields, the claim is an abstract what economists call an efficiency orbehavioral protocol and there is not much worth learning deadweight loss.”50from the description). See also infra Ch. 4(II)(B)(3)(questions about whether software disclosures are adequate,because no requirement to disclose source code). most patent applications 18 months after filing. 48 See Cohen 2/20 at 36-39; Wesley M. Cohen, 49Patents: Their Effectiveness and Role (2/20/02) (slides) at See infra Ch. 1(III)(A)(1). See also Edmund24, at http://www.ftc.gov/opp/intellect/cohen.pdf W. Kitch, Elementary and Persistent Errors in the(hereinafter Cohen Presentation), Cohen 10/30 at 84-85, Economic Analysis of Intellectual Property, 53 VAND. L.123-24 (finding patents to be the most important R&D REV. 1727, 1729-38 (2000).information source in Japan but just “in middle of pack” in 50the United States); Wesley Cohen et al., R&D Spillovers, Dam, 23 J. LEGAL STUDIES at 248; see alsoPatents and the Incentives to Innovate in Japan and the Langenfeld 2/20 at 10-13, 64-66 and James Langenfeld,United States, 31 RESEARCH POLICY 1349, 1355-56 (2002) Innovation, Competition, and Intellectual Property:(survey findings suggest that patents more effectively serve Providing an Economic Framework (2/20/02) (slides) at 4the information disclosure function in Japan than in the (arguing that strong IP rights reduce price competition, andU.S.); Janusz Ordover, A Patent System for Both Diffusion that partial IP protection would maximize economicand Exclusion, 5 J. OF ECON. PERSP. 43, 45 (1991) (stating welfare), atthat the Japanese patent system is designed to induce earlier http://www.ftc.gov/opp/intellect/langenfeld.pdf; Farrelldisclosure than the American patent system). The Japanese 2/28 at 596 (stating that IP rights can come at the cost ofpatent system apparently induces disclosure by a variety of monopoly price); Kushan 10/25 at 131 (inventors pursuemeans. For example, it awards patent rights to those who patents to try to “exploit exclusivity to a commercialfile first, inducing innovators to disclose their inventions in advantage”). Many other participants recognized suchpatent applications earlier than does the American system potential market power effects. See, e.g., ABA (Economicsof awarding patent rights to the first to invent. See, e.g., stmt) 11 (describing the exercise of market power as aOrdover, 5 J. OF ECON. PERSP. at 45; Cohen 10/30 at 123. possible cost of patent protection); Stoner 2/26 at 108-09;Moreover, Japan’s patent system also generally grants Hall 2/26 at 181, 184; Farrell 2/28 at 596; Katsh 4/10 atnarrower patents, such that there are “more patents per 25-26; MacKie-Mason 5/1 at 171; Gambrell 10/25 at 38-product” – fostering more cross licensing and related 39; Farrell 11/6 at 109-11; Ordover 11/6 at 114; Hansnegotiations and information sharing – than in the United Lennros, Question Regarding Competition & IntellectualStates. Cohen 2/20 at 37; see also Cohen et al., 31 Property (Public Comment) 1, atRESEARCH POLICY at 1356-62; Ordover, 5 J. OF ECON. http://www.ftc.gov/os/comments/intelpropertycomments/lePERSP. at 48. Two other explanations that affected survey nnroshans.htm; see also Louis Kaplow, The Patent-results – Japan’s pre-grant opposition system and its Antitrust Intersection: A Reappraisal, 97 HARV. L. REV.publication of patent applications 18 months after filing, 1813, 1821-23 (1984) (noting that “the patentee’s reward issee Cohen 10/30 at 123; Cohen et al., 31 RESEARCH made possible through monopolistic restrictions” andPOLICY at 1356; Ordover, 5 J. ECON. PERSP. at 45-46 – may discussing the difficulty of striking a balance betweenno longer be relevant. Japan has abandoned pre-grant rewarding patent holders and limiting anticompetitiveopposition, and the United States has begun publishing harm). 7
  • 78. Moreover, in the rational exercise of cause unjustified injury to consumers.55its self-interest, a patentee may sue would-be “[T]his economy is founded on the privilegerivals for infringement, deterring entry to to compete. That is the fundamental,compete. Patentee suits against entrants for bedrock principle of our capitalist economy.infringement can “tax” entry.51 The threat of . . . [W]e simply must be very concernedbeing sued for infringement by an incumbent when we manipulate our markets to restrain– even on a meritless claim – may “scare . . . competition.”56 For these reasons, oneaway” venture capital financing.52 Likewise, panelist cautioned that “[w]e should be waryaccording to panelists, a patentee may of creating unwarranted market power byprolong its market power by precluding granting unwarranted patents.”57access to technology necessary for the nextgeneration of products to emerge.53 II. COMPETITION’S EFFECTS ON INITIAL To the extent that the promise of AND FOLLOW-ONpatent protection is necessary to stimulateinvention, disclosure, or investment, then INNOVATIONsociety accepts these costs as necessary tomaximize long-term economic welfare.54 If Like patent policy, competition playsthe promise of patent protection is not an important role in spurring thenecessary for those purposes, however, then development of technologies and sequencesthe costs – which may include higher prices of related, follow-on technology.58 Thisor retarded follow-on innovation – may section discusses how a greater level of competition can affect the level of innovation, holding patent policy constant. Panelists noted that competition can 51 spur innovation in several ways, but that Lerner 2/20 at 158-61, 187; see also Weinstein2/27 at 451-52 (discussing patents as barriers to entry); economic theory and empirical evidenceStoner 10/30 at 9 (discussing patents as potential entry suggest that the effect of an increase inbarriers); Stallman 4/9 at 21, 38 (arguing that patents can competition on innovation will vary fromexclude firms from standards); Josh Lerner, Patenting inthe Shadow of Competitors, 38 J. LAW & ECON. 463, 465,489-490 (1995) (finding that high litigation costs deter 55biotechnology firms from seeking patents when rivals See, e.g., Farrell 11/6 at 109-11 (noting thatalready hold patents). costs of temporary monopoly become a matter for concern if the patents in some sense are not valid or deserved); 52 Lerner 2/20 at 189. Farrell 2/28 at 596-97 (because protecting intellectual property is a “costly way” to stimulate innovation since it 53 See Arrow 2/25 at 59-61, 64-65; see also infra sometimes allows monopoly pricing, IP protection shouldBox 2-1. be used “judiciously”). 56 54 See, e.g., Hall 2/26 at 181 (noting the trade-off Thomas 4/11 at 56.between short-term monopoly in return for incentive to 57innovate and disclose); Lunney 7/10 at 97-98 (noting that R. Levin 2/6 at 102.traditional trade-off balances incentives to innovate against 58monopoly deadweight loss). See also supra Ch. See, e.g., Partha Dasgupta & J.E. Stiglitz,1(I)(C)(1)(a) (recognizing that statutory standards for Uncertainty, Industrial Structure and the Speed of R&D,patentability govern, and that in any event, it would not 11 BELL J. ECON. 1, 25 (1980) (finding that competition inusually be possible to use a “but for” test for patentability). research and development raises the level of R&D). 8
  • 79. one context to another.59 For example, some duplicate each others’ research, which somepanelists stated that firms in a competitive believe to be a wasteful process.65 Eachmarket generally have greater incentives to point is addressed in turn below.innovate than a monopolist who does notface the threat of entry.60 Likewise, A. Competition Can Spurcompetition may drive a race to innovate, Innovation, Whether Initialspurring invention faster. The firm that or Follow-Oninnovates first may gain a patent that allowsit to exclude others, or may reap the benefits 1. Cannibalizationof its work by taking advantage of itscompetitive lead (at least when, among other Competition can drive innovation,things, copying the innovation is expensive and its power to do so may depend onor time-consuming).61 Some panelists market structure. To be sure, even acritiqued – and others defended – the so- monopolist that faces no competition has ancalled Schumpeterian hypothesis that large incentive to innovate to expand the demandfirms innovate more than small firms, and for its products and to reduce its costs.that firms in concentrated markets innovate Other things being equal, however, amore than firms in competitive markets.62 monopolist that does not face the threat ofFinally, some noted that firms competing to entry has less incentive to engage in costlyinnovate will approach research problems R&D to develop new products than does adifferently, increasing the chances of firm facing competition, some contend. Tosuccessful innovation.63 There are costs and the extent that new products wouldlimits, however, to competition’s power to cannibalize the monopolist’s existing sales,spur innovation. Patent grants are the monopolist would be less likely to findsometimes crucial to avoiding the kind of R&D expenditures worthwhile, theyfree riding that could erode incentives to maintain.66 By contrast, firms in ainnovate.64 Moreover, the innovation races competitive market have incentives tothat competition can incite can lead firms to innovate in hopes of acquiring market power, some argue.67 Similarly, the 59 See, e.g., Nelson 2/20 at 123-36 (summarizing monopolist that does face a threat of entrythe literature and concluding that “there is no simple may have more incentive to invest in R&Drelationship”). 60 See infra Ch. 2(II)(A)(1) (discussingcannibalization). 65 See infra Ch. 2(II)(B)(2) (discussing 61 See infra Ch. 2(II)(B)(2) (discussing races to duplication of effort).innovate). 66 See generally Kenneth J. Arrow, Economic 62 See infra Ch. 2(II)(B)(3) (discussing Welfare & the Allocation of Resources for Invention, inSchumpeterian hypothesis and its critics). THE RATE AND DIRECTION OF INVENTIVE ACTIVITY 609, 619 (1962), Nelson 2/20 at 126. 63 See infra Ch. 2(II)(A)(4) (discussing diversityof R&D efforts). 67 See, e.g., Shane Greenstein, Market Structure and Innovation: A Brief Synopsis (2/20/02) (slides) at 2, at 64 See infra Ch. 2(II)(B)(1) (discussing http://www.ftc.gov/opp/intellect/greenstein.pdf (hereinafterappropriability problems). Greenstein Presentation). 9
  • 80. than a prospective entrant would have, relative to that of its rivals.71 Onebecause the monopolist may have more to commentator who has studied the disk drivelose from entry than a potential entrant has industry has concluded that its patternsto gain.68 regarding competition and innovation show that “firms that trail the leader innovate2. Races to Innovate more.”72 On the other hand, some state that races to innovate may lead to wasteful The role of competition in expenditures and risky cutting of corners,stimulating R&D expenditures is perhaps and they are not necessarily efficient.73most obvious when there is a race to patent,as, for example, when two companies are Some panelists observed that whenattempting to solve the same problem and imitation is costly or time-consuming, a firmthe one that solves it first can win a patent can reap substantial benefits from innovationand exclude the other from the market. by exploiting its head start on competitors toLured by this possibility, potential inventors further develop the innovation and themay race to innovate.69 means to market it. It might enjoy a short- term monopoly on the innovation until other A number of studies have examined firms can copy it, and even after they enter,different settings where competitors race to the innovator’s established position mayachieve innovations and have concluded that help it maintain market share.74 In somethe results vary by context. For example, industries, it is enough if an innovationanalyses indicate that the effects of “permit[s] the firm to reach the market firstcompetition on innovation will vary with a product (or in most industries a newaccording to the nature of the inventiveprocess70 and a firm’s efficiency level Jennifer Reinganum, Uncertain Innovation and the Persistence of Monopoly, 73 AM. ECON. REV. 741 (1983) 68 See, e.g., Greenstein 2/20 at 140-141; (arguing that with uncertainty in the relationship betweenGreenstein Presentation at 2; FTC Staff Report, investment and the success of innovation efforts, potentialAnticipating the 21st Century: Competition Policy in the entrants have greater incentives than incumbents to seekNew High-Tech, Global Marketplace, Ch. 6, p. 12, n. 54 “drastic” (revolutionary) innovations).(May 1996) (summarizing testimony of Prof. Carlton that 71monopolists who fear the loss of their monopoly profits See, e.g., Jan Boone, Competitive Pressure:have an even greater incentive to innovate than a The Effects on Investments in Product and Processcompetitive firm), available at Innovation, 31 RAND J. ECON. 549 (2000).http://www.ftc.gov/opp/global/report/gc_v1.pdf. 72 Josh Lerner, An Empirical Exploration of a 69 See, e.g., GEORGE E. FROST, THE PATENT Technology Race, 28 RAND J. ECON. 228 (1997).SYSTEM AND THE MODERN ECONOMY, STUDY OF THE 73SUBCOMM. ON PATENTS, TRADEMARKS , AND COPYRIGHTS See infra Ch. 2(II)(B)(2).OF THE SENATE COMM. ON THE JUDICIARY, 84TH CONG., 2DSESS. 34 (Comm. Print 1957). 74 See, e.g., Edwin Mansfield, Patents and Innovation: An Empirical Study, 32 MGMT. SCIENCE 173, 70 Compare Richard J. Gilbert & David M. G. 176 (1986) (noting that patents might not seem worthwhileNewbery, Preemptive Patenting and the Persistence of in industries in which imitation is costly or difficult);Monopoly, 72 AM. ECON. REV. 514, 515-17 (1982) Stoner 2/26 at 111 (noting that a simple head start on a(arguing that when invention follows without uncertainty product can yield large profits); Michael L. Katz & Carlfrom investment, a monopolistic incumbent has an Shapiro, Systems Competition and Network Effects, 8 J.incentive to out-bid entrants in a race to invent) with ECON. PERSP. 93, 105-07 (1994). 10
  • 81. feature of an established product); other than patents were.79 With regard to newfirms are sure to follow, but only after the products, patents ranked ahead of secrecytime required for copying or reverse but behind the other three mechanisms.80engineering.”75 Again, the results showed substantial variation among industries, with patents Empirical study has shown that in proving particularly useful with regard tosome industries, firms often innovate to pharmaceutical drugs, pesticides, andexploit first-mover advantages, learning- industrial organic chemicals.81curve advantages, and other advantages, notto gain patent protection. One early study The most recent study confirms theshowed that in only two of the twelve earlier findings; it found that patents trailedsurveyed industries – pharmaceuticals and secrecy, lead time, investments inchemicals – did the firms believe patents to complementary manufacturing capabilities,be essential for developing or introducing and investments in complementary sales andthirty percent or more of the inventions.76 services as appropriability mechanisms that“[I]n office equipment, motor vehicles, businesses preferred.82 “[P]atents arerubber, and textiles, the firms were unambiguously the least central of the majorunanimous in reporting that patent appropriability mechanisms overall,” theprotection was not essential for the study concludes.83 Again, patentdevelopment or introduction of any of their significance varied sharply by industry. Forinventions during this period.”77 By example, in the medical equipment andcontrast, pharmaceutical industry pharmaceutical drug industries patents wereparticipants reported that 60% of inventions effective appropriability mechanisms forwould not have been developed and 65% more than 50% of all product innovations,would not have been commercially but for semiconductors and communicationsintroduced absent patent protection.78 A equipment patents were effective less thanlater study found that lead time, learningcurve advantages, complementary sales or 79 See Richard C. Levin et al., Appropriating theservice efforts, and secrecy were all more Returns from Industrial R&D, in BROOKINGS PAPERS ONeffective means of protecting the ECONOMIC ACTIVITY 783 (1987). This study analyzedcompetitive advantages of new processes survey responses from 650 R&D managers representing 130 lines of business. 80 Id. at 794-95. 81 Id. at 795-96. 75 Dam, 23 J. LEGAL STUDIES at 263. 82 See W.M. COHEN ET AL., PROTECTING THEIR INTELLECTUAL ASSETS: APPROPRIABILITY CONDITIONS AND 76 See Edwin Mansfield, Patents and Innovation: WHY U.S. MANUFACTURING FIRMS PATENT (OR NOT)An Empirical Study, 32 MGMT. SCIENCE 173 (1986). This (National Bureau of Econ. Research Working Paper No.study involved a random sample of 100 firms (excluding 7552, 2000), at http://papersdev.nber.org/papers/w7552very small firms) from twelve broadly defined industries (hereinafter COHEN ET AL., PROTECTING THEIRfrom 1981-1983. INTELLECTUAL ASSETS). 77 Id. at 174. 83 Id. at 9 (discussing product innovations), Figures 1 and 2 (reporting similar results for product and 78 Id. at 175. process innovations). 11
  • 82. Bo x 2-1 . Com petition for m ono poly. Allowing price to rise above marginal cost through a succession of temporary monopo lies can sp ur dynamic com petition, some have a sserted . Som e analysts argue that rapid innovation, increased importance of declining average costs, and network externalities have created conditions ideal for “dynamic” competition for monopoly, in which temporary monopolies rise and fall in the rhythms of rapid entry and exit. See, e.g., Janusz A O rdover, Antitrust for the New E cono my o r New Econ omics for Antitrust (2/20/02 ) 5, at http://www.ftc.gov/opp/intellect/020220januszord over.pdf (hereinafter Ordover (stmt)); Richard A. Posner, Antitrust in the New Economy, 68 A NTITRUST L.J. 925, 929-30 (2001). This type of competition can increase innovation, acco rding to som e observers. Low barriers to entry are critical to many of the se analyses. As noted above, several observers have stated that a monopo list threatened by entry has more to lose than any potential entrant has to gain and will therefore invest more in innovation. See Greenstein 2/20 at 140-141; D ENNIS W . C A R LT O N & J E FF R EY M . P ERLOFF , M O D E R N I ND US TRIAL O R G A NIZ AT IO N 538-40 (3rd ed. 1999). See gen erally Dasgupta & Stiglitz, 11 B ELL J. E C O N . at 25 (finding that the threat of entry may lead a monopolist to increase the pace o f research). Another panelist explained that an incumb ent mo nopolist can create barriers to entry by acquiring broad patents on critical technology. The very existence of such barriers to entry may have offsetting effects, however, because the value of winning the better-protected monopoly rises and the prospect of successful entry becomes m ore attractive. K enneth Arro w, FTC/DOJ Hearings on Competition and Intellectual Property Law and Policy in the Knowledge-Based Economy (2/25 /02) 1, at http://www.ftc.gov/opp/intellect/020225kennethjarrow.pdf; Arrow 2/25 at 64-65.27% of the time.84 3. Schumpeterian Hypothesis and its These points do not suggest that Criticspatents are unimportant. Research regardingthe relationship between patent effectiveness Panelists debated the hypothesis, originallyand R&D investments indicates that espoused by Joseph Schumpeter, that “large“[w]hile patents are not as featured as other and often monopolistic enterprises” are “themechanisms, they do stimulate R&D principal engines of technologicalbroadly, though more in some industries progress.”87 Participants discussed twothan others.”85 These three studies do dimensions of Schumpeter’s hypothesis:suggest, however, that competition also larger firms innovate more than smallerplays an important role in spurring firms, and firms in concentrated marketsinnovation.86 innovate more than firms in competitive markets.88 Economists developing 84 Schumpeter’s ideas have noted that Id. at Table 1. 85 Cohen 2/20 at 43. See ASHISH ARORA ET AL.,R&D AND THE PATENT PREMIUM 35 (National Bureau of from rivals).Econ. Research Working Paper No. 9431, 2003), at 87http://www.nber.org/papers/w9431. For evidence of a See JOSEPH SCHUMPETER, CAPITALISM,strong correlation between R&D investment and patenting SOCIALISM AND DEMOCRACY (1942); see also DENNIS W.activity, see Zvi Griliches, Patent Statistics as Economic CARLTON & JEFFREY M. PERLOFF, MODERN INDUSTRIALIndicators: A Survey, 28 J. OF ECON. LITERATURE 1661, ORGANIZATION 532-33 (3rd ed. 1999); Jennifer1673-74 (1990); see also Evenson 2/20 at 51-52 (surveying Reinganum, The Timing of Innovation: Research,international data). Development, and Diffusion, in 1 HANDBOOK OF INDUSTRIAL ORGANIZATION 849 (Richard L. Schmalensee 86 Cf. Hoerner 7/11 at 54 (stating that many & Robert D. Willig, eds. 1989).companies would engage in the same level of R&D evenwithout the patent system, because they must innovate to 88continue offering products that attract consumers away ABA (Economics stmt) 29. 12
  • 83. economies of scale may make innovation access to investment capital.93 Moreover, aless costly for a large firm.89 Specifically, number of studies have found that R&Dthey contend, large firms sponsoring spending rises proportionally to firm size inconsiderable R&D can reduce the marginal most industries, but that R&D spending bycosts of innovation by using “more large firms generates less innovation perspecialized resources;” can spread the fixed dollar than does spending by smaller firms.94costs of any R&D over a wider base of And some have stated that the weight ofoutput; can spread the risk of unsuccessful economic theory and evidence shows thatR&D efforts by sponsoring many R&D there is a non-linear, inverted-U-shapedprojects simultaneously; and have access to relationship between concentration andinexpensive investment capital, drawn from innovation. In their view, low concentrationthe firm itself or from capital markets.90 may not be conducive to innovation, butMoreover, some commentators state that “very high concentration has a positivelarge firms benefit from their own effect only in rare cases, and more often it isinnovative efforts more than smaller firms apt to retard progress by restricting thedo: large firms can apply their process number of independent sources of initiativeinnovations to large production operations, and by dampening firms’ incentive to gaingaining greater savings; the chances that an market position through acceleratedinnovation will be useful to one of their R&D.”95 Under this view, “[w]hat is neededmany businesses is greater; and their for rapid technical progress is a subtle blendabilities to market their innovations to others of competition and monopoly, with moremay be greater.91 Studies also have revealed emphasis in general on the former than thea positive correlation between concentration latter, and with the role of monopolisticand industry R&D/sales ratios, although that elements diminishing when richcorrelation may break down at high levels of technological opportunities exist.”96concentration.92 Some panelists critiqued theSchumpeterian hypothesis directly. They 93 See, e.g., Teece 2/26 at 195; SCHERER & ROSS,noted, for example, that venture capital INDUSTRIAL MARKET STRUCTURE AND ECONOMICbreaks the link between innovation and the PERFORMANCE at 630, 652 (noting that growth of a venturefinancial resources of a firm, undermining capital industry in United States that can “channel[] investment into new high-technology firms shows that pastthe argument that large firms have unique monopoly profits are no sine qua non for supporting innovation”). 94 See Wesley Cohen & Steven Klepper, A 89 Id. Reprise of Size and R&D, 106 ECON. J. 925, 927-30 (summarizing prior research), 947 (suggesting that large 90 Id. at 29-30. firms may have greater incentives to undertake marginal research projects) (1996); see also Shelanski 2/25 at 25-36 91 See, e.g., Id. (summarizing these arguments). (critiquing Schumpeter theory and noting lack of good empirical support). 92 See, e.g., F.M. SCHERER & DAVID ROSS, 95INDUSTRIAL MARKET STRUCTURE AND ECONOMIC SCHERER & ROSS, INDUSTRIAL MARKETPERFORMANCE 646 (3d ed. 1990). For a discussion of the STRUCTURE AND ECONOMIC PERFORMANCE at 660.possible implications of Schumpeterian theories for 96dynamic competition for monopoly, see Box 2-1. Id. 13
  • 84. Other panelists contended that the concentration and innovation.101 To theSchumpeterian hypothesis is true for some contrary, many studies seem to suggest thatindustries and markets but not true in others. the effect of concentration on innovationFor example, one panelist stated that depends on many factors.102 For example,industry conditions are so varied that it some statistical evidence suggests that thewould be surprising to find a “simple existence of an inverted-U relationshipSchumpeterian relationship” across all between concentration and innovationindustries.97 Likewise, another panelist depends on industry characteristics.103 Somestated that “result[s] vary a lot depending onthe structure and nature of the industry.”98Indeed, two studies that controlled for inter- 101 See, e.g., Gilbert 2/25 at 12-14; SCHERER &industry differences found reason to ROSS, INDUSTRIAL MARKET STRUCTURE AND ECONOMIC PERFORMANCE at 645-51 (noting that some statisticalquestion various facets of the Schumpeterian evidence points to a positive relationship between industryhypothesis.99 In a similar vein, some have concentration and R&D/sales ratios, although that correlation may break down at high levels ofsuggested that policymakers examine “the concentration); Richard J. Gilbert & Steven C. Sunshine,relationship between concentration, R&D Incorporating Dynamic Efficiency Concerns in Mergeractivity, and innovation” in particular Analysis: The Use of Innovation Markets, 63 ANTITRUST L.J. 569 (1995) (stating that firm- and industry-specificindustries, because “industries probably vary factors complicate the relationship between concentrationtoo much for one theory to fit all.”100 and innovation); Richard Rapp, The Misapplication of the Innovation Market Approach to Merger Analysis, 64 ANTITRUST L.J. 19 (1995) (citing the inconclusive Statistical cross-section studies economic literature on the relationship betweenexamining multiple industries have not concentration and innovation); Richard J. Gilbert & Stevenidentified any clear relationship between C. Sunshine, The Use of Innovation Markets: A Reply to Hay, Rapp, and Hoerner, 64 ANTITRUST L.J. 75, 76-77 (1995) (suggesting that industry-specific factors obscure the statistical relationships); Shelanski 2/25 at 32 (stating that the “empirical data do not resolve any of the ambiguity in the relationship between competition and innovation,” and that the “empirical evidence is really quite ambivalent”). 97 Nelson 2/20 at 132-36. 102 See, e.g., Gilbert & Sunshine, 64 ANTITRUST L.J. at 76-77 (stating that “many factors influence the 98 Rubinfeld 2/25 at 20. incentive to invest in the development of new products and processes”). 99 See Wesley Cohen et al., Firm Size and R&DIntensity: A Re-examination, 35 J. INDUS. ECON. 543, 543- 103 See, e.g., SCHERER & ROSS, INDUSTRIAL544 (1987) (questioning linkage between firm size and MARKET STRUCTURE AND ECONOMIC PERFORMANCE atintensity of R&D); P.A. Geroski, Innovation, 645-51 (noting that including variables such as R&DTechnological Opportunity, and Market Structure, 42 performed outside the industry, the pace of innovation, andOXFORD ECON. PAPERS 586, 586 (1990) (finding, based on the strength of appropriation mechanisms weakens thedata from the United Kingdom and a variety of measures of inverted-U relationship in some industries, which points tomarket structure, “fairly strong evidence against the the importance of firm and industry effects in qualifying thehypothesis that increases in competitive rivalry decrease relationship), Nelson 2/20 at 128 (noting potential data andinnovativeness”). statistical problems with at least some of the studies that have found inverted-U relationships). For a recent working 100 DENNIS W. CARLTON & ROBERT H. GERTNER, paper finding inverted-U relationships in data involvingINTELLECTUAL PROPERTY, ANTITRUST AND STRATEGIC United Kingdom firms, see PHILIPPE AGHION ET AL.,BEHAVIOR 14 (National Bureau of Econ. Research COMPETITION AND INNOVATION: AN INVERTED UWorking Paper No. 8976, 2002), at RELATIONSHIP (National Bureau of Econ. Researchhttp://www.nber.org/papers/w8976. Working Paper No. 9269, 2002) (describing the inverted-U 14
  • 85. 4. Diversity of R&D Efforts Bo x 2-2 . Add itional Pricing Strategies. Panelists described a number of non-patent pricing strategies that firms may use to recoup fixed costs, including R&D Several panelists discussed the spending. importance of diverse research efforts in producing innovation. One panelist noted Lon g-run averag e costs. One panelist noted that firms that when many firms devote R&D efforts with declining average costs will not price at marginal co st because the y must recover their substantial fixed costs. to tackling the same problem, the public Ordover (stmt) 2. Another panelist echoed that a firm benefits.105 Likewise, another panelist charging a flat price must set it higher than ma rginal co st if noted that “if you have fewer innovators it has returns to scale. Varian 2 /25 at 76. One pa nelist [and] less diversity, you are likely to have suggested that long-run average costs may be a useful less innovation or higher prices or lower analytical benchmark, but adde d that it is difficult to determine w hich of a firm’s fixed costs correspond to quality products.”106 He illustrated his individual pro ducts and that som e temp orary returns in point by discussing a proposed merger excess of that benchmark may be necessary for adequate that, he stated, might have stifled incentives to innovate. See Ordover (stmt) 3. innovation in a market “where the strategy of innovation is highly unpredictable [and Price discrimination. Some m aintain that, rather than use where] path-breaking innovations . . . are constant per-unit prices, firms have begun to adopt more “sophisticated” models of pricing – such as volume or made by niche players and not by the loyalty discounts, bundling, and self-selective price leading incumbents.”107 Indeed, some discrimination – as a means of covering substantial up- commentators have observed that under front inve stments, such as R& D sp ending. See Varian 2/25 certain conditions, rates of innovation are at 76-79. positively correlated with rates of entry.108 Nevertheless, others suggested that the ability of diverse R&D efforts to affectindustry case studies indicate that innovation depends on a key industrycompetition drives innovation in particularindustries.104 105 See Arrow 2/25 at 58-59 (stating that “diversity is good” with respect to “differing sources of R&D”); see also Thomas M. Jorde & David J. Teece, Innovation and Cooperation: Implications for Competition and Antitrust, 4 J. ECON. PERSPECTIVES 75, 81relationship), at http://nber.org/papers/w9269.pdf. (1990) (acknowledging that “horizontal cooperation” in research “may reduce diversity”). 104 See, e.g., Lerner, 28 RAND J. ECON. at 244 106(empirical study of the computer disk drive industry Rubinfeld 2/25 at 19.showing that “the greatest innovative activity is shown by 107firms that follow the leader”); Gilbert 2/25 at 12 (noting Rubinfeld 2/25 at 22-23; see also Danielthat the correlation between competition promoting Rubinfeld & John Hoven, Innovation and Antitrustinnovation characterizes “almost any [sector of ] the Enforcement, in DYNAMIC COMPETITION AND PUBLICsoftware industry,” including operating systems and POLICY 65, 87-88 (Jerry Ellig, ed. 2001) (noting need forInternet browsers, as well as semiconductors); Gilbert & diversity of innovation).Sunshine 63 ANTITRUST L.J. at 580-81 (noting evidence 108and industry case studies that “support the stronger See James Bessen & Eric Maskin, Sequentialconclusion that protection from competition i[s] inimical to Innovation, Patents, and Imitation (Public Comment) 13-technological progress”); MICHAEL E. PORTER, THE 15, atCOMPETITIVE ADVANTAGE OF NATIONS 143 (1990) http://www.ftc.gov/os/comments/intelpropertycomments/ji(“[R]ivalry has a direct role in stimulating improvements mbessenericmaskin.pdf (hereinafter Bessen & Maskinand innovation”). (stmt)). 15
  • 86. characteristic: the predictability of successful inventor.”113subsequent R&D paths.109 Yet what some deem wastefulB. Costs Of, and Limits To, duplicative efforts is what others deem Competition’s Power to Spur useful competition.114 Firms compete via their R&D efforts, and such competition Innovation generates better consumer products and lower prices, benefits that may outstrip any1. Appropriability Problems social loss from the patent race, some observe.115 Some have noted that the As discussed above, however, benefits accruing from diverse efforts atpanelists noted that competition cannot innovation may outweigh the waste involvedserve as the sole driver of innovation. in competitive innovation.116 They argueInventors sometimes cannot appropriate that the potential wastefulness of parallelvalue from the invention without the grant of R&D efforts should not influence publica patent, making patents an important policy decisions:incentive for innovation in such settings.110 [W]e do not normally consider the2. Duplication of Effort opening of a new gasoline station or grocery store near an existing one to Some analysts have underscored one be an example of waste, or at leastof the costs of competition to innovate: not one with which public policyduplication of effort involved in parallel should be concerned, even thoughresearch efforts.111 “Independent research we believe that only one can surviveactivities often proceed down identical or and we know that some economicnear-identical technological paths,” making rent of location may accrue to thea policy of encouraging diversity in R&D survivor. Rather, we consider thepaths unhelpful, in their view.112 They argue competition induced by the newthat excess efforts at innovation generate entrant to lead to a better outcome“wasteful patent race[s] to be the first 113 Stoner 2/26 at 108-09; see also Mark F. Grady & Jay I. Alexander, Patent Law and Rent 109 See infra Ch. 2(III)(A)(1). Dissipation, 78 VA. L. REV. 305, 308 (1992) (positing that competition to be the first to develop pioneer and follow-on 110 See supra Ch. 2(I)(A) (discussing patents’ innovations causes overinvestment that “dissipates,” orpower to internalize externalities and protect against free eliminates, the benefit to society of the innovation or itsriding). For a discussion of non-patent pricing strategies improvement).that firms may use to recover fixed research and 114development costs, see Box 2-2. See, e.g., Dam, 23 J. LEGAL STUDIES at 263 (making this point). 111 Cf. William M. Landes & Richard A. Posner, 115Trademark Law: An Economic Perspective, 30 J. LAW & See, e.g., Dam, 23 J. LEGAL STUDIES at 252,ECON. 265, 267-68 (1987) (arguing that rent seeking is as 263.wasteful as having many parties search for lost treasure). 116 See Robert P. Merges & Richard R. Nelson, 112 Jorde & Teece, 4 J. ECON. PERSPECTIVES at On the Complex Economics of Patent Scope, 90 COLUM. L.81. REV. 839, 873, 877 (1990). 16
  • 87. than would accrue thorough legal on innovation organized by the initial protection of the exiting firm. So, innovator versus that conducted by too, we cannot have much independent innovators. On the one hand, confidence that some of the natural some argued that strong initial patent rights alternatives to competition in R&D can facilitate follow-on innovation by, or would increase social welfare.117 under the management of, the initial innovator. For example, some haveIII. PATENTS’ EFFECTS ON contended that broad initial patent grants can FOLLOW-ON allow the original patentee to organize its licensees’ research into the patent’s INNOVATION prospects, avoiding wasteful patent races.119 Others, however, disagreed, stating that Finally, it is appropriate to address subsequent researchers acting independentlythe effects of patent grants on follow-on of the original inventor and competinginnovation. Innovation is often an ongoing, against each other may foster greatercumulative process, with each generation of innovation – and may have less marketinnovations building on what came before.118 power in any resulting innovation.120For example, knowledge gained throughbasic research may serve as a foundation for Second, this section considers thesubsequent applied activities; new products implications for independent follow-onor services may go through multiple innovation of a single, blocking, initialgenerations of improvements and extensions patent. Sometimes the follow-on innovatorof use; initial research may produce tools – will seek to design around the initial patent,from laser technology through specialized potentially generating new technologies, butsoftware programs and isolated, purified also incurring R&D costs.121 Other times thegenetic material – that follow-on research follow-on innovator will license the patentedthen applies to develop products and technology. This section examines theservices for end-use consumers. In each division of rewards between initial andcase, the question arises whether policies follow-on innovators through such licensingand laws suitable for fostering a single and considers some of the impediments thatgeneration of inventions also maximize might interfere with achieving licensingwelfare in the more dynamic, cumulative arrangements that adequately reward bothinnovation settings actually observed. This generations of innovators.122section explores these issues. First, this section identifies the 119relative strengths and weaknesses of follow- See infra Ch. 2(III)(A)(1) (discussing follow- on innovation organized by the initial innovator). 120 See infra Ch. 2(III)(A)(2) (discussing follow- 117 Dam, 23 J. LEGAL STUDIES at 263. on activities by independent follow-on innovators). 121 118 See, e.g., ABA (Economics stmt) 20, 24; See infra Ch.2(III)(B)(1) (discussing design-Lemley 2/25 at 37; Scotchmer 2/26 at 128-29; Suzanne around innovation).Scotchmer, Standing on the Shoulders of Giants: 122Cumulative Research and the Patent Law, 5 J. ECON. See infra Ch.2(III)(B)(2)-(3) (discussingPERSP. 29 (1991). division of rewards and licensing). 17
  • 88. Finally, this section considers innovator’s future development efforts andimplications for follow-on innovation in the force it to license away its technology.128face of multiple existing patents. Each point is discussed in turn below.Sometimes, the need to attain access tomultiple existing patents in the hands of A. The Roles of Managed andmultiple patentees can pose difficulties for Independent Follow-Onindependent follow-on innovators. This Innovationproblem may flow just from the transactioncosts of negotiating multiple licenses.123 Panelists discussed follow-onMoreover, the necessary patents may be too innovation from the perspective of twonumerous to identify and license; follow-on general models. Under one model, follow-innovators may almost inevitably risk suit on innovation proceeds under the controlfor infringement once they sink costs into and management of the initial innovator.their research or production efforts.124 An That innovator might conduct follow-onadditional problem may affect the level of activities itself. It also might effectivelythe multiple royalties: the patentees, acting “hire” others to do some of the follow-onindependently, may seek a higher total work, licensing them to use its technologyroyalty than would a single patentee for follow-on research and development.charging a package price.125 Furthermore, Both mechanisms are forms of “managed”some argue that oligopolists holding a follow-on innovation. Alternatively, follow-collection of necessary patents can injure on activity may proceed independently of theand block follow-on innovation by refusing initial innovator’s coordination or control,to license, or charging high royalty rates, to with an array of outside researchers eachentrants.126 A patentee may use multiple seeking to build upon prior discoveries, apatents on near substitutes for its original model that this Report terms “independentwork to retard independent follow-on follow-on innovation.” When a priorinnovation and impede entry, some discovery is patented, an independentcontend.127 Finally, some suggest that, follow-on innovator may need a license, andunder certain conditions, the initial the patentee may or may not wish to grant it.innovator’s rivals might use multiple patentson trivial variants to constrain the initial 1. Follow-On Innovation Organized by the Initial Innovator 123 See infra Ch.2(III)(C)(1) (discussingtransaction costs). In some instances, an initial innovator with a broad patent covering 124 See infra Ch.2(III)(C)(2) (discussing hold up future development opportunities mightin the patent thicket). pursue, or organize others to pursue, the 125 See infra Ch.2(III)(C)(3) (discussing royalty follow-on innovations. Professor Edmundstacking and the Cournot complements problem). Kitch emphasized several advantages of 126 See infra Ch.2(III)(C)(4) (discussing such arrangements. Broad, initial patentoligopoly and group boycotts). 127 128 See infra Ch.2(III)(C)(5) (discussing patent See infra Ch.2(III)(C)(6) (discussing patentfences). flooding). 18
  • 89. rights can protect appropriability, not just for questioned whether initial innovators areinitial inventions but for the full range of likely to provide effective centralfollow-on activities needed to bring management;132 no one decision maker mayproducts to market. Broad initial patent have the range of knowledge necessary torights enable the innovator to provide choose the best follow-up opportunities or toefficient, central management of the select the ideal follow-up researchers.133subsequent development efforts, avoiding Others noted that the theory depends onunnecessary duplication of R&D activity and efficient licensing of follow-onwasteful racing for follow-on patent rights. opportunities, but that licensing negotiationsBroad initial patent rights permit innovators may be lengthy and costly or break downto disclose information without fear of free due to differences in valuations.134 Stillriding, thereby facilitating access to others stressed that the efficiencies realizedfinancing, complementary technology, and may be private, not social – arguing thatspecialized supplies.129 follow-on patent races, although costly, may benefit consumers by yielding products These considerations are key sooner and with more certainty, and thatelements of what has come to be known as coordination may eliminate desirablethe “prospect theory” of patent rights. Theprospect theory focuses on exploration of 132technological opportunities, referred to as See, e.g., Lemley 2/25 at 37-38 (central management by initial innovator an unwise “gamble” when“prospects.”130 It emphasizes the effect of innovation is likely to be cumulative); Frederick M.patents on commercialization, as opposed to Scherer, The Economics of Human Gene Patents, 77a view that emphasizes the effect of patents ACADEMIC MEDICINE 1348, 1362 (2002).on incentives to invent. Its perspective is 133 See, e.g., Rubinfeld 2/25 at 20 (“very hard exforward looking, focusing on the efficient ante to know who is going to be successful”); see alsocoordination of, and incentives for, follow- Mark A. Lemley, The Economics of Improvement inon activities.131 Intellectual Property Law, 75 TEX. L. REV. 989, 1048-52 (1997); Merges & Nelson, 90 COLUM. L. REV. at 873 (“[N]o one knows for sure what possible inventions are in Several panelists identified potential the technological pool. . . . The only way to find out whatshortcomings of this prospect theory. Some works and what does not is to let a variety of minds try.”); Rebecca S. Eisenberg, Patents and the Progress of Science: Exclusive Rights and Experimental Use, 56 U. CHI. L. REV. 1017, 1066-69 (1989) (arguing that central 129 See Kitch 2/20 at 79-87; Edmund W. Kitch, management by an initial innovator is least likely to beThe Nature and Function of the Patent System, 20 J. LAW successful when follow-on research is likely to “lead[]& ECON. 265 (1977). According to Professor Kitch, a down unexpected paths” and when it “depends on insightsbroad patent places its owner in a position “to coordinate and creativity that may differ from one investigator to thethe search for technological and market enhancement of the next, ” and concluding from this that the prospect theorypatent’s value so that duplicative investments are not made may work better in contexts involving applied, rather thanand so that information is exchanged among the searchers.” basic, research).Id. at 276. Broad patents also permit the owner “to makeinvestments to maximize the value of the patent without 134 See, e.g., O’Rourke 2/20 at 98; Hall 2/26 atfear that the fruits of the investment will produce 182-83; Merges & Nelson, 90 COLUM. L. REV. at 874-75unpatentable information appropriable by competitors.” Id. (noting the “steep transaction costs of technologyat 276. licensing”); see also infra Ch. 2(III)(B)(3). Professor Kitch acknowledged these concerns but argued that the fact that 130 See Kitch, 20 J. LAW & ECON. at 266. negotiation sometimes breaks down or proves costly does not reveal how well the process works as a whole. Kitch 131 See id.; Scotchmer 2/26 at 129. 2/20 at 109. 19
  • 90. competition in the market for follow-on complementary assets.139 When research isproducts.135 complete and follow-on products enter the market, their derivation from independent2. Follow-On Activities by lines of development may result in less Independent Follow-On market power than when the initial Innovators innovator controls follow-on innovation.140 Follow-on innovation often proceeds Independent follow-on innovation, ofthrough the activities of inventors course, might entail substantial duplicationindependent of the initial innovator.136 of effort.141 Some scholars condemn thisIndependent follow-on innovation has all of duplication as wasteful, rent-seekingthe potential benefits identified supra in Ch. activity. Professor Mark Grady, for2(II)(A), discussing the role of competition example, explains that when an initialin spurring innovation. Competition may innovation signals opportunities for follow-prod follow-on innovation efforts to proceed on inventions, hopeful inventors maymore quickly.137 It may foster greater “redundantly waste efforts to find anddiversity of R&D activity, providing broader capitalize on that method ofrange for identifying research opportunities, improvement.”142 Others caution, however,designing and pursuing research paths, and that what to the firms involved is wastefulrecognizing and acting upon the implications duplication of effort may have socialof research results.138 It may overcome benefit.143 As Professor Suzanne Scotchmerbiases in the initial innovator’s choice of explained, coordinating follow-on activitiesfollow-on research projects attributable to itsfirm-specific skills or investments in 139 See generally Gilbert & Sunshine, 63 ANTITRUST L.J. at 577; Merges & Nelson, 90 COLUM. L. REV. at 873 (“Once a firm develops and becomes competent in one part of a ‘prospect,’ it may be very hard for it to give much attention to other parts, even though in 135 See Scotchmer 2/26 at 136-39; Suzanne the eyes of others, there may be great promise there.”).Scotchmer, Competition Policy and Innovation: The 140Context of Cumulative Innovation (2/26/02) (slides) at 7, at Scotchmer 2/26 at 136-39 and Scotchmerhttp://www.ftc.gov/opp/intellect/020226suzanneandersonsc 2/26 Presentation at 7.otchmer.pdf (hereinafter Scotchmer 2/26 Presentation);Bessen & Maskin (stmt) 4 (“increasing the number of firms 141 See Stoner 2/26 at 112-13; Kitch, 20 J. LAWin pursuit of a solution raises the probability that someone & ECON. at 276. See also supra Ch. 2(II)(B)(2).will succeed”) (emphasis in original). 142 Grady & Alexander, 78 VA. L. REV. at 308. 136 Indeed, one analyst finds independent follow- The authors argue that many aspects of patent law can beon efforts the predominant pattern. See Scherer, 77 explained as reflecting a desire to limit rent dissipation. Id.ACADEMIC MEDICINE at 1362 (“It is more the norm than at 308-10. They note, though, that this effort may provethe exception in the history of technology for the firms complex: a system that awards a broad initial patent tointroducing significant derivatives of and improvements discourage wasteful follow-on races could unintentionallyupon a basic discovery to be other than the original encourage duplicative efforts to win the initial patent. Id.discoverer.”). at 308 (“The obvious compromise is to grant protection broad enough to prevent a race to improve . . . but not so 137 See Scotchmer 2/26 at 137-38. broad as to create wasteful races for other patent goldmines.”). 138 See, e.g., Arrow 2/25 at 58-59; Barton 2/26 at 143172-73. See supra Ch. 2(II)(B)(2). 20
  • 91. by eliminating patent races may increase the 1. Design-Around Innovationresearch firms’ profits but harm consumers.“[T]ypically, the patent race will get us the Several panelists stressed that aproduct sooner, and may get us the product significant benefit of the patent system is itswith higher probability,” she stated.144 Over role in directing R&D away from imitationall, the debate suggests that duplication may by forcing competitors to design aroundentail elements of both social benefit and existing patents. In the long run, theyundesirable waste.145 argued, re-directing R&D toward more innovative goals encourages greaterB. Follow-On Innovation in the technological progress.146 One panelist, for Face of a Single Blocking, example, explained that patent protection of the ulcer-treating drug Tagamet forced Initial Patent design-around efforts that led to the development of another successful drug, The Hearings identified two distinct Zantac;147 others cited Xerox’s photocopyingsets of issues that the patent system raises technology, which developed out of an effortfor independent follow-on innovation. First, to design around Kodak’s silver halideinitial innovation may give rise to individual photography patents and which, in turn, gavepatents that block certain follow-on impetus to design-around research thatactivities. This section discusses two generated ink-jet technology.148potential responses: (i) directing follow-oninnovation around the blocking patent or (ii) Other panelists pointed to the design-negotiating with the initial patentee for a around theory’s limits. In some settingslicense to permit the follow-on activities to design-around may be technicallygo forward. Second, in some settings, impossible.149 In other settings, such asfollow-on activities may require numerous,distinct pieces of patented technology toproceed; the special problems this may pose 146 See, e.g., Myrick 3/19 at 20 and 10/30 at 40-here are analyzed infra in Chapter 2(III)(C). 42; Frankel 4/10 at 7; Banner 10/30 at 71; Frederick J. Telecky, Statement of Frederick J. Telecky, Jr., Senior Vice President and General Patent Counsel, Texas Instruments: 144 Scotchmer 2/26 at 137 (terming this “a FTC/DOJ Hearings on “Competition and Intellectualconflict between the private incentives to cut back on R&D Property Law and Policy in the Knowledge-Basedand the social incentives”). Economy” (2/28/02) 3, at http://www.ftc.gov/opp/intellect/020228telecky.pdf 145 A focus on duplicative research efforts reveals (hereinafter Telecky (stmt)).both facets. On the one hand, Professor Rebecca Eisenberg 147finds merit in overlapping research, arguing that “different See Armitage 3/19 at 230.investigators are likely to make different observations and 148have different ideas for follow-up experiments, improving See Varian 2/25 at 94; Sobel 7/10 at 175.the chances for serendipitous discoveries” and that “[e]ven 149completely duplicative research efforts may serve a See, e.g., Barr 10/30 at 90 (broad patents canvaluable function by confirming research results and prevent design-around); Detkin 2/28 at 668 (“unavoidableenhancing the likelihood that a discovery will be noticed.” overlap of IP” in semiconductor technology); RichardSee Eisenberg, 56 U. CHI. L. REV. at 1068-69. On the other Stallman, The Danger of Software Patents Speech byhand, Professor Kitch finds unnecessary waste when initial Richard Stallman at Cambridge University, March 25research is kept secret and follow-on researchers must tread 2002 (Public Comment) 4 (“no way around that patent. . . .the same ground without knowing of or learning from the nothing else you could do like that”), atprior failed efforts. See Kitch, 20 J. LAW & ECON. at 276. http://www.ftc.gov/os/comments/intelpropertycomments/st 21
  • 92. when the patented technology is needed to initial innovation provides a benefit to theconform to a standard or consumers are follow-on innovator, and the full socialotherwise locked in or when the infringing benefit of the initial innovation includes aapproach is already built into a competitor’s portion of the follow-on benefits that itproduct before the patent issues, design- confers. Consequently, providing the initialaround may be economically impossible.150 innovator some share in the returns from theIn still other contexts the design-around may follow-on activity may be efficient.153add little value, merely requiring thatcompetitors “work harder to get to the same Optimal sharing arrangements,place.”151 Indeed, analysts emphasize that however, may prove elusive, for shiftingdesign-around is not costless, but rather rewards from one generation to another mayconsumes resources that, absent the initial reduce incentives at the disadvantagedpatent, might be more fruitfully employed.152 generation. “The challenge is to rewardWithout a clear basis for assessing the net early innovators for the technologicalvalue of design-around activity, general foundation they provide to later innovators,conclusions are difficult. but to reward later innovators adequately for their improvements and new products as2. Division of Rewards well.”154 Rather than designing around an Royalty payments from the follow-initial patent, an independent follow-on on innovator are a means for implementinginnovator may acquire a license to the the sharing arrangements. Standards ofpatented technology and proceed with patentability, discussed in Chapter 4, shapedevelopment of products or processes within the backdrop against which licensingthe patent’s coverage. In such cases, thedivision of rewards between the initial andfollow-on innovators becomes crucial, 153 See Scotchmer 2/26 at 128-29; Scotchmer, 5because it determines the level of incentives J. ECON. PERSP. at 31 (“First innovators will have correctfor each generation of innovation. The incentives to invest only if they receive some of the social surplus provided by second generation products.”). Of course, as already noted, innovation may be continuous, so that the “follow-on” innovator at one stage in the cycle becomes the “initial” innovator at the next. See Scotchmerallmanrichard.pdf. 2/26 at 170. 150 See, e.g., Stallman 4/9 at 19, 88-89; Barr 154 Scotchmer, 5 J. ECON. PERSP. at 30.10/30 at 79. Substantial additional literature explores the economically optimal division of profit between initial and follow-on 151 See Stallman 4/9 at 38; cf. F. M. SCHERER, innovators. See, e.g., HUGO A. HOPENHAYN & MATTHEWINDUSTRIAL MARKET STRUCTURE AND ECONOMIC F. MITCHELL, INNOVATION FERTILITY AND PATENT DESIGNPERFORMANCE 446 (2d ed. 1980) (noting both “examples (National Bureau of Econ. Research Working Paper No.and counterexamples” of valuable and essentially 7070, 1999), at http://www.nber.org/papers/w7070.pdf;duplicative design-around research). Suzanne Scotchmer, Protecting Early Innovators: Should Second-Generation Products be Patentable, 27 RAND J. 152 See id.; Kitch, 20 J. LAW & ECON. at 278-79. ECON. 322 (1996); Jerry R. Green & Suzanne Scotchmer,Stated differently, the design-around process may re- On the Division of Profit in Sequential Innovation, 26introduce some of the same duplications of effort outside RAND J. ECON. 20 (1995); Howard Chang, Patent Scope,the scope of an initial patent that are discouraged within the Antitrust Policy, and Cumulative Innovation, 26 RAND J.patent’s coverage. ECON. 34 (1995). 22
  • 93. negotiations occur. When the initial exposed: it must secure a license now, afterinnovator obtains a narrow patent, so that its investments are sunk. Faced withthe follow-on innovation does not infringe, opportunistic demands, the follow-onthe initial innovator will receive no royalty. innovator may not receive rewards adequateIt may still benefit if the follow-on for its contribution.157 If this were theinnovation is a complement that increases established pattern, socially efficient levelsthe value of the initial innovation, but the of independent follow-on innovation couldinitial innovator will suffer without not be sustained.158compensation if the follow-on innovation isa substitute. If instead the initial innovator Negotiation is more likely to dividereceives a broad patent, so that the follow-on rewards to support efficient follow-oninnovation infringes, the initial innovator activity if licensing occurs ex ante, that is,can force the follow-on innovator to take a before the follow-on innovator makes itslicense for the initial technology and share sunk investments.159 Although incentives tosome of the follow-on benefits through the enter ex ante licenses often may beensuing royalties.155 If the follow-on present,160 the Hearings and relatedinnovator garners a patent on its scholarship suggested reasons that licensingimprovement, it may have some negotiating may not occur ex ante in someleverage of its own; the patents are mutually circumstances.161blocking, and if the initial innovator wantsaccess to the improvement, it will need togive as well as take.156 157 See, e.g., Scotchmer 2/26 at 135; Rai 4/10 at 19; Green & Scotchmer, 26 RAND J. ECON. at 21, 23-24.3. Licensing Nevertheless, the initial innovator generally would not have an incentive to charge a royalty so high that the follow-on company would exit. The timing of negotiations affectswhether licensing arrangements will 158 Of course, follow-on innovation that is veryadequately reward both initial and follow-on valuable, and patent-protected, may still be profitable. Seeinnovation. Results are most likely to be Green & Scotchmer, 26 RAND J. ECON. at 25. Thus, some panelists argued that when improvements are significantproblematic when licensing occurs ex post, and adequate information is available, awarding a blockingthat is, after the follow-on innovator has position to the follow-on innovator may sufficiently protectincurred the sunk costs of its R&D efforts. that innovator even when licensing negotiations are conducted ex post. See Parkhurst 4/10 at 93-94; Kieff 4/10At that point, the follow-on innovator is at 163-64. 159 See Stoner 2/26 at 118-19; Scotchmer 2/26 at 155 See, e.g., Green & Scotchmer, 26 RAND J. 135.ECON. at 21 (“Because the breadth of the first patent 160determines whether a product infringes, it thus determines See, e.g., Kieff 4/10 at 163 (“let’s assume Ithe division of profit.”). have no idea where the big commercial utility is – I want to license everyone in the room in the hope that they find a 156 See, e.g., O’Rourke 2/20 at 103-04 commercial utility, because then I get a piece of that pie”);(describing the mutual infringement situation as the Blackburn 2/26 at 264-65 (“when you cannot predict ahead“blocking patent doctrine”); American Intellectual Property of time the incentive is there to broadly license”).Law Association (AIPLA), AIPLA Testimony (Public 161Comment) 3, at For full discussion of many of the possiblehttp://www.ftc.gov/os/comments/intelpropertycomments/ai licensing impediments, see Lemley, 75 TEX. L. REV. atpla.pdf. 1050-61; Eisenberg, 56 U. CHI. L. REV. at 1073-74. 23
  • 94. • Some analysts stress the agreement.166 One analyst highlights thepotential licensee’s exposure in bringing a potential for bargaining stalemates when thefollow-on idea to the initial innovator. In an initial innovation involves basic researchex ante context, before the follow-on with little commercial value itself and theinnovator has made its R&D investment, the follow-on innovations require substantialfollow-on idea would not be patent- investment.167protected, and the initial innovator mightmisappropriate it. Contracts to protect • In some circumstances, theagainst such conduct may prove initial innovator may not have a privateinadequate.162 incentive to license. Some panelists cautioned that firms may be reluctant to • Some analysts suggest that license others who may eventually prove totransaction costs of ex ante licensing may be competitors. When in-houseprove high.163 The negotiations may be development works to enhance or maintainfraught with uncertainty because the subject market power, the initial innovator maymatter entails research that has not yet been serve its self-interest by forgoing sociallyconducted.164 There may also be substantial beneficial licensing.168uncertainty regarding the validity and scopeof the initial innovator’s patent rights.165 Others responded that transaction costs and the effects of uncertainty usually can be • Divergent views regarding overcome, that the holder of an upstreamthe relative value of initial and follow-on patent has the incentive to assure thatcontributions may prevent reaching downstream products reach the market,169 and that if licensing to follow-on innovators 162 166 See Lemley, 75 TEX. L. REV. at 1051; See, e.g., Rai 4/10 at 19. One mechanism forScotchmer, 5 J. ECON. PERSP. at 36 n.11; Eisenberg, 56 U. resolving uncertainties and divergent views regarding theCHI. L. REV. at 1063, 1073. The argument is an illustration likely value of follow-on research involves the use ofof the point that patents facilitate efficient transfers of licenses with reach-through royalties, that is, royaltiesinformation. See infra Ch. 2(I)(A)(2). measured as a percentage of the sales of the follow-on product or service. For discussion of the use of reach- 163 See, e.g, O’Rourke 2/20 at 98; Lemley, 75 through royalties in biotechnology contexts, see infra Ch.TEXAS L. REV. at 1053-55; Merges & Nelson, 90 COLUM. 3(III)(E)(1). For discussion of some of the legal andL. REV. at 874. economic issues posed by reach-through royalties, see Second Report (forthcoming). 164 See, e.g., Lemley, 75 TEXAS L. REV. at 1053 167(“if it is hard to value an invention that has already been See e.g., Scherer 7/10 at 56 (noting themade, it is well-nigh impossible to value one that might be combination of technical and market uncertainty); Scherer,made in the future”); Eisenberg, 56 U. CHI. L. REV. at 77 ACADEMIC MEDICINE at 1362.1073. 168 See, e.g., Rubinfeld 2/25 at 19-20 (discussing 165 See, e.g., Teece 2/26 at 202-04, 210 “in-house bias”); Cohen 10/30 at 151-52; Shapiro 11/6 at(observing that unclear boundaries “foul up” the market for 164; McFalls 11/6 at 182-83.know-how, but concluding that solutions to licensing 169problems eventually emerge); see infra Ch. 5(I). See, e.g., Blackburn 2/26 at 264. 24
  • 95. Bo x 2-3 a. Patent Thickets. The potential eco nom ic pro blems assoc iated with paten t thickets are diverse. First, in a patent thicket where inno vation dep ends on having access to existing patents held by many different owners, the transaction costs of access can rise substantially because of the costs of negotiating with each of many individual patentees. See infra Ch. 2(III)(C)(1). Second, in some situations, the transaction costs of learning about and individually licensing all existing relevant patents are high enough to significantly undermine the economic incentive to develop follow-on innovation and production. In other situations, uncertainty surrounding pending p atents hampe rs reaching licen sing agreements. Unless a firm can m itigate the problem, it may have to choose between the risk of being sued for infringement after it sinks costs into its invention or production, or dropping its innovative or productive efforts altogethe r. See infra Ch.2(III)(C)(2). Third, a follow-on innovator in a patent thicket generally need s to acc ess multiple patentees’ intellectual prop erty to develo p his inve ntion. Follo wing C ournot’s prediction , each p atentee will dem and a higher royalty than its patent would command if it were licensed as part of a pa ckage. See infra Ch. 2(III)(C)(3). Finally, in patent thickets in which follow-on innovation depend s on having access to many p atents held by a gro up of oligop olists, the oligopolists may use the patents to prevent entry. See infra Ch. 2(III)(C)(4).would be beneficial, it is likely to occur.170 innovative work.173Prof. David Teece, for example, explainedthat although there may be “battles around Indeed, in some industries, there haspatents” in the early days of an industry, been a proliferation of patents.174“what tends to happen is that these problems Commentators have noted five factors thatget solved.”171 Anecdotal information and contribute to patent proliferation. First, thecase studies point in both directions.172 technology developed in industries such as semiconductors, computer hardware, andC. Follow-On Innovation in the software can contain a large number of Face of Multiple Existing Patents In some circumstances, the need for 173 See infra Ch. 3(III)(D)(4), (IV)(E)(2) andaccess to multiple existing patent-protected (V)(E)(2). For a summary of problems of the patent thicket and techniques for mitigating it, see Boxes 2-3a and 2-3b.technologies may also hinder subsequent 174 See Carl Shapiro, Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard- Setting, in 1 INNOVATION POLICY AND THE ECONOMY 119, 120 (Adam Jaffe et al. eds., 2001) (hereinafter Navigating 170 the Patent Thicket); Mowery 2/27 at 427; Stallman 4/9 at See, e.g., Teece 2/26 at 210-11; Kieff 4/10 at 20; Burk 3/20 at 149; Greenhall 2/27 at 375-76; Detkin159-60, 199-200; cf. Arora 2/25 at 88-89 (transaction costs 2/28 at 667-68 and Peter N. Detkin, A Semiconductorgenerally are not large enough to prevent licensing Patent Survey (2/28/02) (slides) at 5, atprovided that the proper incentives are present). http://www.ftc.gov/opp/intellect/020228peterndetkin.pdf. The introduction of patent maintenance payments may help 171 Teece 2/26 at 210. somewhat to clear patent thickets because a significant number of patent holders do not maintain their patents for 172 For example, compare O’Rourke 2/20 at 98 the full term. See Mark A. Lemley, Rational Ignorance atand Rai 4/10 at 19 (both citing Robert P. Merges & the Patent Office, 95 NW. U. L. REV. 1494, 1503-04 (2001)Richard R. Nelson, On the Complex Economics of Patent (50 percent of patents gone by twelfth year). Many of theScope, 90 COLUM. L. REV. 839 (1990), for its case studies patents so eliminated may never have had commercialof breakdowns in licensing opportunities) with Boulware significance, however. See Lemley, 95 Nw. U. L. REV. at10/30 at 175-76 (citing two broad, basic biotech patents 1503 (“Most of these [lapsed] patents aren’t litigated orthat have been widely licensed). licensed during the short time they are in force.”). 25
  • 96. incremental innovations.175 One panelist number of procedural and substantive rulesfrom the software industry noted that that collectively strengthened the rights ofprograms can contain millions of lines of US patent owners”179 – produced this shift.code and include “potentially hundreds of Hall and Ziedonis also identified two eventsthousands” of patentable inventions.176 The that arose out of the “pro-patent” shift andcomplex nature of such technology creates a signaled the strength of the new patenttechnology thicket over which a patent regime: (i) Polaroid’s patent infringementthicket develops.177 suit against Kodak, which resulted in almost $1 billion in damages and an injunction against Kodak’s participation in the instant- Bo x 2-3 b. Mitigating the Patent Thicket. film camera business,”180 and (ii) higher royalty rates obtained by Texas Instruments Techniques that companies use for handling the from an aggressive licensing strategy, which patent thicket include assuring mutual destruction, demonstrated to other firms the revenue see infra Ch. 2(III)(C)(2)(b), patent pooling, potential of mining a large patent cross-licensing, and p ackage licensing, see infra Ch. 2(III)(C )(3). portfolio.181 Third, in the semiconductor, computer hardware, and software industries, Second, in their research, Hall and defensive patenting strategies can driveZiedonis contend that a “pro-patent” shift in firms to patent even more. As more patentsthe U.S. legal environment in the 1980s was issue, the likelihood of “unintentional andthe stimulus for patent proliferation.178 The sometimes unavoidable patent infringement”authors believe that a series of congressional increases.182 Some firms respond to this byreforms in the early 1980s – including the “fil[ing] hundreds of patents each year”creation of the Court of Appeals for the themselves, patents they can use defensivelyFederal Circuit, which “put in place a against firms threatening infringement actions.183 The result of this, of course, is 175 See Detkin 2/28 at 669-70, 710-11; Poppen2/28 at 684, 712; Barr 2/28 at 713-14; Fox 2/28 at 714; 179Mowery 2/27 at 427; Armbrecht 3/19 at 54; Cohen 10/30 at Rosemarie Ham Ziedonis & Bronwyn H.91. Hall, The Effects of Strengthening Patent Rights on Firms Engaged in Cumulative Innovation: Insights from the 176 See Kohn 2/27 at 351-52; Pooley 2/27 at 382. Semiconductor Industry 12 (June 2001), at http://emlab.berkeley.edu/users/bhhall/papers/HallZiedonis 177 See Teece 2/27 at 500 (“the right question to 01%20libecap.pdf (draft version).ask is not whether or not theres a patent thicket, but 180whether or not the patent thicket, if there is one, is Hall & Ziedonis, 32 RAND J. OF ECON. atundergirded by a technology thicket”). 109. 181 178 Bronwyn H. Hall & Rosemarie Ham Id.Ziedonis, The Patent Paradox Revisited: An Empirical 182Study of Patenting in the U.S. Semiconductor Industry, Barr 2/28 at 677.1979-1995, 32 RAND J. OF ECON. 101, 105 (2001). See 183also Jeffrey T. Macher, David C. Mowery & David A. Id.; see also Detkin 2/28 at 668 (“there’s anHodges, Semiconductors, in U.S. INDUSTRY IN 2000: unavoidable overlap of IP . . . people are tripping over eachSTUDIES IN COMPETITIVE PERFORMANCE 279-81 (1999), at other’s patents right and left”); Hart 4/9 at 42-42; Hall 2/28http://www.nap.edu/books/0309061792/html/245.html. at 661; Hall & Ziedonis, 32 RAND J. OF ECON. at 125. 26
  • 97. yet more patenting. his business, he found 120 patents that appeared to overlap each other.189 The cost Fourth, in some industries, increased of access rises in such situations because ofpatenting levels may reflect increases in the costs of negotiating with each of manyR&D activity.184 Finally, the issuance of individual patentees.190unwarranted patents may be a contributingfactor to patent proliferation.185 One panelist b. Stemming from Lack ofcited interviews conducted with participants Benchmarksin the semiconductor industry in which theparticipants voiced concern regarding the Moreover, transaction costs may bepatenting of “very trivial inventions.”186 greater where bargainers lack benchmarks for the deal they are trying to reach. In1. High Transaction Costs general, incomplete or asymmetrical information in bilateral bargaining situationsa. Stemming From Number of Patents raises transaction costs by lengthening negotiations.191 Many licensing agreements When follow-on innovation depends are kept confidential, panelists noted,192 andon having access to patents held by many in any event, different transactions maydifferent owners, the transaction costs of involve unique elements that makeaccess can rise substantially. In industries comparisons difficult. As a result, when twowith incremental innovation, such as the parties wish to create a licensing agreement,software industry, innovation often dependson access to many patents.187 There can be“potentially dozens or hundreds of patents 189 See Greenhall 2/27 at 375-76.covering individual components of aproduct” in such an industry.188 One 190 Some have called this problem an aspect of the “tragedy of the anticommons.” See, e.g., Michaelpanelist’s experience illustrates the concern: Heller & Rebecca Eisenberg, Can Patents Deterin searching the patent landscape Innovation? The Anticommons in Biomedical Research,surrounding a particular patent relevant to 280 SCIENCE 698, 699 (1998); Hall 2/26 at 182-83; cf. Dickinson 2/6 at 61-62 (referring to this transaction cost concern as “patent layering”). For a description of the 184 tragedy of the anticommons and business panelists’ See, e.g., Ziedonis 3/20 at 13-14 and perceptions of whether it actually occurs, see infra Ch.Rosemarie Ziedonis, The Role of Patents in 3(III)(D)(4).Semiconductors: Insights from Two Recent Studies(3/20/02) (slides) at 3, at 191 See, e.g., RICHARD A. POSNER, ECONOMIChttp://www.ftc.gov/opp/intellect/020320rosemarieziedonis. ANALYSIS OF THE LAW 68-69 (5th ed. 1998) (using examplepdf; Telecky 2/28 at 711 (increasing patents reflect of litigation settlement to demonstrate the transaction costsincreasing research budgets); Mossinghoff 2/6 at 82-83 incurred by incomplete information, arguing that a(pharmaceutical R&D expenditures have increased at a potential litigant who does not know the price at which itsgreater rate than pharmaceutical patents). counterpart would prefer litigation to settlement will find it 185 expensive to determine the correct settlement terms, and See Ziedonis 3/20 at 15-16. will expend “much time and resources” trying to bargain once it has determined its counterpart’s price “range”); 186 See Ziedonis 3/20 at 15-16. Joseph Farrell, Information and the Coase Theorem, 1 J. ECON. PERSP. 113, 115 (1987) (sketching the difficulties 187 See Telecky (stmt) 3; Teece 2/27 at 500. that incomplete information raises for bilateral bargaining). 188 192 Mowery 2/27 at 427. See, e.g., Pooley 2/27 at 436-37. 27
  • 98. they may lack “a market-driven assessment concerns.of the value of the patent” in question,according to one panelist,193 and that In some situations, the transactionignorance can raise transaction costs. costs of learning about and individuallyIndeed, some would-be licensors feel the licensing all existing relevant patents areneed to threaten litigation in order to have high enough to undermine significantly theaccess to others’ confidential licensing economic incentive to develop follow-onterms,194 further heightening transaction innovation and production.195 For example,costs. one panelist noted that in industries such as semiconductors in which the ratio of patents2. Hold Up in the Patent Thicket to products is high, a firm cannot make a new product “without infringing hundreds ifa. Hold Up not thousands of patents.”196 Another commentator concurred: participants in the Sometimes, follow-on innovation semiconductor industry receive “thousandsand production depends on having access to of patents . . . each year and manufacturerspatents that are economically infeasible to can potentially infringe on hundreds oflicense because they are too numerous to patents with a single product.”197 Anotherlicense individually or even to learn about. panelist observed that “the large number ofIn other situations, uncertainty surrounding issued patents in [the computer hardwarepending patents hampers the reaching of industry] makes it virtually impossible tolicensing agreements. Unless downstream search all potentially relevant patents,actors – whether innovators or review the claims, and evaluate themanufacturers – can mitigate the problem, possibility of an infringement claim or thethey may have to choose between the risk of need for a license.”198being sued for infringement after they sinkcosts into invention or production, or In other situations, secrecydropping innovative or productive effortsaltogether. Either option can injure 195 High transaction costs can render licensingeconomic welfare. Below is a discussion of from multiple intellectual-property holders economicallythe economic theories behind these infeasible. See, e.g., Broadcast Music, Inc. v. Columbia Broadcasting System, Inc., 441 U.S. 1 (1979) (determining that blanket license warranted review under the rule of reason). Some have called transaction costs the problem of 193 Friedman 2/27 at 439 (noting that knowledge the “patent thicket.”of others’ licensing terms would be of limited use, since 196“markets with few people in it are extraordinarily inexact”); See Lemley 2/25 at 37-39 (noting that in suchsee also Pooley 2/27 at 437 (noting that if would-be industries, patents are awarded on “inventions [that] arelicensees could review confidential terms of other licensing small changes in process, they are small changes inagreements, they could get an objective sense of the worth product, they are circuit design innovations, they are littleof the patent by evaluating not just the royalty rate but who pieces of the innovation,” and that in such industries withthe licensees were, how much “they are actually paying high ratios of patents to products, “hold-up problems arewhen weighed against other contributions that they’re much greater than they are in other industries”).making or obligations they are taking,” and the like). 197 Shapiro, Navigating the Patent Thicket at 194 See, e.g., Pooley 2/27 at 437-38 (suggesting 125.that discovery in litigation could disclose such 198information). Barr (stmt) 1; see also Barr 2/28 at 676-7. 28
  • 99. surrounding a patent makes it very difficult rate. But after lock in, the downstream actorfor downstream actors to avoid it. Indeed, no longer has that option. Redesigning athe holder of a yet-unpublished patent can product after significant costs have been(once it issues) use it to hold up follow-on sunk may not be economically viable.203innovators and producers who unknowingly And the cost of being preliminarily enjoinedinfringed it.199 One panelist stated that “the is high: as one industry participant noted,long delays in the patent office work to losing a motion for a preliminary injunction[some firms’] benefit by keeping the in an infringement lawsuit “would beeventual coverage of their patents indefinite detrimental to a firm if it means shuttingwhile others produce products.”200 Some down a high-volume manufacturing facilitynoted that improving “information [since the] loss of one week’s production[available] at an earlier stage about patents alone can cost millions of dollars.”204likely to issue” could help ameliorate holdup,201 but hold up may persist because of Hold up can injure innovation anduncertainty about the scope of claims that competition. First, such a demand foreventually will issue.202 payment after lock in can compel the downstream actor to pay the patentee a “far If an innovator or producer learns greater” royalty rate.205 That higher rate,that it has infringed a patent only after it has one scholar noted, can be passed along tocommitted sunk costs to its innovation and consumers in the form of higher prices.206production – and thus locked in to the effort Second, the threat of hold up may reduce– the patentee may be in a position to overall levels of innovation, because somedemand supra-competitive royalty rates. If, companies will “refrain from introducingbefore lock in, the downstream actor had certain products for fear of holdup.”207known about the patent and could havedesigned its product or innovation around it,then the firm might have used theopportunity to adopt alternative designs as 203 See, e.g., Shapiro, Navigating the Patentleverage for seeking a competitive royalty Thicket at 125; Barr (stmt) 2-3; Rosemarie Ziedonis, When the Giants’ Shoulders are Crowded: Fragmented Rights and Patent Strategies in Semiconductors 8-9 (July 2002) 199 See Barr 2/28 at 676. (unpublished manuscript), at http://www.isnie.org/ISNIE02/Papers02/ziedonis.pdf. 200 Barr (stmt) 2; see also Ch. 4(II)(C)(1) 204(discussing continuations). Hall & Ziedonis, 32 RAND J. OF ECON. at 109 (paraphrasing the statement of an industry participant 201 See, e.g., Shapiro, Navigating the Patent whom they interviewed).Thicket at 126. For example, ninety percent of patent 205applications are now published within 18 months, pursuant Shapiro, Navigating the Patent Thicket atto the requirements of the American Invention Protection 125.Act; and a 1995 patent term change from 17 years after 206issuance to twenty years after filing may reduce incentives See id. at 126; see also Poppen 2/28 at 690.to prolong examinations. See infra Ch. 4(II)(C)(1). Somepanelists believed that these developments can mitigate 207 Shapiro, Navigating the Patent Thicket athold up; others pointed out that they would not completely 126; see also Peter C. Grindley & David J. Teece,cure the problem. Id. Managing Intellectual Capital: Licensing and Cross- Licensing in Semiconductors and Electronics, 39 CAL. 202 Barr 2/28 at 676; see also infra Ch. 4. MGMT. REV. 8, 20 (1997). 29
  • 100. b. Strategies to Mitigate ensures detente, and design freedom, for such firms.212 Each firm takes into account(i). Mitigation Strategy of Amassing that, if it tried to extract excessive royalties Patents and Assuring Mutual or impede the other’s innovation efforts Destruction through threats of patent infringement litigation, the other firm could retaliate by To mitigate such hold up in the suing it for patent infringement andcontext of a patent thicket, some firms in enjoining its production. This leads thecertain industries have accumulated large firms to reach licensing agreements withpatent portfolios.208 Panelists noted that a each other, often portfolio cross-licensingfirm with a large patent portfolio is in a agreements.213 Such agreements can givebetter position to raise patent infringement each firm the freedom to design and operatecounterclaims against a firm that tries to without fear of being sued by the other.214hold it up.209 It is also better able to forceothers to license their patents (or perhaps (ii). Costs and Limits of MAD Mitigationportfolios of patents),210 or to demand that Strategyother firms agree not to assert blockingpatents against it (often called “non- (A). Costly Arms Raceassertion agreements”).211 The prospect ofmutually assured destruction (or “MAD”) Amassing patent portfolios may mitigate hold up, but it also carries costs. It is, as one commentator noted, a “rather 208 See, e.g., Hall & Ziedonis, 32 RAND J. OF costly arms race.”215 It generates a “lot ofECON. at 104 (describing semiconductor industry);Ziedonis, When the Giants’ Shoulders are Crowded: resource waste,” some panelists noted,216Fragmented Rights and Patent Strategies in since firms spend “a significant amount onSemiconductors at 4 (describing semiconductor industry); legal bills to apply for patents” to use inBarton 2/26 at 150 (predicting evolution of mutual-assured-destruction strategy in financial services industryand biotech industry, and noting that such strategies are“not going to be an uncommon situation”). 209 See, e.g., COHEN ET AL., PROTECTING THEIR 212INTELLECTUAL ASSETS at 26-27 (stating that one of the See Hall 2/28 at 662; Hall & Ziedonis, 32most important uses of patents across all industries is to RAND J. OF ECON. at 109; Friedman 2/27 at 356prevent infringement lawsuits); Hall 2/28 at 662 (describing one goal of amassing large patent portfolios as(“Basically we pile up a lot of patents because the other maintaining detente). MAD strategies apply only to firmsguy has a lot of patents and that, when we, if we, do get that are vulnerable; those that are not are discussed below,threatened, we can engage in a cross-licensing see infra Ch. 2(III)(C)(2)(b)(ii)(B) (discussing undeterrednegotiation.”); League for Programming Freedom, Against NPEs).Software Patents (Public Comment) 6, at 213http://www.ftc.gov/os/comments/intelpropertycomments/lp See, e.g., Cohen 2/20 at 63-64.f.pdf. 214 See, e.g., id. See also Second Report 210 See, e.g., COHEN ET AL., PROTECTING THEIR (forthcoming) (discussing portfolio cross-licensingINTELLECTUAL ASSETS at 20-24 (stating that one reason agreements).firms in complex product industries obtain patents is to 215strengthen their position in cross-licensing negotiations). Cohen 2/20 at 33-34. 211 216 Non-assertion agreements are discussed in Hall 2/26 at 178-79 (reporting semiconductorSecond Report (forthcoming). patent executives’ views). 30
  • 101. these MAD strategies.217 One panelist but to assert against others.221 Since NPEsissued a directive to his company’s staff are not vulnerable to an infringementrequiring that they “reallocate roughly 20 to counter attack, MAD strategies threatening35 percent of [their] developers resources infringement actions do little to constrainand sign on two separate law firms to their willingness to seek high royalty ratesincrease [their] patent portfolio.”218 The from locked-in downstream actors.222 Thus,engineers’ time dedicated to assisting in the NPEs can threaten other firms with patentfiling of patents, which “have no . . . infringement actions, which, if successful,innovative value in and of themselves,” could inflict substantial losses, without fearcould have been spent on developing new of retaliation.223 In short, MAD strategies dotechnologies, the panelist noted.219 nothing to mitigate NPE hold up.224(B). NPEs Undeterred One panelist hypothesized that NPEs’ invulnerability may create a In addition to being expensive, MAD competitive problem if it prevents the typestrategies are not always effective. Firms of cross-licensing that has evolved as acannot use their patent portfolios defensively “safety valve” due to the prevalence ofagainst companies referred to as non- overlapping and cumulative patents.225practicing entities (NPEs).220 NPEs are Under this theory, a cross-licensing “safetyfirms that are, for a variety of reasons, valve” may be necessary for markets to workinvulnerable to a countersuit for patent efficiently when there are large numbers ofinfringement. They may be design firms overlapping and cumulative patents. If thethat patent their inventions but do not market-created safety valve relies on allpractice them or patent assertion firms thatbuy patents from other companies 221 See, e.g., Poppen 2/28 at 685-88; Detkin 2/28(particularly bankrupt ones) not to practice at 672. 222 See, e.g., Rhoden 2/28 at 723-24 (“There’s nothing that they need that you have and so they’re basically in the position where they have something perhaps that you need. Since there’s no mutually assured 217 Barton 2/26 at 177-78. destruction . . . they can come in and assert and shut your business down and you have no option against them.”); 218 Greenhall 2/27 at 376. McCurdy 3/20 at 72 (you cannot negotiate reasonable royalties from NPEs because “there is no counterassertion 219 Id. at 377, 420. capacity.”); cf. Ziedonis 3/20 at 71-72 (“The Lemelson Foundation, I think, has made a very successful business 220 Participants in the Hearings also used the from setting licensing fees so that balancing payment, youterms “non-vertically integrated” intellectual property set it low enough to where it’s below the cost of actuallyholders and “trolls” to refer to NPEs. For purposes of going to court or the managerial time that it would take toclarity, this Report uses the neutral term “NPE.” See basically fend off the lawsuit.”).Poppen 2/28 at 685-88; Detkin 2/28 at 672; Carl Shapiro, 223The FTC’s Challenge to Intel’s Cross-Licensing Practices, See Poppen 2/28 at 685-89; Detkin 2/28 atInstitute of Business and Economic Research Competition 671-72; Hall & Ziedonis, 32 RAND J. OF ECON. at 109.Policy Center Paper CPC02-029, at 7 (2002), at 224http://repositories.cdlib.org/cgi/viewcontent.cgi?article=10 See infra Ch. 3(IV)(E)(2)(c) for a description28&context=iber/cpc; see also infra Ch. 3(IV)(E) (noting of the recent rise in NPE activity.view that NPEs are merely exercising legitimate patent 225rights). Farrell 11/6 at 174-75. 31
  • 102. parties wishing to bring products to the prices of complementary products have onmarket, then a patent holder that is not each other’s sales, and would set a packagevulnerable to countersuit for infringement price that would maximize total profit.230may “gum[] up the safety valve.”226 Thus, if monopolistic producers of complementary products packaged their3. Royalty Stacking and Cournot’s products and extracted a single monopoly Complements Problem rent, prices would fall, output would increase, and profits would rise.231 In addition, the so-called“complements problem” can raise costs for The complements problem isinnovators who depend on access to multiple relevant to the problem of blocking patents,patents. First identified in 1838 by Antoine one panelist argued. A follow-on innovatorCournot and echoed by subsequent frequently needs to access multiple patentsobservers, the complements problem refers to develop his invention. When actingto the welfare loss stemming from individual alone, patent holders – like individualmonopolists each selling complementary monopolists of complementary technologygoods for a given use.227 Profit maximizing or information inputs – will demand higherbehavior will lead each producer to extract a aggregate royalties than they would if theymonopoly price for his good, resulting in acted as a group.232 Such behavior imposescumulative monopoly rents proportional to a financial burden on prospective licenseesthe number of complements.228 In contrast, that might deter further innovation.233if a single firm controlled the production ofall complementary inputs, it would extract a Indeed, some have argued that over-single monopoly rent, and the price would generous granting of patent rights in thebe lower than the aggregate of individual biomedical industry has resulted in follow-monopoly prices.229 This is because the firmwould take into account the effect that the 230 See Carl Shapiro, Theories of Oligopoly Behavior, in 1 HANDBOOK OF INDUSTRIAL ORGANIZATION 339 (Richard L. Schmalensee & Robert D. Willig eds. 226 Id. at 175. 1989) (noting that competing monopolists of complementary goods would not take “negative 227 See generally ANTOINE COURNOT, externalities” into their pricing decisions).RESEARCHES INTO THE MATHEMATICAL PRINCIPLES OF THE 231THEORY OF WEALTH (1838), tr. Nathaniel Bacon (1895); cf. See, e.g., Shapiro, Navigating the PatentShapiro, Navigating the Patent Thicket at 145 n.6 (noting Thicket at 123 (noting that “monopolist suppliers will findassumption that the complementary inputs are used in fixed it in their joint interests to offer a package price that is lessquantities and cannot substitute for each other). than the[] two components sold for when priced separately”); Nirvikar Singh & Xavier Vives, Price and 228 See Shapiro, Navigating the Patent Thicket at Competition in a Differentiated Duopoly, 15 RAND J. OF149 (applying economic theory to show that the aggregate ECON. 546, 547 (1984) (showing that rational firmsmonopoly “markup” of competing complement producers supplying complementary goods will cooperate to offer ais equal to the number of producers multiplied by the high enough quantity to “reinforce” one another’s market).markup for a single product). 232 See Shapiro, Navigating the Patent Thicket at 229 See, e.g., id. at 123 (observing that prices 123 (applying general Cournot theory of complements towould be lower and profits higher if a single producer blocking patents).controlled all complements than if each were controlled by 233individual monopolists). See id. at 124. 32
  • 103. on innovators’ under-utilization of existing of relevant intellectual property. 238 Anotherresearch. Biomedical researchers face a commentator has suggested that patentmaze of overlapping patents in the hands of pooling, cross-licensing, and packagedifferent owners, the critics state.234 They licensing can ameliorate the complementsargue that these conditions can raise to problem: when two or more patent holdersprohibitive levels the cost of licensing all of predict that other firms might wish licensesthe relevant patents for a useful advance.235 to their patents, they can organize patentPerversely, it can also divert research to pools or package licenses to facilitateunpromising areas that are relatively barren orderly transfer of intellectual property atof patents.236 One commentator states that lower combined royalty rates and higherthis situation provides an example of combined profits.239 Such mechanisms serveCournot’s complements problem: each a similar function to the impleading proposalbiotechnology patent holder, acting in its by allowing producers of complementaryown self-interest, holds royalty rates goods to set a mutually beneficial price.inefficiently high, raising the costs of further Such packages might also reduce theinnovation.237 transaction costs faced by prospective follow-on innovators.240 One panelist has suggested that an“impleading” mechanism could help cure 4. Oligopoly/Group Boycottthe Cournot problem. Under a systemsimilar to the one governing stakeholder When follow-on innovation dependslawsuits, a follow-on innovator could offer a on having access to many patents held by a“reasonable” royalty rate to all of the holders group of oligopolists, the oligopolists can use the patents to prevent entry. Specifically, some argued that patentees can foster “MAD oligopolies” that deter entry. They noted that a group of patentees, each 234 See Heller & Eisenberg, 280 SCIENCE at 698(discussing rising number of patents granted in biomedical fearing an infringement counterclaim fromresearch). The authors have deemed this an aspect of the the other, can tacitly agree not to sue each“tragedy of the anticommons.” See id. at 699. For a other for infringement.241 The patenteesdescription of the tragedy of the anticommons, see infraCh. 3(III). could “give each other at least a tacit license [or an] explicit license with some kind of 235 See Heller & Eisenberg, 280 SCIENCE at 698-99, 701 (describing the disincentive to follow-oninnovation created by overlapping patents). The authors 238also noted that individual patent holders tend to See Pooley 2/27 at 415-16 (also notingoverestimate the likelihood that their patent will be used in industry consortia or government intervention as potentialthe final invention, which induces some to charge a higher solutions to the problem).royalty rate than their patent deserves, raising the direct 239costs of licensing for follow-on innovations. See id. at 701. See Shapiro, Navigating the Patent Thicket at 123 (calling such solutions an “ideal outcome” under the 236 See id. at 699 (describing distortionary effects right circumstances).of the anticommons problem). 240 See supra Ch. 2(III)(C)(1)(a). 237 See Shapiro, Navigating the Patent Thicket at 241124 (linking tragedy of the anticommons to the See, e.g., Barton 2/26 at 151; see also Barton,complements problem). 65 ANTITRUST L.J. at 464. 33
  • 104. formal cross-license.”242 The group could the potential strategic uses of multipledeter entry either by refusing to license the patents. One branch of the discussionnew entrant or by charging the entrant high focused on an initial innovator’s efforts toroyalty rates.243 For example, one panelist accumulate patents to buttress a threatenednoted that “in Japan, . . . the leading firms in position of market power. Thus, an initialthe industry . . . agglomerate[d] huge innovator may seek to build a “fence”portfolios which they were swapping with around its position by securing additionaleach other, but which they were unwilling to patents on near substitutes, thereby blockingtrade with the outside players.”244 follow-on innovators from designing around the initial patent or raising their R&D On the other hand, what looks like a costs.248 Under a pure “fence” strategy, theMAD oligopoly may really be a pro- patentee would have no intention either tocompetitive way of rewarding those who license the substitute patent technologies ortook the initial risks, some observed,245 or of to develop them on its own; the only goalcutting through patent thickets.246 Moreover, would be to keep rivals out.249 Somecross-licensing arrangements in the analysts suggest that preemptive patenting ofsemiconductor industry appear not to have this type is likely to be a useful strategy onlyslowed innovation, one panelist argued.247 in exceptional circumstances, given the costs and the potential for multiple routes to5. Patent Fences and Patent entry.250 Nonetheless, some recent survey Extensions evidence suggests that “fence” strategies may be frequently employed in “discrete Hearing discussion raised some of products” industries – which entail relatively few patents per commercial product – in which individual patents fail to prevent 242 Barton 2/26 at 151. 243 Id. at 152 (noting that patentee group couldcharge outsiders royalty rates “that were not simply enoughto cover a reasonable share of the research costs and soforth, but [were] so big as to knock everybody else out ofthe industry”). 248 See Cohen 2/20 at 31; COHEN ET AL., PROTECTING THEIR INTELLECTUAL ASSETS at 22, 25; 244 Ordover 11/6 at 105. Richard J. Gilbert, Patents, Sleeping Patents and Entry Deterrence, printed in STRATEGY , PREDATION, AND 245 See, e.g., Teece 2/26 at 176-77 (arguing that ANTITRUST ANALYSIS 223-25 (S. Salop ed. 1981);often, patentees cross-license “as a way to extract a fee. So SCHERER, INDUSTRIAL MARKET STRUCTURE ANDthe latecomers who didnt . . . incur a lot of those early ECONOMIC PERFORMANCE at 451 (1980).expenses end up . . . having to pay something, and youseem to me that youve solved the classic sort of free-rider 249 See Cohen 2/20 at 32; COHEN ET AL.,problem”). PROTECTING THEIR INTELLECTUAL ASSETS at 22, 25. 246 See, e.g., Barton 2/26 at 152 (noting that 250 See Gilbert & Newbery, 72 AMER. ECON.some cross-licensing agreements are “appropriate because REV. at 514-15; Gilbert, Patents, Sleeping Patents andwe have zillions of mutually-blocking patents”); Shapiro Entry Deterrence at 227 (observing that if there are many11/6 at 111. See generally supra Ch. 2(III)(C)(2)(b) patentable alternatives at comparable development costs,(discussing MAD strategies). then “the use of preemptive patenting to fence in a monopoly is about as effective as holding back a flood with 247 See Teece 2/26 at 177. a sieve”), 268-69. 34
  • 105. imitation or substitution.251 patents on improvements. During the Hearings, Professor F.M. Scherer argued Panelists did not suggest that patent that Xerox’s strategy for photocopyingfences developed by a firm’s own research illustrated this approach, stating that “byare, or should be, antitrust violations.252 amassing this continuing portfolio ofSome scholarship, however, raises concerns improvement patents, Xerox was going toregarding the effects of patent fences on monopolize the industry, not for 17 years,follow-on innovation and efficiency. Some but forever.”255 For the strategy to becommentators suggest that using patents to effective however, there must be somebuild fences departs from traditional patent reason to expect that, following expirationgoals; rather than securing a defined reward of the initial patent, competitors offering thefor beneficial innovation, it expands that no-longer-patent-protected core productinitial reward by broadening the zone of would not adequately constrain pricing ofexclusivity – and the possible impact on the improved version.256entry and independent follow-on innovation– without conferring additional social 6. Patent Floodingbenefits through new products orprocesses.253 Other analysts have contended Efforts to build a sufficient patentthat socially wasteful expenditures of portfolio to induce others to share theirresources flow from fencing activities.254 technology through cross licenses may shade into more aggressive strategies. When rivals A related strategy, which might be obtain patents on trivial variants of an initialdesignated “patent extensions,” involves innovation, “patent flooding” becomesefforts to extend patent protection beyond possible. Under this strategy, “[t]he flooderthe life of an initial patent by accumulating ‘surrounds’ a competitor’s patent or technology . . . so that over time, the competitor finds itself ‘unable to 251 See Cohen 2/20 at 32 and Cohen Presentation maneuver.’”257 Lacking the breathing roomat 14 and 15; COHEN ET AL., PROTECTING THEIRINTELLECTUAL ASSETS at 22-25 (treating as indicative of afence strategy survey responses that reported blocking, but 255 Scherer 7/10 at 180. Professor Scherer’snot negotiating or licensing, as among the motives for textbook cites a similar prolongation of control through apatenting and finding such responses 44-45% of the time in series of improvement patents on the electric light bulb.discrete product industries). See SCHERER, INDUSTRIAL MARKET STRUCTURE AND ECONOMIC PERFORMANCE at 451-52 (1980). 252 Building a fence through acquisitions ofpatents, however, could raise issues under Section 7 of the 256 See Sheila F. Anthony, Riddles and LessonsClayton Act, 15 U.S.C. § 18. See, e.g., McFalls 11/6 at from the Prescription Drug Wars: Antitrust Implications of183-84. Certain Types of Agreements Involving Intellectual Property, Before the American Bar Association Program on 253 See COHEN ET AL., PROTECTING THEIR “Antitrust and Intellectual Property: The Crossroads”INTELLECTUAL ASSETS at 28; cf. SCHERER, INDUSTRIAL (June 1, 2000) (analyzing analogous issues raised by EliMARKET STRUCTURE AND ECONOMIC PERFORMANCE at Lilly’s acquisition of an exclusive license from Sepracor),451-52 (1980) (patent fencing a way “to extend and at http://www.ftc.gov/speeches/anthony/sfip000601.htm.pyramid . . . monopoly power”). 257 Sri Krishna Sankaran, Patent Flooding in the 254 See HOPENHAYN & MITCHELL, INNOVATION United States and Japan, 40 IDEA 393, 394 (2000),FERTILITY & PATENT DESIGN at 4. quoting Dan Rosen & Chikako Usui, The Social Structure 35
  • 106. to develop improvements or to find new Conclusion. Competition policy anduses for its invention, the initial innovator patent policy enhance economic welfare ineventually must accede to demands that it complementary ways. Yet neithershare its technology through a cross license competition nor patent policy can, alone,with the flooder.258 Critics of patent promote innovation fully. Competitionflooding argue that these cross licenses are alone is not a perfect tool for fosteringone-sided, extracting valuable intellectual innovation. For example, the award ofproperty from the targets and undermining patents is often necessary to remedy freeinitial innovators’ incentives to innovate riding on others’ innovations. But patentwithout contributing significant follow-on policy alone also is not a perfect tool forbenefits.259 Typically they point to examples fostering innovation. Indeed, patent rightsin Japan, where the patent system appears can in some circumstances hinder follow-onmore conducive to flooding strategies than innovation and competition. Rather, the twoin the United States.260 The hearing record means of promoting innovation must workdoes not suggest that patent flooding is in tandem with each other.currently a widespread practice in thiscountry.261 The balance of this Report explores how they can best do so. Chapter 3 provides extensive real-world illustrations of theof Japanese Intellectual Property Law, 13 UCLA PAC. economic phenomena as voiced by businessBASIN L. J. 32, 44 (1994); Jeffrey A. Wolfson, Patent representatives from the pharmaceutical,Flooding in the Japanese Patent Office: Methods forReducing Patent Flooding and Obtaining Effective Patent biotechnology, computer hardware, andProtection, 27 GEO. WASH. J. INT’L L. & ECON. 531 (1994). software/Internet industries. Chapters 4 and 5 translate the core economic concerns into a 258 See, e.g., Wolfson, 27 GEO. WASH. J. INT’L L. detailed examination of patent system& ECON. at 533 (describing cross-licensing as the flooder’stypical goal). standards and procedures. 259 See, e.g., id. at 533, 554-55. 260 See, e.g., Sankaran, 40 IDEA at 399-404(emphasizing that patent applicants in Japan may deferexamination for up to seven years, allowing them to assertthe claimed rights coercively for a prolonged periodwithout ever having to demonstrate patentability). Severalanalysts note the perception of a proclivity in Japan to issuenarrow patents to initial innovators and to grant patents onrelatively minor variations on prior inventions; they arguethat this would increase the potential for flooding. See,e.g., John Gladstone Mills III, A Transnational Patent instances in which it has been alleged. See Sankaran, 40Convention for the Acquisition and Enforcement of IDEA at 411-17. As discussed below, the patent law’sInternational Rights, 84 J. PAT. & TRADEMARK OFFICE obviousness doctrine, which deals with the size ofSOC’Y 83, 110-11 (2002); Sankaran, 40 IDEA at 395; inventive step necessary for obtaining a follow-on patent,Wolfson, 27 GEO. WASH. J. L & ECON. at 539-41; Ordover, affects opportunities to employ flooding strategies. See5 J. ECON. PERSP. at 48. infra Ch. 4(II)(A)(1). In the United States, some aspects of this doctrine work against flooding. Cf. Merges 2/26 at 261 See Kunin 7/11 at 181-83 (describing patent 162-64 (explaining that the double patenting doctrine in theflooding as a product of the Japanese patent system). Even United States gives initial innovators greater ability thanan analyst who argues that patent flooding may have improvers to patent what would otherwise be obviousoccurred in the United States identifies at most a handful of variations). 36
  • 107. CHAPTER 3 BUSINESS TESTIMONY: CURRENT INNOVATION LANDSCAPE IN SELECTED INDUSTRIESI. SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1II. THE PHARMACEUTICAL INDUSTRY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 B. Industry Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1. Discrete Research and Development for NCEs . . . . . . . . . . . . . . . . . . . . . 4 2. The Demanding Nature of the NCE Development Process . . . . . . . . . . . . 6 3. The Implications of Clinical Trials for Effective Patent Term of NCEs . . 7 4. Incremental Innovation for the Development of IMDs . . . . . . . . . . . . . . . 8 C. The Role of Patents In Spurring Pharmaceutical Innovation . . . . . . . . . . . . . . . . . 9 D. The Role of Competition in Spurring Pharmaceutical Innovation . . . . . . . . . . . 10 E. The FTC’s Pharmaceutical Industry Enforcement Actions and Generic Drug Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 F. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14III. THE BIOTECHNOLOGY INDUSTRY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 B. Industry Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 C. The Role of Competition in Spurring Biotechnology Innovation . . . . . . . . . . . . 17 D. The Implications of Patent Protection for Innovation . . . . . . . . . . . . . . . . . . . . . 17 1. The Role of Patents in Spurring Innovation in the Biotechnology Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 a. Patentability Encourages Investment in R&D . . . . . . . . . . . . . . . 17 b. The Role of Patent Disclosures in Fostering Biotechnology Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 c. Patenting of Biotechnology Research Tools . . . . . . . . . . . . . . . . 18 2. The Quality of Biotechnology Patents . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3. The Mechanisms Available for Challenging Questionable Patents . . . . 21 a. Challenging Questionable Patents Through Litigation . . . . . . . . 21 b. Challenging Questionable Patents Through Reexamination Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
  • 108. c. Challenging Questionable Patents Through a New Opposition System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4. The Potential for Patents to Impede Innovation in the Biotechnology Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 a. The Development of an Anticommons . . . . . . . . . . . . . . . . . . . . 24 b. Access to Existing Technologies Needed for Follow-On Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 E. Licensing Practices for Biotechnology Research Tools . . . . . . . . . . . . . . . . . . . 26 1. Reach-Through License Agreements . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2. Patent Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3. Non-Exclusive Licensing of Patented Research Tools . . . . . . . . . . . . . . 28 F. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29IV. THE COMPUTER HARDWARE INDUSTRIES, INCLUDING SEMICONDUCTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 B. Industry Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 C. The Role of Competition in Spurring Computer Hardware Innovation . . . . . . . 31 D. Alternative Means of Fostering Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 E. The Implications of Patent Protection for Innovation . . . . . . . . . . . . . . . . . . . . . 33 1. The Role of Patents in Spurring Innovation . . . . . . . . . . . . . . . . . . . . . . . 33 2. The Potential for Patents to Impede Innovation . . . . . . . . . . . . . . . . . . . . 34 a. Patent Thickets in the Computer Hardware Industries . . . . . . . . 34 b. The Potential for Patent Thickets to Harm Innovation . . . . . . . . 36 c. The Strategic Use of Patents in Licensing Negotiations . . . . . . . 37 F. Tools to Navigate the Patent Thicket . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 1. Cross-Licensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2. Patent Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 3. Standard-Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 G. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43V. THE SOFTWARE AND INTERNET INDUSTRIES . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 B. Industry Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
  • 109. C. The Role of Competition in Spurring Software and Internet Innovation . . . . . . . 46D. Alternative Means of Fostering Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 1. Copyright . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 2. Open Source Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47E. The Implications of Patent Protection for Innovation . . . . . . . . . . . . . . . . . . . . . 48 1. The Role of Patents in Spurring Innovation in the Software and Internet Industries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 a. The Role of Patents in Preventing Free Riding and Encouraging Investment in Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 b. The Role of Patents in Fostering Innovation Through Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 c. The Role of Patents in Fostering Design-Around Innovation . . . 50 2. The Potential for Patents to Impede Innovation in the Software and Internet Industries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 a. Patents May Impede Independent Follow-On Innovation . . . . . . 50 b. Patents May Increase the Costs of Entry . . . . . . . . . . . . . . . . . . . 51 c. Avoiding Patent Infringement Is Costly and Uncertain . . . . . . . . 51 d. Questionable Patents Create Uncertainty and Hinder Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53F. Licensing Strategies to Navigate the Patent Thicket . . . . . . . . . . . . . . . . . . . . . . 55G. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
  • 110. CHAPTER 3 BUSINESS TESTIMONY: CURRENT INNOVATION LANDSCAPE IN SELECTED INDUSTRIESI. SUMMARY role of competition and patents in spurring or discouraging innovation in their Over six days of Hearings, business industries.representatives from four high-techindustries discussed the drivers of Pharmaceutical and biotechnologyinnovation in their industries. representatives testified that strong patentRepresentatives from the pharmaceutical, protection is essential to innovation in theirbiotechnology, Internet, and computer industries. Business representativeshardware and software industries described characterized innovation in these industriestheir real-world experience with how patents as costly and unpredictable, requiringand competition affect incentives to significant amounts of pioneering researchinnovate. Their discussions confirmed many to discover and test new drug products. Byof the principles summarized in Chapter 2 preventing rival firms from free riding onand sometimes shed additional light and discoveries, patents allow pharmaceuticaloffered new perspectives on the topics. firms to recoup the substantial capitalThey highlighted both the benefits and costs investments made to discover, test, andof current patent and antitrust policies obtain regulatory approval of new drugapplied in their industries. This chapter products. Biotech representativesdiscusses the diverse views presented by the emphasized that patent protection is criticalpanelists, and also incorporates the results of to attract the capital necessary to fund thisbusiness surveys and other industry-specific high-risk investment. Indeed, firms believedscholarship. that the biotech industry would not exist but for patents. One concern involved patents The panelists identified various on the research tools used to assist in theattributes that characterized innovation in discovery of new drug products. Biotechthe different industries. Panelists discussed representatives expressed concern that suchwhether innovation in their industries tends patents could obstruct the commercializationto be discrete or cumulative, building of new products, thereby hindering follow-incrementally on prior discoveries. Panelists on innovation. To date, however, evidencealso addressed sources and amounts of suggests that such problems have notcapital required for entry, barriers to entry, emerged.the extent to which industries are verticallyintegrated, and difficulties in Pharmaceutical and biotechcommercializing new products. They raised representatives testified that they use patentissues of fixed cost recovery, alternative information disclosures required by theappropriability mechanisms, and patent statutes to direct their research andrelationships between initial and follow-on development (R&D) into areas not claimedinnovation, adding business insights and by the patents. Representatives from genericpractical experience to the analysis of pharmaceutical firms discussed how patentChapter 2. According to both panelists and disclosures guide their efforts to “design-academics, factors such as these shape the around” patents, so that they can develop
  • 111. non-infringing generic versions of brand- rapidly increasing rates. They explained thatname drug products. the increased likelihood of firms holding overlapping intellectual property rights By contrast, computer hardware and creates a “patent thicket” that they mustsoftware industry representatives generally clear away to commercialize newemphasized competition to develop more technology. They discussed how patentadvanced technologies as a driver of thickets divert funds away from R&D, makeinnovation in these rapidly changing it difficult to commercialize new products,industries. These representatives, and raise uncertainty and investment risks.particularly those from the software Some computer hardware and softwareindustry, described an innovation process representatives highlighted their growingthat is generally significantly less costly than concern that companies operating in a patentin the pharmaceutical and biotech industries, thicket are increasingly vulnerable to threatsand they spoke of a product life cycle that is to enjoin their production from non-generally much shorter. Some software practicing entities that hold patentsrepresentatives observed that copyrights or necessary to make the manufacturer’sopen source code policies facilitate the product.incremental and dynamic nature of softwareinnovation. They discounted the value of A global concern that representativespatent disclosures, because they do not from each of the four industries describedrequire the disclosure of a software was that poor patent quality (e.g., a patentproduct’s underlying source code. for which there is invalidating prior art, or a patent broader than was enabled) can blunt Computer hardware manufacturers incentives to innovate. They described thenoted that they often use trade secrets, rather costly nature of litigation to invalidate thesethan patents, to protect their inventions, patents, both in terms of dollars andbecause it is difficult to discover whether a resources diverted from R&D. They alsorival firm has infringed a patented discussed how a timely, less costlymanufacturing invention. Computer mechanism to review poor quality patentshardware manufacturers generally would would enhance innovation in theirrather keep the invention secret than publicly industries.disclose it and risk third partymisappropriation of patent rights that they These representatives also describedwill be unable to discover. By contrast, how each industry has developed licensingcomputer hardware firms that specialize practices to extract value from their patentssolely in hardware design and have no or, in some cases, to obviate some of themanufacturing responsibilities valued patent problems raised by patent thickets. Theyprotection as a way to raise venture capital. raised concerns that uncertainty about the parameters of antitrust enforcement may be Representatives from both the hindering the use of certain methods tocomputer hardware and software industries extract patent value. For example, biotechobserved that firms in their industries areobtaining patents for defensive purposes at 2
  • 112. representatives noted Bo x 3-1 . Independent Inventors and the FTC’s Invention Promotion Casesthat antitrust concernshave contributed to One cross-industry concern raised by a specific sub-group was theuncertainty about the vulnerability of ind ependent inventors to fraudulent prac tices as they seek p atentspropriety of using and offer licenses on those p atents. T his pro blem has be en, and continues to b e, a matter of FTC concern. Two panelists representing the independent inventionreach-through royalty com munity m entioned the defrauding of inventors by invention pro motio n firms.provisions in research See Ud ell 2/28 at 568-69 (“the FT C has don e a ma gnificent job o f not onlytool licenses. educating inventors, but also getting the scam organizations that have been bleed ing inventors for decades out of the pockets o f the poor inve ntors in America.”); Hayes-Rines 3/19 at 61-62 (urging enhanced FTC enforcement Firms in the efforts).computer hardware andsoftware industries In 1997, the FTC launched a consumer education program and a law-indicated that antitrust enforcement sweep en titled “Project M ousetrap” beca use a “numb er of firms inconcerns may be the invention pro motion industry are perpetrating a massive fraud” against indepen dent inventors. As a result of this sweep and o ther enforcem ent actions,inhibiting joint the Commission brought eight cases against invention promoters during thediscussions of licensing 1990s. The complaints have named 41 defendants, consisting of 21 companiesterms during the and 20 ind ividuals. In some cases, the Co mmission alleged that the defendantsstandard-setting represented that they would obtain patents for their customers’ inventions withoutprocess. They noted clarifying that these would be d esign patents, which typically have less commercial value than utility patents. The Commission generally alleged that thethat antitrust has defendants repre sented that their research and marketing services were likely totraditionally been secure profitable licenses for their customers’ inventions. The Commissionsuspicious of joint further alleged that, in fact, the defendants were rarely successful at securingdiscussions of licensing licensing agreements, and that the few licenses that the defendants did secureterms arising prior to seldom resulted in appreciable income for the inventors.the adoption of a In six cases, the Commission obtained consent orders that required thestandard. Some defendants to pay consumer redress and to make affirmative disclosures topanelists suggested, prospective custo mers about the promo ters’ past success rates. O ne case is still inhowever, that such litigation and the eighth case was dismissed after the U.S. Attorney’s office filedconduct is necessary for criminal charges. More recently, the Commission has expanded its consumer education program, in coo peration with the PT O, to include rights av ailable tothe efficient invento rs under the A merican Inventors Pro tection Act of 199 9. Further de tailsestablishment of new on the Commission’s consumer education efforts and enforcement actions arestandards because some ava ilable a t http://www.ftc.gov/bcp/conline/edcams/invention/ andcompanies are using http://www.ftc.gov/opa/1997/07/mouse.htm.patents strategically. 3
  • 113. II. THE PHARMACEUTICAL B. Industry Description INDUSTRY R&D in the pharmaceutical industryA. Introduction generally produces two main types of innovation: (1) discrete innovation, which Representatives from the means, in general terms, that the inventionpharmaceutical industry stated that patent might be improved, but does not point theprotection is indispensable in promoting way to wide-ranging, subsequent discoveriespharmaceutical innovation for drug products of new chemical entities (NCEs);1 and (2)containing new chemical entities. The sunk incremental innovation, which describes thecost of engaging in research projects aimed development of improvements to existingtoward the development of these drugs is drug products, often referred to as productextremely high. By preventing rival firms line-extensions.2 Obviously, innovation canfrom free riding on the innovating firms’ occur at many points along the continuum,discoveries, patents can enable from discrete to incremental, but thesepharmaceutical firms to cover their fixed categories are useful in identifying certaincosts and regain the capital they invest in characteristics associated with innovation inR&D efforts. Moreover, the patenting the pharmaceutical industry.process requires disclosure of the underlyinginvention covered by the patent, potentially 1. Discrete Research andencouraging further innovation. Generic Development for NCEsdrug companies report they use disclosedpatents as a basis on which to “invent- Discrete R&D in the pharmaceuticalaround” patented, brand-name products in industry focuses on the discovery andorder to develop generic variations. development of new chemical or molecular The panelists who representedpharmaceutical firms or organizations at theHearings were Robert A. Armitage,representing Eli Lilly and Company; Monte 1 See Robert P. Merges & Richard R. Nelson, OnR. Browder, representing Ivax Corporation; the Complex Economics of Patent Scope, 90 COLUM. L.David Coffin-Beach, representing REV. 839, 880 (1990) (discussing types of innovation);Torpharm, Inc.; Gregory J. Glover, Counsel FTC/DOJ Hearings on Competition and Intellectual Property Law and Policy in the Knowledge-Basedto Pharmaceutical Research and Economy, Mark Lemley Testimony Feb. 25, 2002, at pageManufacturers of America; Nancy J. Linck, 37 (hereinafter, citations to transcripts of these Hearingsrepresenting Guilford Pharmaceuticals; and state the speaker’s last name, the date of testimony, and relevant page(s)); Richard C. Levin, Testimony of RichardRoss Oehler, representing Aventis C. Levin, President, Yale University (2/6/02), atPharmaceuticals Inc. One scholar, Edward http://www.ftc.gov/os/comments/intelpropertycomments/leA. Snyder, from the University of Chicago, vinrichardc.htm (hereinafter R. Levin (stmt)). But cf. Browder 3/19 at 174 (noting the potential need forand one attorney, Rochelle K. Seide, from progression from generic compound to specific compoundBaker Botts, LLP, also participated in a to unique formulation).business perspective panel on the 2pharmaceutical industry. For an overview of the different types of pharmaceutical patents, see Box 3-2. 4
  • 114. entities to make small molecule drug Box 3-2 . Pharm aceutical Patentsproducts.3 The discovery of a chemicalmolecule that is both efficacious and safe for Pharmaceutical patents are issued for fourhuman usage can result in a totally new drug different categories: drug substance, method of use,product. Such discoveries typically require formulation, and process. Dru g substance patents cover the compound or active ingredient in the drugsignificant amounts of pioneering research, product, such as fluoxetine hydrochloride, which is theand both fixed costs and risks of failing to active ingredient in Prozac. M ethod of use p atentsdevelop a marketable product, consequently, cover the use of the p roduct to treat certain healthare very high. Brand-name companies spend problems, such as depression or asthma. Formulationa substantial amount in development costs patents cover the physical composition or delivery mechanism of the drug product, such as an extendedover the course of 10 to 15 years to bring a release tablet or capsule. Process patents gene rallyproduct involving an NCE to market from cover the procedure used to make the activethe initial research stage.4 The brand-name ingredient. For further details on pharmaceuticalcompanies’ trade association reports that patents, see Federal T rade Commission, Generic Drugmost newly marketed drugs do not cover Entry Prior to Patent Expiration: An FTC Study (July 200 2) (hereinafter, FTC, Generic Drug Study), attheir average development costs.5 Brand- http://www.ftc.gov/os/2002/07/genericdrugstudy.pdf .name companies typically rely on a smallnumber of “blockbuster” drugs to recoup 3 their overall investment in innovation, This contrasts with the biotechnology industry,which focuses instead on cells and large biological including R&D costs for failed products.6molecules (such as DNA and proteins). See Beier 2/26 at248. Relatively few patents are required to 4 protect a product with an NCE.7 One See Gregory J. Glover, Competition in thePharmaceutical Marketplace (3/19/02) 3 (stating that the panelist noted that an actual drug productaverage cost to develop a new drug is $802 million), at can be based on between four and 15http://www.ftc.gov/opp/intellect/020319gregoryjglover.pdf patents.8 The low number of patents(hereinafter Glover (stmt)); Armitage 3/19 at 127-28; seealso Bureau of Economics Staff Report, Federal Trade contained in a pharmaceutical productCommission, The Pharmaceutical Industry: A Discussionof Competitive and Antitrust Issues in an Environment ofChange (Mar. 1999) (discussing development risk), 6available at See The National Institute for Health Carehttp://www.ftc.gov/reports/pharmaceutical/drugrep.pdf Management, Changing Patterns Of Pharmaceutical(hereinafter BE Staff Report, The Pharmaceutical Innovation 4 (2002), atIndustry); Arthur D. Little, Examining the Relationship http://www.nihcm.org/innovations.pdf (hereinafter NIHCMBetween Market-Based Pricing and Bio-Pharmaceutical Innovation Report); IMS Health, IMS HEALTH DataInnovation (Public Comment) 3, at Reveal Dramatic Growth in Megabrands, athttp://www.ftc.gov/os/comments/intelpropertycomments/lit http://secure.imshealth.com/public/structure/dispcontent/1,tlearthurd2.pdf). 2779,1362-1362-143992,00.html; PhRMA (stmt) 11. 7 5 See Pharmaceutical Research and One panelist defined discrete product industriesManufacturers of America, Delivering on the Promise of as those that require relatively few patents to protect aPharmaceutical Innovation: The Need to Maintain Strong product, and complex product industries as those thatand Predictable Intellectual Property Rights (Public require a relatively large number. See Cohen 2/20 at 30Comment) 9, at and Wesley M. Cohen, Patents: Their Effectiveness andhttp://www.ftc.gov/os/comments/intelpropertycomments/ph Role (2/20/02) (slides) at 13, atrma020422.pdf (hereinafter PhRMA (stmt)); see also http://www.ftc.gov/opp/intellect/cohen.pdf.Glover (stmt) 4; Armitage 3/19 at 129; BE Staff Report, 8The Pharmaceutical Industry (discussing market risk). See Browder 3/19 at 174. 5
  • 115. means that, as panelists noted, the and expensive.13 The first stage involves thedevelopment of patent thickets is generally identification of chemical compounds thatnot a concern.9 Although brand-name might treat an indication or disease.14 Incompanies may compete with each other in general, the brand-name companies’ tradethe same therapeutic class, such as anti- association reported, “only 20 in 5,000depressants or blood-pressure-lowering compounds that are screened enterdrugs, and may seek to obtain a number of preclinical testing,” which involvespatents in a particular area to ensure freedom laboratory and animal testing.15to operate, such behavior has not given riseto so many overlapping sets of patent rights The second stage begins when theas to hinder the commercialization of new company sponsoring the drug submits antechnologies.10 From 1989 to 2000, the NDA to the FDA. Three phases of clinicalFood and Drug Administration (FDA) testing then follow, which the drug-approved 1,035 New Drug Applications sponsoring company undertakes and the(NDAs), 361 of which were for NCEs.11 FDA’s Center for Drug Evaluation andThe remaining 674 NDAs that FDA Research oversees. Brand-name companiesapproved during this period were conduct Phase I clinical studies on healthyincrementally modified drugs (IMDs).12 human beings to determine side effects and gather preliminary evidence of effectiveness.2. The Demanding Nature of the Phase II studies “are designed to obtain data NCE Development Process on the effectiveness of the drug for a particular indication or indications in Panelists provided an overview of patients with the disease or condition.”16the two-stage process to determine whether Phase III studies are expanded controlledan NCE is safe and efficacious to market – a and uncontrolled trials and can involveprocess that is time-consuming, uncertain, thousands of patients. These clinical trials are often very resource and time-intensive.17 9 See Glover (stmt) 8. A patent thicket is a “dense web of overlappingintellectual property rights that a company must hack itsway through in order to actually commercialize newtechnology.” Carl Shapiro, Navigating the Patent Thicket: 13Cross Licenses, Patent Pools, and Standard-Setting, in 1 See Armitage 3/19 at 127-28.INNOVATION POLICY AND THE ECONOMY 119, 120 (Adam 14Jaffe et al. eds., 2001) (hereinafter Navigating the Patent See id. “Indication” means disease, illness, orThicket). disorder. 10 15 See Glover (stmt) 6, 8; Armitage 3/19 at 230. See Glover (stmt) 3; Armitage 3/19 at 127. 11 16 See NIHCM Innovation Report at 3. Tufts Center for the Study of Drug Development, How New Drugs Move through the 12 See NIHCM Innovation Report at 3. IMDs are Development and Approval Process (2001), atdrugs which rely on an active ingredient present in a drug http://csdd.tufts.edu/NewsEvents/RecentNews.asp?newsid=already approved for the U.S. market, or a closely related 4.chemical derivative of such an ingredient, that has been 17modified by the manufacturer. Id. at 5. See Glover (stmt) 3. 6
  • 116. 3. The Implications of Clinical Trials years from the date of FDA approval.22 for Effective Patent Term of NCEs Pharmaceutical companies report, however, that by the time clinical trials are complete The time-consuming nature of and a drug product is ready to market, theclinical trials to evaluate a drug product’s effective patent life for a drug patent – evensafety and efficacy may limit the length of with patent term restoration – is typicallyeffective patent term that brand-name about 11 years,23 substantially shorter thancompanies can realize. Panelists testified the 20-year statutory patent term.24that brand-name companies seek to obtain Congress also has provided other marketpatents early in the R&D process – usually exclusivity periods for brand-namebefore clinical trials have commenced.18One panelist stated that the initial patent(s)to be issued for a totally new drug productare on the drug substance (i.e., the NCE ormolecule).19 This panelist contended thatdrug substance patents are typically the most 22 35 U.S.C. § 156 (c)(3). Another approach tovaluable for the brand-name company, restoring the patent term that elapses during FDA review would be to reduce FDA approval time. One study hasbecause they are much more difficult for found that reductions in regulatory approval times arepotential competitors (including generic somewhat more effective in increasing cash flow for acompanies) to design around than brand-name company, because such reductions add years to the less heavily discounted beginning of the product lifeformulation or method of use patents.20 cycle, rather than the end. See James W. Hughes et al., “Napsterizing” Pharmaceuticals: Access, Innovation, and In the Hatch-Waxman Amendments Consumer Welfare (Public Comment) 8-9, at http://www.ftc.gov/os/comments/intelpropertycomments/snto the Federal Food, Drug and Cosmetic Act, ydermoorehughes.pdf.Congress provided for restoration of aportion of the patent term that elapses while 23 See PhRMA (stmt) 9-10 (stating that “theclinical trials and FDA review are under [average] effective patent life for drugs introduced from 1984-1995 that received patent term restoration, includingway.21 The Hatch-Waxman Amendments such restoration, was only about 11 years” and citing Sheilacan restore patent term up to a maximum of R. Shulman et al., Patent Term Restoration The Impact offive years, depending on how long clinical the Waxman-Hatch Act on New Drugs and Biologics Approved 1984-1995, 2 J. BIOLAW AND BUS. 63, 66trials and FDA review take. Total effective (1999)); see also Linck 4/9 at 97; Browder 3/19 at 174-75;patent term may not exceed more than 14 Seide 3/19 at 176; Armitage 3/19 at 176-77. But see NIHCM Foundation Issue Brief, Prescription Drugs and Intellectual Property Protection: Finding the Right Balance Between Access and Innovation 1, 3 (Aug. 2000) (arguing that the effective patent term has increased by at 18 least 50% since the passage of the Hatch-Waxman See Glover 3/19 at 172-74; Armitage 3/19 at Amendments), at http://www.nihcm.org/prescription.pdf.176-77. 24 19 A patent’s term is 20 years from the date of See Armitage 3/19 at 178. filing the application. Due to the time-consuming nature of the patent examination process, most patents are unlikely to 20 See id.; McCurdy 3/20 at 36-37. have an effective patent term of 19 or 20 years. See 35 U.S.C. § 154(a)(2), as amended by the Uruguay Round 21 Drug Price Competition and Patent Agreements Act of 1994, Pub. L. No. 103-465, whichRestoration Act of 1984, Pub. L. No. 98-417, 98 Stat. 1585 changed patent term from 17 years measured from date of a(1984) (codified as amended 21 U.S.C. § 355 (1994)). patent’s issuance to 20 years measured from date of filing the patent application. 7
  • 117. companies.25 build on and interact with many other features of existing technology.”26 In the Bo x 3-3 . Generic Drug Entry Prior to Patent pharmaceutical industry, incremental Expiration: An FTC Study innovation generally falls into one of three categories. The modified product may use a In light of the questions its various generic new formulation, such as a transdermal drug investigations raised, the Commission began an industry-w ide stud y of generic drug co mpe tition in patch instead of a pill, may combine two October 2000. The Generic Drug Study focused previously approved active ingredients, or solely on the procedures used to facilitate generic drug may use a new salt or esther, which is a entry prior to expiration of the patent(s) that protect more purified form of the original chemical the brand-name drug product. The Commission issued entity. 27 Several panelists suggested that nearly 80 special orders - pursuant to Section 6(b) of the FT C Act - to brand-name comp anies and to generic brand-name companies have responded to drug manu facturers, seeking inform ation about certain effective patent term reduction and the practices. The Commission staff compiled the increasing cost of discovering and information received to provide a factual description developing NCEs by implementing product of how the 180-day marketing exclusivity and 30- life-cycle management, including the use of mon th stay pro visions affect the timing of generic entry prior to patent expiration. Based on this data, IMDs.28 Some have noted that IMDs the Commission made two primary recomme ndations “provide a high return on investment.”29 concerning the 30-month stay provision and the 180- day exclusivity to mitigate the possibility of abuse that Participants in the Hearings deters more generic drugs from be com ing available. expressed differing views about the benefits The Gene ric Drug S tudy is available at http://www .ftc.gov/o s/200 2/07 /genericdrugstudy.p df. of these modified drugs. Some testified that IMDs benefit consumers by providing more convenient dosing or “superior therapeutic4. Incremental Innovation for the Development of IMDs 26 See Merges & Nelson, 90 COLUM. L. REV. at The other main type of innovation in 881.the pharmaceutical industry consists of 27 See NIHCM Innovation Report at 5; Armitageenhancing known chemical entities by 3/19 at 217.formulating new dosage forms or additional 28 See Linck 4/9 at 97-98; Aventismethods of use for existing chemical Pharmaceuticals Inc., Comments of Dr. Nahed Ahmed, Viceentities. This type of innovation is generally President, Productivity, Portfolio & Project Managementdescribed as “incremental,” which, in Drug Innovation & Approval Aventis Pharmaceuticals Inc. (Public Comment) 3-4 (contending that there are stronggeneral terms, means that “todays advances economic incentives for brand-name companies to implement IMDs, because they are “safer, faster, and more cost effective for the development as an incremental 25 For example, the safety and efficacy data for a improvement rather than an original product.”), atproduct may not be relied upon by another company for http://www.ftc.gov/os/comments/intelpropertycomments/avfive years if the product contains an NCE and for three entis.pdf (hereinafter Aventis (stmt)); Armitage 3/19 atyears if the product involves a new use of an existing 216-218; Snyder 3/19 at 224; NIHCM Innovation Report atcompound. 21 U.S.C. § 355(c)(3)(D)(ii). A drug product 3.also can obtain an additional six months of market 29exclusivity if it conducts studies showing the product is NIHCM Innovation Report at 4; see alsosafe and effective for children. 21 U.S.C. § 355a. Aventis (stmt) 4. 8
  • 118. properties than the original formulation,”30 clinical improvement.”36or by serving certain patient populationsbetter than the original product.31 The C. The Role of Patents Inbrand-name companies’ trade association Spurring Pharmaceuticalstated that if physicians and consumers Innovationchoose IMDs in preference to genericalternatives of the original brand-name Panelists reported that patentproduct, the modified drug is warranted.32 protection promotes innovation in theIn contrast, a generic drug manufacturer pharmaceutical industry by creatingsuggested that IMDs might be a tactic incentives for brand-name companies toemployed by brand-name companies “to innovate, and by disclosing inventions,extend patent monopolies beyond the patent thereby encouraging generic companies toexpiry of the new chemical entity . . . by a innovate by designing around brand-namematter of years, not days or weeks or company patents.months.”33 This panelist also argued that thePTO issues too many questionable patents, Participants in the Hearingswhich create a gridlock of patent litigation in overwhelmingly expressed the view thatthe district court system and thereby delay patent rights for pharmaceuticals aregeneric entry.34 The FTC’s Generic Drug essential for brand-name companies toStudy found that over time, for blockbuster prevent free riding and recoup theirproducts, brand name companies are suing significant investments in research andfor infringement on more patents, and those development of NCEs.37 One panelist notedsuits take longer on average than suits that patents are particularly important in theinvolving a single patent.35 Others have pharmaceutical industry, because the Hatch-reported that “the FDA view[s] the vast Waxman Amendments permit genericmajority of IMDs as providing no significant applicants to rely on the brand-name company’s proprietary data demonstrating the safety and efficacy of the brand-name drug product.38 30 Glover (stmt) 7. 36 NIHCM Innovation Report at 7; see also 31 Coffin-Beach 3/19 at 201-05 (stating that IMDs may have See Snyder 3/19 at 224. “questionable therapeutic merit.”). 32 See PhRMA (stmt) 29-30; see also Glover 37 See PhRMA (stmt) 10-13; Glover (stmt) 2, 4(stmt) 7. (describing the cost of new drug development and generic entry); Linck 4/9 at 48-49; Armitage 3/19 at 165; see supra 33 Coffin-Beach 3/19 at 201-05, 212-213 Ch. 2(B)(1)(b) (discussing economic studies on the role of(suggesting that brand-name companies time their patents in protecting against free riding in differentincremental modifications to maximize their product’s industries).franchise, for example, by waiting 10 years to develop asustained-release version of an NCE). 38 See Armitage 3/19 at 133, 165. The FDA considered retesting of generic drugs to be wasteful if the 34 Coffin-Beach 3/19 at 204-205. underlying drug is safe and effective. Moreover, such retesting is unethical because it requires that some sick 35 See FTC, Generic Drug Study at 47-48. patients take placebos and be denied treatment known to be 9
  • 119. Patent law requires applicants to that cover the drug product is to designdisclose the inventions for which they seek around those patents.43 Representatives ofpatents. The purpose of the disclosure generic companies observed that the processobligation is to foster further innovation by of designing around brand-name patents canenabling a person skilled in the particular art give rise to innovation.44 In someto learn from another’s invention.39 This circumstances a generic company maydisclosure obligation is a trade-off for obtain a patent for its design-aroundobtaining the right to exclude others from innovations.45making, using, offering for sale or selling aninvention.40 Several panelists observed that D. The Role of Competition inthe disclosure requirement fosters Spurring Pharmaceuticalinnovation in the pharmaceutical industry byenabling both brand-name and generic Innovationcompanies to discern the development plansand scientific development of rival Panelists described competitioncompanies.41 One panelist reported that among brand-name companies and the rolepatent literature is an important source of of the Hatch-Waxman Amendments ininformation on technological advances for fostering competition and innovation in thethe pharmaceutical industry, whereas pharmaceutical industry. One panelistscientific literature, much of which is observed that the granting of aenabled by patents, is more important in the pharmaceutical patent does not necessarilybiotechnology industry.42 confer a “monopoly on the treatment of any specific disease;” brand-name companies One way in which a generic company may compete with each other in the samecan compete with a particular brand-name therapeutic class, such as drugs that reduceproduct prior to the expiration of the patents cholesterol.46 Moreover, according to the brand-name companies’ trade association, competition among brand-name companieseffective. See H.R. REP. No. 98-857, Part I at 16 (1984), is increasing, because the period of marketreprinted in 1984 U.S.C.C.A.N. 2647, 2649. The ability of exclusivity between the introduction ofother companies to rely on that data and developbioequivalent generic versions of NCEs at much lower breakthrough medicine and competingcosts significantly reduces the profits for the branded innovators has been consistently shrinkingproduct. One panelist stated that once a certain drug has ageneric counterpart, the result is a “more rapid decline inthe pioneer share of the market” because pharmacy benefitmanagers and formulary managers require that physiciansand patients use generic drugs, as opposed to the more 43 For further details, see FTC, Generic Drugexpensive branded drugs. See Glover 3/19 at 171. Study. For discussion of design-around innovation by brand-name companies, see Armitage 3/19 at 230. 39 See supra Ch. 2(I)(A)(3). 44 See, e.g., Browder 3/19 at 228. 40 Rogan 2/6 at 21. 45 See Coffin-Beach 3/19 at 225. 41 See Coffin-Beach 3/19 at 212; Glover 3/19 at224-25; Seide 3/19 at 226; Browder 3/19 at 238; Oehler 46 See Glover (stmt) 6. But see NIHCM2/26 at 319. Innovation Report at 3 (suggesting that price competition among several new drugs products in a therapeutic class is 42 See Blackburn 2/26 at 319-20. limited.). 10
  • 120. since 1965.47 None of the panelists believed, obstacles to generic competition, the Hatch-however, that competition alone could Waxman Amendments “stimulated thegenerate sufficient innovation in the development of a generic pharmaceuticalpharmaceutical industry.48 industry in the United States. Since the law’s passage, the generic industry’s share One of the unique aspects of the of the prescription drug market has jumpedpharmaceutical industry is how the from less than 20 percent to almost 50regulatory structure governing the approval percent today.”50 The Hatch-Waxmanof new brand-name and generic drug Amendments have fostered significant priceproducts has spurred additional competition competition in those markets with genericand innovation. In this case, the Hatch- entry.51 The generic competition spurred byWaxman Amendments sought to balance Hatch-Waxman has forced brand-nameincentives for continued innovation by firms to come up with new products toresearch-based pharmaceutical companies replenish their revenue streams.52 Brand-and opportunities for market entry by name companies often have introducedgeneric drug manufacturers. The IMDs for which they can seek patentstreamlined approval process gives generic protection to lessen the impact of thisdrug applicants the opportunity to obtain generic competition.53FDA approval of their generic drug productsprior to patent expiration.49 By removing Congress also encouraged generic 47 See Glover (stmt) 7; PhRMA (stmt) 28. But infringed by the generic product.see Sal Ricciardi, Comments Re: Competition and 50Intellectual Property Law and Policy in the Knowledge- See Glover (stmt) 7; see also AshokeBased Economy Public Hearings, Spring and Summer, Bhattacharjya, FTC Health Care Workshop: Panel on2002 (Public Comment) 10 (discussing restraints on Branded and Generic Pharmaceuticals 5 (stmt presented atsecondary market competition), at the FTC’s Healthcare Workshop Sept. 10, 2002), athttp://www.ftc.gov/os/comments/intelpropertycomments/pd http://www.ftc.gov/ogc/healthcare/bhatta.pdf; FTC,a.pdf. Generic Drug Study at (i) (identifying these figures as shares of prescriptions filled). 48 Panelists disagreed on the extent to which 51innovation would occur in the pharmaceutical industry Studies indicate that the first generic typicallyabsent patent protection, although all believed that it would enters the market at 70 to 80 percent of the price of thedecline markedly. Professor Snyder, who has conducted corresponding brand and rapidly secures as much as a two-research into this particular issue, cited findings indicating thirds market share. See, e.g., Congressional Budget Office,that in the absence of patent protection for pharmaceuticals, How Increased Competition from Generic Drugs Hasinnovation would decrease by approximately 60%. Affected Prices and Returns in the PharmaceuticalArmitage disagreed with Snyder, asserting that the absence Industry 28 (July 1998), atof patents would eliminate innovation in the http://www.cbo.gov/showdoc.cfm?index=655&sequence=0pharmaceutical industry. Compare Snyder 3/19 at 170 with ; DAVID REIFFEN & MICHAEL R. WARD, GENERIC DRUGArmitage 3/19 at 180. INDUSTRY DYNAMICS (Federal Trade Commission Bureau of Econ. Working Paper No. 248, 2002), at 49 Brand-name companies must provide the FDA http://www.ftc.gov/be/econwork.htm; see also BE Staffwith information regarding patents that cover their drug Report, The Pharmaceutical Industry.products, which the FDA then lists in a publication 52commonly known as the “Orange Book.” See 21 U.S.C. § See, e.g., Glover 3/19 at 146 (noting that355(j)(7)(A) and FTC, Generic Drug Study at Ch. 3. “even major companies must develop a blockbuster everyGeneric drug companies who seek FDA approval prior to two to three years or face massive financial contraction”).patent expiration must give notice to brand-name 53companies stating that the listed patents are invalid or not Browder 3/9 at 227-28. 11
  • 121. entry by granting 180 days of marketing E. The FTC’s Pharmaceuticalexclusivity to the first generic applicant to Industry Enforcementfile an application for a generic drug product Actions and Generic Drugthat does not infringe the brand-nameproduct or that challenges the validity of the Studybrand-name company’s patents.54 The 180-day exclusivity period increases the The Commission has pursuedeconomic incentives for a generic company numerous antitrust enforcement actionsto be the first to file, because the generic affecting both brand-name and generic drugapplicant has the potential to reap the reward manufacturers when it had reason to believeof marketing the only generic product (and, that a company abused its patent rights inthus, to charge a higher price until more violation of the antitrust laws. Thegeneric products enter). Through this 180- Commission has addressed conduct that itday provision, the Amendments provide an alleged would have the effect of delayingincreased incentive for companies to generic entry, including certain patentchallenge patents and develop alternatives to settlement agreements between brand-namepatented drugs.55 Indeed, one generic companies and generic applicants,58 a brand-panelist reported that competition among name company’s acquisition of an exclusivegeneric companies for the 180 days of license to a particular patent,59 the purportedexclusivity has become “acute.”56 Once a brand-name company isnotified of the filing of such a genericapplication, it has a 45-day window in whichto sue the generic applicant for patent 58 Abbott Laboratories, No. C-3945 (FTC Mayinfringement. The initiation of the patent 22, 2000) (consent order), available atinfringement suit triggers a 30-month stay of http://www.ftc.gov/os/2000/05/c3945.do.htm. Geneva Pharmaceuticals, No. C-3946 (FTC May 22, 2000)FDA approval of the generic drug (consent order), available atapplication. According to the legislative http://www.ftc.gov/os/2000/05/c3946.do.htm. Hoechst/Andrx, No. 9293 (FTC May 8, 2001) (consenthistory, the stay allows for the order), available atcommencement of a lawsuit and takes into http://www.ftc.gov/os/2001/05/hoechstdo.htm. In anotheraccount the patent owner’s rights while still matter, Schering-Plough, the Commission resolved all claims against one of three respondents, American Homeencouraging generic entry.57 Products (AHP), by issuing a final consent order. Schering-Plough Corp., No. 9297 (FTC Apr. 2, 2002) (consent order as to AHP), available at http://www.ftc.gov/os/2002/04/scheringplough_do.htm. 54 For a fuller discussion of the effect of the 180- The case against the other two respondents is inday marketing exclusivity provision on competition, see litigation before the Commission. See Schering-PloughFTC, Generic Drug Study at Ch. 3. Corp., No. 9297 (FTC July 2, 2002) (Initial Decision), available at 55 See Granutec, Inc. v. Shalala, 139 F.3d 889, http://www.ftc.gov/os/2002/07/scheringinitialdecisionp1.pd891 (4th Cir. 1998). f. 56 Coffin-Beach 3/19 at 239. 59 Biovail Corp., No. C-4060 (FTC Oct. 2, 2002) (consent order), available at 57 H. REP. NO. 98-857, at 27 (1984). http://www.ftc.gov/os/2002/10/biovaildo.pdf. 12
  • 122. use of sham litigation,60 and an agreement addressed allegations that individual brand-between generic drug manufacturers.61 It name manufacturers have delayed genericalso has addressed conduct that the competition through the use of improperCommission contended would eliminate a Orange Book listings65 that trigger thepotential competitor for an NCE in the Hatch-Waxman provision prohibiting themerger context.62 FDA from approving a generic applicant for 30 months.66 Over the past few years theCommission also has observed through its Brand-name companies previouslyinvestigations, law enforcement actions, and could obtain additional 30-month stays byGeneric Drug Study that some brand-name obtaining additional patents that claimedand generic drug manufacturers may have their brand-name products. There were“gamed” the 180-day marketing exclusivity opportunities for “gaming” the 30-monthand the 30-month stay provisions, stay because the FDA does not overseeattempting to restrict competition beyond whether these additional patents meet thewhat the Hatch-Waxman Amendments requirements for listing with the FDA, andintended.63 The Commission has undertaken there is no private right of action for a courttwo main types of enforcement activities in to make such a determination. Notthis area. It has addressed patent settlement surprisingly, given the amount of revenue atagreements between brand-name companies stake, the FTC found in its Generic Drugand generic applicants that the Commission Study that some brand-name companiesalleged had delayed the entry of one or more have “gamed” the 30-month stay provision,generic applicants through manipulation of and that it had the potential to be “gamed” inthe 180-day exclusivity period.64 It also has the future, absent reform.67 The FDA changed its rule to prevent brand-name companies from obtaining additional 30- 60 Bristol-Myers Squibb, No. C-4076 (FTC Mar. month stays. This rule change was based7, 2003) (consent order), available athttp://www.ftc.gov/os/2003/03/bristolmyersconsent.pdf. 61 Biovail Corp. and Elan Corp. PLC, No. C-4057 (FTC Aug. 20, 2002) (consent order), available at 65 The Commission first raised concerns abouthttp://www.ftc.gov/os/2002/06/biovailelando.pdf. the potential anticompetitive impact of improper Orange Book listings in American Bioscience, Inc. v. Bristol-Myers 62 In the Matter of Glaxo Wellcome plc, and Squibb Co., No. CV-00-08577 (C.D. Cal. Sept. 7, 2000).SmtihKline Beecham PLC, No. C-3990 (FTC Jan. 31, See Brief of Amicus Curiae Federal Trade Commission,2001) (consent order) (requiring divestiture of certain Am. Bioscience, Inc. v. Bristol Myers Squibb Co., 2000intellectual property rights on NCEs), available at U.S. Dist. LEXIS 21067 (No. CV-00-08577), available athttp://www.ftc.gov/os/2001/01/glaxosmithklinedo.pdf. http://www.ftc.gov/os/2000/09/amicusbrief.pdf. 63 66 For further details on the Generic Drug Study Biovail Corp., No. C-4060 (FTC Oct. 2, 2002)see Box 3-3. (consent order), available at http://www.ftc.gov/os/2002/10/biovaildo.pdf; Bristol- 64 Abbott Laboratories, No. C-3945 (FTC May Myers Squibb, No. C-4076 (FTC Mar. 7, 2003) (consent22, 2000) (consent order), available at order), available athttp://www.ftc.gov/os/2000/05/c3945.do.htm. Geneva http://www.ftc.gov/os/2003/03/bristolmyersconsent.pdf.Pharmaceuticals, No. C-3946 (FTC May 22, 2000) 67(consent order), available at See FTC, Generic Drug Study at (ii)-(iv) andhttp://www.ftc.gov/os/2000/05/c3946.do.htm. Ch. 3. 13
  • 123. largely on the FTC’s recommendation.68 spurs for the discovery and commercialization of NCEs in theF. Conclusion pharmaceutical industry in many ways showcases the patent system’s benefits. Representatives from the Such innovation entails the high fixedpharmaceutical industry emphasized that research costs, relative ease of imitation, andpatents are critical for promoting free riding problems that patent protectionpharmaceutical innovation of NCEs. Brand- effectively manages. Fewer patent thicketname companies depend on patents to issues arise in the pharmaceutical contextrecoup their substantial investment in the than in industries where innovation is lessdiscrete innovation that leads to the discrete and individual products are covereddevelopment of new drug products. Also, by many patents. Subsequent sectionsbrand-name companies make and patent examine how the roles of patents andincremental improvements to their products competition vary in industries that exhibitto manage them on a life-cycle basis. different characteristics.Panelists differed as to the extent to whichsuch IMDs benefit consumers. Competition in the pharmaceuticalindustry occurs in two primary ways:between brand-name companies that haveproducts in the same therapeutic class andbetween brand-name and generic companies.Competition between and among brand-name companies and generics can fosterinnovation, as well as other benefits ofcompetition. Patent disclosure requirementscan enable brand-name and genericcompetitors to design around some patentscovering brand-name drug products in orderto bring competing products to market. TheCommission has brought enforcementactions in the pharmaceutical industry toprotect competition, including incentives toinnovate. The innovation that the patent system 68 See Applications for FDA Approval to Marketa New Drug; Patent Listing Requirements and Applicationof 30-Month Stays on Approval of Abbreviated New DrugApplications Certifying That a Patent Claiming a Drug isInvalid or Will Not be Infringed, 68 Fed. Reg. 36,675(2003) (to be codified at 21 C.F.R. § 314). 14
  • 124. III. THE BIOTECHNOLOGY Lee Bendekgey, representing Incyte INDUSTRY Genomics; Robert Blackburn, representing Chiron Corp.; Monte R. Browder, representing Ivax Corporation; BarbaraA. Introduction Caulfield, representing Affymetrix, Inc.; David Coffin-Beach, representing The biotechnology industry also Torpharm, Inc.; David J. Earp, representingrelies primarily on patents to provide Geron Corp.; Michael K. Kirschner,incentives to invest in innovation. representing Immunex Corp.; and RossBiotechnology companies seek patent Oehler representing Aventis Corp. Rochelleprotection to appropriate the value of their K. Seide, from Baker Botts, LLP, alsoinventions, to attract investment from capital participated in a business perspective panelmarkets, which funds their costly research, on the biotechnology industry.and to facilitate inter-firm relationshipsnecessary for commercial development oftheir inventions. Patent disclosures can B. Industry Descriptionassist biotechnology firms in focusing theirR&D efforts on areas not covered by The biotechnology industry usespatents. Competition also encourages cellular and molecular (i.e., biological)innovation, for example, as firms race to processes to address problems or makedevelop new technologies. products. R&D in the biotechnology industry focuses on cells and large biological Although panelists generally agreed molecules (such as DNA and proteins)on the benefits of patents in the rather than the chemical compounds that thebiotechnology industry, many panelists also pharmaceutical industry uses to make smallstated that the issuance of questionable molecule drug products.69patents is harming innovation in theindustry, and that the mechanisms for Cells are the basic building blocks ofchallenging such patents, including all living things. Plants, animals, andlitigation, are inadequate. Some also humans are incredibly diverse, yet there areexpressed concern that the need for multiple, remarkable similarities among the speciespatented research tools has the potential to that are invisible to the naked eye. All livingcreate difficulties for follow-on innovation. things use essentially the same cellularOthers discussed how licensing practices, processes and speak the same geneticsuch as reach-through license agreements language.70 This unity at the cell level ofand patent pools, can be used to surmount different species provides the foundation forsome of these difficulties by facilitating biotechnology research.access to research tools that promote furtherinnovation. Participants asserted that R&D The panelists who represented 69 See Beier 2/26 at 248.biotechnology firms or organizations at theHearings were David W. Beier, Counsel to 70 See Biotechnology Industry Organization,the Biotechnology Industry Organization; Biotechnology: A Collection of Technologies, at http://www.bio.org/er/technology_collection.asp. 15
  • 125. spending in the biotechnology industry “is biotechnology industry participants aremore than double the average of the small, particularly relative to thepharmaceutical industry (both on a per pharmaceutical industry, and lack internalemployee basis and as a percentage of sales), financial resources sufficient for undertakingand the pharmaceutical industry is several extensive drug development.76times more R&D intensive than any otherindustry.”71 R&D is particularly lengthy for Although innovation in the biotechbiotechnology firms, because biotechnology industry has many facets, it generally resultsinnovation is more uncertain than innovation in two classes of inventions.77 One classin other industries.72 Panelists also noted relates to newly discovered and isolatedthat the commercialization of biotechnology genes or proteins or to pharmaceuticalresearch is particularly difficult, due to three inventions based on those genes or proteins.factors. First, as discussed above in relation Although one cannot patent a naturally-to the pharmaceutical industry, the drug occurring gene or protein as it exists in adevelopment process is time-consuming, plant, animal, or human, one can patent auncertain, and expensive. One panelist gene or protein that has been isolated fromnoted that his company took 10 years to the body and is useful in that form as abring its first product to market, and another pharmaceutical drug or other application.786 years before it brought its second product The other class of biotechnology inventionsto market.73 Second, much biotechnology relates to methods of treating patients with aresearch is basic, at least a step removed given disease through the use of a particularfrom the more applied research that is gene or protein. Even if someone has adirectly susceptible to commercialization.74 patent on a gene or protein, a researcher whoBiotechnology thus highlights the issues that discovers a new method of use for that genelie at the core of the prospect theory or protein can patent the new method ofregarding incentives for, and efficiencies in, use.79bridging the gap between basic research andultimate commercial sales.75 Third, most The biotechnology industry is closely related to the pharmaceutical industry. One panelist observed that both industries try to 71 Biotechnology Industry Organization,Testimony (2/26/02) 2, athttp://www.ftc.gov/opp/intellect/020226davidwbeier.pdf(hereinafter BIO (stmt)); Kirschner 2/26 at 240. 76 See Earp 2/26 at 252. 72 See Beier 2/26 at 248-49; Kirschner 2/26 at 77 See generally Biotechnology Industry240. Organization, Primer: Genome and Genetic Research, Patent Protection and 21st Century Medicine, at 73 See Kirschner 2/26 at 239. http://www.bio.org/genomics/primer.html (hereinafter BIO, Primer). 74 See Earp 2/26 at 252; Seide 3/19 at 167. 78 See United States Patent and Trademark 75 See Rai 4/10 at 21 (citing bio-pharmaceuticals Office, Manual of Patent Examining Procedure § 2105 (8thas a context in which “patents serve not only the traditional ed. 2001), available atincentive function but also serve the function of http://www.uspto.gov/web/offices/pac/mpep/mpep.htmincentivizing further commercialization and (hereinafter MPEP); Seide 3/19 at 167-68.development”); see generally supra Ch. 2(III)(A)(1) 79(discussing Professor Kitch’s prospect theory). See BIO, Primer at 6. 16
  • 126. discover end-use products.80 Indeed, small their initial success by introducing a productmolecule-type research, the aim of which is with no comparable or rival product.86 Afterto produce a traditional pharmaceutical drug this success, much bigger and better fundedproduct, has become much more efficient competitors entered the market, thus addingthrough the use of biotechnology tools such competitive pressure to keep innovating.87as proteins and genomic sequences.81 Also, In general, however, although panelistsmany biotechnology companies conduct found competitive forces important, theybasic research to identify promising placed emphasis on the role of patents asproducts, and then partner with a drivers of innovation in the biotech industry.pharmaceutical company to test andcommercialize the product.82 Patents D. The Implications of Patentfacilitate this process; there is a tremendous Protection for Innovationamount of licensing, as well as acquisitionactivity, between the two industries 1. The Role of Patents in Spurringsearching for synergies to bring products to Innovation in the Biotechnologymarket.83 IndustryC. The Role of Competition in a. Patentability Encourages Spurring Biotechnology Investment in R&D Innovation In 1980, the Supreme Court in Several panelists discussed the role Diamond v. Chakrabarty88 decided thatof competition in spurring biotechnology living organisms produced by humaninnovation.84 One panelist commented that intervention are patentable. Participants“one thing that competition does is, it sure stated that the biotechnology industry wouldmakes you hurry up.”85 Drawing on his not have emerged “but for the existence ofexperience in the biotech industry, he predictable patents,”89 and that Chakrabartyobserved that companies typically found spurred significant growth in the biotechnology industry.90 Their discussion describes the role of patents in an industry 80 See Blackburn 2/26 at 250-51. with a very costly, high-risk R&D process and a structure consisting significantly of 81 See Seide 3/19 at 188-89, 244-45 (discussing“rational drug design”); Blackburn 2/26 at 250, 261-62. 82 See, e.g., Blackburn 2/26 at 251; Earp 2/26 at252. 86 See Bendekgey 2/26 at 285-86. 83 See Bendekgey 2/26 at 257-59; Oehler 2/26 at 87254. See id. 88 84 See, e.g., Caulfield 3/19 at 242-43. Diamond v. Chakrabarty, 447 US 303 (1980). 85 See Bendekgey 2/26 at 286. Patent races may 89lead to excessive R&D in a particular area, although See Kirschner 2/26 at 240-41, 328.distinguishing beneficial from wasteful overlapping efforts 90may prove difficult. See supra Ch. 2(III). BIO (stmt) 4. 17
  • 127. small, not-yet-profitable firms.91 research.97 By contrast, a panelist from a pharmaceutical firm with a biotechnology A biotechnology trade association affiliate noted that “there is value to behighlighted one particular role of patents in found in patents as literature.”98 Anotherthis setting: patentability of biotech panelist noted that “the information transferinventions enables the biotechnology happens in the scientific literature [ratherindustry “to attract venture capital.”92 than] the patent literature,” but added thatBiotechnology companies overwhelmingly “quite a bit of the scientific literature isunderscored the importance of patents for enabled by the fact that there’s been a patentattracting venture capital.93 As one of these filed on it.”99 This panelist observed thatpanelists stated, “patents are indeed the key patent literature is a more important sourceasset for us. They enable us to have access of information in the pharmaceuticalto the capital markets and to continue our industry than the biotechnology industry.100innovation and development.”94 Theventure capital accessed through patents thus c. Patenting of Biotechnologyenables not-yet-profitable companies to Research Tools“sustain . . . innovation through massiveinvestments in research and development.”95 A research tool is a technology that is used by pharmaceutical and biotechnologyb. The Role of Patent Disclosures in companies to find, refine, or otherwise Fostering Biotechnology design and identify a potential product or Innovation properties of a potential drug product.101 As such, it serves as a springboard for follow- The panelists differed on the extent on innovation. Examples of these types ofto which required patent disclosures enabling tools include high-throughputencourage the dissemination of information screening technologies, micro-array-typeand, therefore, foster follow-on innovation technologies, genomic databases, andin biotech.96 One panelist stated that thepatent literature “has not been a significantsource of ideas” for the company’s 91 See id. at 2, 4; Beier 2/26 at 265-66;Blackburn 2/26 at 275-76. 97 Kirschner 2/26 at 318. 92 BIO (stmt) 4. 98 Oehler 2/26 at 319. 93 See Earp 2/26 at 237; Bendekgey 2/26 at 256;Blackburn 2/26 at 263. 99 See Blackburn 2/26 at 319. 94 See Earp 2/26 at 326. 100 See id. at 320. 95 BIO (stmt) 4. 101 See Blackburn 2/26 at 250, 260 (noting that there are likely to be slightly varying definitions of research 96 See Kirschner 2/26 at 318; Blackburn 2/26 at tools); Bendekgey 2/26 at 267-68, Cohen 10/30 at 150,319; Oehler 2/26 at 319. McGarey 11/6 at 160. 18
  • 128. Bo x 3-4 . Effects of Research Tool Patents and Licensing on Biomedical Innovation John P. W alsh, Ashish Arora & W esley M. Cohen in P ATENTS IN THE K NOWLEDGE -B A S ED E C O N O M Y 285 (W esley M . Cohen & Stephen A . Merrill eds. 200 3), available at http://books.nap.edu/books/0309086361/html/285.html#pagetop. John P. W alsh, Ashish Arora, and W esley M. Cohen conduc ted an empirical study of the implications for innovation of patenting and licensing practices in the pharmaceutical and biotech industries. The authors conducted “70 interviews with IP attorneys, business managers and scientists from 10 p harmaceutical firms and 15 biotech firms, as well as university researchers and technology transfer officers from 6 universities, patent lawyers and government and trade association personnel.” The authors found that patents on research tools have increased, but have not significantly hindered drug disco very. The increased comp lexity of the patent landscape, they concluded, has not resulted in a tragedy of the anticommons. (See Box 3-5 for further explanation of this theory.) They noted that some university research has been delayed by restriction s on the use of p atented genetic diagnostics, and that there have been some delays or access restrictions to research tools or other foundational discoveries. In some instances, research was re-directed to areas where there were fewer patents. Overall, however, the researchers found that no valuable research projects were halted as a result of limited access to a research tool. The authors cautioned, however, that the potential exists and ongoing scrutiny is wa rranted . See infra Ch. 3(III)(D )(4). The authors also concluded that firms and universities use a range of strategies to avoid breakdo wn and restricted access to research tools, including taking licenses, inventing around patents, infringement (often informally invoking a research exem ption), developing and using public tools and challenging patents in court. New PT O guidelines, active intervention by the NIH, and overall shifts in the courts’ attitudes towards research tool patents also have lessened these potential threats, they found. A new Federal Circuit case that stated a narrow scope of the research exemption available to universities led the authors to question the extent to which some of these findings will rema in app licable. The relevant Fed eral Circuit case , Ma dey v. Du ke U niversity, 307 F.3d 135 1, 13 62 (Fed . Cir. 20 02), cert denied, 123 S. Ct. 2639 (2003), is discussed infra Ch. 4(II)(D). modeling programs. Research tools are example, one panelist explained that withgenerally patentable. Researchers require a gene chip array technology “what used tolicense to use patented research tools to take a post-doc[toral student] in theidentify and develop inventions, but laboratory approximately six months withtypically do not require a license from the proper front-end research can now be doneresearch tool patent holder to practice the in 20 minutes.”104 Another panelistensuing inventions.102 suggested that research tools have led to a considerable reduction in the cost and time Several commentators discussed the required for the targeting of therapeuticbenefits to innovation derived from using antibodies during the initial stages of newand patenting research tools.103 For drug research. He mentioned “a very small pre-IPO firm that has moved into a phase two product in three years based on research 102 See Blackburn 2/26 at 260. tool technology” and went on to state that 103 this would have been “inconceivable to have See id. at 262; Bendekgey 2/26 at 258-59 and267-68; Seide 3/19 at 167. For discussion of issues raisedby research tool patents, see John P. Walsh et al., Effects ofResearch Tool Patents and Licensing on Biomedical http://books.nap.edu/books/0309086361/html/285.html#paInnovation, in PATENTS IN THE KNOWLEDGE -BASED getop (hereinafter Research Tool) and Box 3-4.ECONOMY 285 (Wesley M. Cohen & Stephen A. Merrill 104eds. 2003), available at See Caulfield 3/19 at 135. 19
  • 129. happened 20 years ago, before the invention particular area of research.110 Two panelistsof research tools.”105 observed that questionable patents create a “significant drag” on competition, and Two panelists stressed the another panelist stated that questionableimportance of patenting research tools.106 patents have a “chilling effect on both publicOne of them asserted, for example, that “if and private sector research.”111there’s anything you want to protect andincent with patents, it’s the research tool One panelist stated his personal viewtechnology.”107 He argued that patent that “the PTO’s ability to provide aprotection will be critical in encouraging meaningful examination of biotechnologyinvestment in the next generation of research patents right now is in a cris[i]s.”112tools, which might reduce the costs and time Acknowledging the dedication and qualityrequired for the clinical trial phases, which of the PTO’s examiners, this panelist notedare the most “expensive part” of the drug that the examiners are under such timedevelopment process.108 constraints that they may be unable to conduct a meaningful patent examination.1132. The Quality of Biotechnology According to this panelist, the PTO should Patents 110 Panelists discussed concerns with the See Earp 2/26 at 238, 290-91; Caulfield 3/19 at 159; Blackburn 2/26 at 296.quality of biotechnology patents. Many ofthe panelists observed that poor quality 111 Caulfield 3/19 at 159; Barbara A. Caulfield,patents can hinder innovation and Business Perspectives on Patents: Biotech and Pharmaceuticals, Federal Trade Commission/Departmentcompetition.109 A number of panelists stated of Justice Hearings (3/19/02) (slides) at 6, atthat poor quality patents can harm http://www.ftc.gov/opp/intellect/020319barbaracaufield.pdfinnovation and competition by deterring a (hereinafter Caulfield Presentation); Blackburn 2/26 at 296, Kirschner 2/26 at 328.rival firm from entering or continuing with a 112 See Kirschner 2/26 at 242. Mr. Kirschner voiced concerns with patents issued to wrong parties or to multiple parties on the same invention; patents that 105 “contain overly-broad claims in view of the prior art or the Blackburn 2/26 at 261, 262 (discussing the scope of what was enabled or the scope of what wasscreening of small molecules); Oehler 2/26 at 277-78 described” id at 242; and patents for which “the best prior(noting that research tools offer “great promise,” but as yet art was not cited to the patent office, was not discovered byhave only reduced the time required for the early phases of the patent office, or was cited to the patent office andresearch). clearly the examiner did not appreciate it.” Id. at 241-42, 106 289. See Blackburn 2/26 at 262; Bendekgey 2/26at 258-59, 267-68. 113 See Kirschner 2/26 at 241-44, 288-90. 107 Similarly, a panelist commented that “examiners have an See Blackburn 2/26 at 262. incentive to move cases along and dispose of them.” See Bendekgey 2/26 at 231 (“I’ve certainly had comments 108 See id. at 262-63. See supra Ch. 3(II)(B) repeated to me to the effect that . . . examiners have an(discussing the phases of pharmaceutical drug incentive to move cases along and dispose of them, anddevelopment). sometimes they think there’s something novel here, they’re not sure what, and so they’re just going to allow it and let 109 See Bendekgey 2/26 at 230; Earp 2/26 at 238; things get sorted out in litigation. And I can tell you, whenKirschner 2/26 at 241; Oehler 2/26 at 292; Blackburn 2/26 you’re at the receiving end of litigation like that it has aat 294. decidedly chilling effect on competition.”). 20
  • 130. “focus on improving quality, at least within 3. The Mechanisms Available for[the biotechnology patent examination Challenging Questionable Patentsgroup],” because patent quality is moreimportant than pendency in the Firms in the biotechnology industrybiotechnology industry.114 Another panelist reported that they avoid infringing evenobserved, “of the issues that people raise . . . questionable patents and therefore refrainin many cases [it] just come[s] down to the from entering or continuing with a particularquality of the examination.”115 field of research.119 Most panelists observed that the two existing mechanisms for Although panelists agreed that poor challenging a questionable patent arepatent quality can adversely affect generally inadequate.120innovation, disagreement existed whetherpatent quality in the biotechnology area was a. Challenging Questionable Patentsany different from that in other industries. Through LitigationOne panelist reported that patent quality isnot a field-specific problem.116 In fact, he Panelists considered litigation to beobserved that biotechnology patents may be an inadequate means of challenging a patentof a higher quality than those in other for three main reasons. First, the pace ofindustries, because of “the concentration of innovation in the biotechnology industry isthe Patent Office on guidelines and so rapid that by the time a court determinesresources in the biotech field” in the last 10 the question of patent validity, the researchyears.117 The representative of a or product opportunity has passed. As onebiotechnology trade association similarly panelist observed, “six months can be anoted that the PTO has responded tremendous amount of time” inaffirmatively to public controversies in biotechnology research, while arelation to biotechnology patents as they biotechnology patent case “takes two tohave arisen and thus has headed off any three years” to litigate.121 Moreover, otherlasting adverse impacts of questionablebiotechnology patents.118 119 See, e.g., Earp 2/26 at 290-91, 238; Blackburn 2/26 at 296; Caulfield 3/19 at 161; see also Alik Widge, Comments Regarding Competition and Intellectual Property (Public Comment), at 114 http://www.ftc.gov/os/comments/intelpropertycomments/wi See Kirschner 2/26 at 243, 329. But cf. dgealik.htm.Armitage 3/19 at 134 (raising concerns with pendencyperiods for biotechnology patent applications). 120 See Bendekgey 2/26 at 231; Earp 2/26 at 238, 115 291, 327; Kirschner 2/26 at 244, 328; Blackburn 2/26 at See Bendekgey 2/26 at 230. 294; Caulfield 3/19 at 160. One panelist noted that a third option exists that permits the public to submit comments to 116 See Oehler 2/26 at 292. the PTO about patent applications published because they have been pending before the PTO for longer than 18 117 See id. months. He also acknowledged this approach was not as “perfect and as targeted as an opposition proceeding, as in 118 See Beier 2/26 at 296 (noting that “the patent Europe.” Oehler 2/26 at 294.system has been remarkably self-correcting.”); see also 121Kirschner 2/26 at 329 (noting the PTO’s responsiveness to Caulfield 3/19 at 160; see also Barton 2/26 atconcerns raised by the industry). 220-21. 21
  • 131. panelists suggested that just because a patent significant costs have been sunk, or avoidis not challenged through litigation does not the area of research.128 Panelists stated theirmean that the patent is not problematic.122 companies usually will choose to avoid the area of research altogether rather than risk Second, the cost of litigation is possible infringement later in the R&Dprohibitively expensive for many firms in process.129 One panelist observed that thethe biotechnology industry. One panelist inability of a company to challenge thereported that a biotechnology patent case validity of a patent unless that companycosts between five and seven million dollars itself has been threatened with litigation byto litigate.123 Such expenditure, this panelist the patent owner results in harm toobserved, on an area that may not end up competition, because “bad patents [are ableproducing revenue is beyond the means of to] . . . sit out there . . . [where] you can’tmost firms in the biotechnology industry.124 touch them.”130According to panelists, most firms tend to besmall and generally have to obtain funding b. Challenging Questionable Patentsfrom the capital markets or venture Through Reexaminationcapitalists because of the difficulties in Procedurescommercializing products.125 Any person at any time may file a Finally, current standing request for reexamination, and if the requestrequirements prevent a potentially infringing raises a substantial new question ofparty from determining in advance the patentability affecting any claim of themerits of a questionable patent.126 A patent, reexamination is commenced.potentially infringing party can seek a Reexamination is available on an ex partedeclaratory judgment to invalidate a patent and inter partes basis.131 The panelistsonly after that party has been threatened with unanimously considered the reexaminationlitigation by the patent owner. Patent procedures as they existed at the time of theowners in the biotechnology industry are hearing inadequate for a third party tocareful to avoid such a situation.127 This challenge the validity of another party’smeans the potentially infringing party has to patent.132 Participants articulated threechoose whether to forge ahead with theresearch, and risk being sued after 128 See id. at 295. 122 See Blackburn 2/26 at 309; Kirschner 2/26 at 129 See Earp 2/26 at 238, 290-291; Caulfield 3/19308. at 159; Caulfield Presentation at 6; Blackburn 2/26 at 296. 123 See Caulfield 3/19 at 160. 130 Blackburn 2/26 at 294-6. 124 See id. 131 For further discussion of reexamination, opposition, and review, see infra Ch. 5(III). 125 See Kirschner 2/26 at 239; Earp 2/26 at 252;Armitage 3/19 at 166; Seide 3/19 at 167. 132 See, e.g., Earp 2/26 at 301, Bendekgey 2/26 at 303, Beier 2/26 at 301, Blackburn 2/26 at 294-96. One 126 See Blackburn 2/26 at 294. panelist wryly observed that as of the time of the hearing the inter partes reexamination procedures had been 127 See id. invoked in only four out of 160,000 cases. See Beier 2/26 22
  • 132. problems with the reexamination system, the European system.136 One panelisttwo of which Congress has addressed by suggested that the best features of thelegislation since the Hearings.133 The existing United States reexamination systemremaining problem panelists cited was that should also be incorporated into anyparticipation in an inter partes opposition system.137reexamination proceeding estops a thirdparty participant from raising a broad Another panelist stated that anspectrum of issues in subsequent court opposition system should be implementedlitigation.134 regardless of whether the problems discussed above in relation to reexaminationc. Challenging Questionable Patents proceedings were addressed by statute.138 In Through a New Opposition System fact, he noted that even if the reexamination proceedings were improved, it “probably Three of the panelists suggested that wouldn’t convince a whole lot more peoplethe United States should implement an to go forward with it.”139 This view was notopposition system for challenging challenged among the panelists.questionable patents.135 These panelistsrecommended that such an opposition 4. The Potential for Patents tosystem draw on the best features of other Impede Innovation in thepatent opposition proceedings, particularly Biotechnology Industry Unlike the pharmaceuticals industry, in which major aspects of the innovation process are relatively discrete, biotechnologyat 301. innovations typically form the basis of, or 133 provide the tools for, independent follow-on These two problems were: a third-party whoinvoked the reexamination procedures was precluded from R&D. Commentators discuss two ways inappealing the PTO’s decision to the federal courts (see BIO which patents have the potential to harm(stmt) 24; Beier 2/26 at 301; Earp 2/26 at 301; Bendekgey follow-on innovation in biotechnology: (1)2/26 at 303); and prior art of record during the patentapplication process could not be the basis for a through the development of anreexamination (see Earp 2/26 at 302). Amendments to the anticommons;140 and (2) through thepatent statute enacted in November 2002 conferred appeal withholding of access to technologiesrights on third party requesters in inter partes patentreexamination proceedings, overruled the decision in In rePortola Packaging Inc., 110 F.3d 786 (Fed. Cir. 1997)(holding that reexamination could not be used if the basis isthe same prior art references that the examiner considered, 136since such references do not raise a substantial new See Earp 2/26 at 238, 291, 327; Bendekgeyquestion of patentability), and clarified that patent 2/26 at 231.reexamination on the basis of previously cited prior art “is 137not precluded.” Patent and Trademark Office See Kirschner 2/26 at 244.Authorization Act of 2002 § 5-6, 35 U.S.C. § 303(a), 312 138(a) 134, and 141-44. See Earp 2/26 at 327. 134 139 See, e.g., Beier 2/26 at 301. See id. 135 140 See Bendekgey 2/26 at 231; Earp 2/26 at 238, For further explanation of this theory, see291, 327; Kirschner 2/26 at 244, 329. Box 3-5. 23
  • 133. needed for follow-on innovation.141 One panelist stated, for example, that the need “to have access to a wide range ofa. The Development of an Anticommons technologies to discover, create, manufacture and market a human Scholars have argued that innovation therapeutic product” means thecan be harmed by the development of an biotechnology industry is “highly vulnerableanticommons, which can arise when to . . . the tragedy of the anticommons”;144 hemultiple property right owners have claims found “the risk of” an anticommonsto separate inputs needed for some product problem.145 He cited the example of Enbrel,or line of research.142 Some panelists which at one time was subject to royaltiesbelieve that an anticommons threatens paid to seven companies.146 The panelistinnovation in the biotechnology industry.143 later noted that the royalty stacking that took place in relation to Enbrel was prior to the advent of research tool patents and reach- 141 One potential limit on such harm may spring through royalties, which, he indicated, havefrom an experimental use defense. Although there is somedebate about its scope, the industry panelists generally increased the likelihood of anticommonsaccepted that an experimental use defense exists at problems.147common law offering some shelter from infringementlitigation to non-commercial research. See Armitage 3/19at 186-87; Polk 3/19 at 190; cf. Thomas 2/8 (Patent In their business survey of theSession) at 30; Sung 2/8 (Patent Session) at 136-38; biotechnology industry, Professors Walsh,Caulfied 3/19 at 163. For further discussion of the researchexemption, see infra Ch. 4(II)(D). In their study of the Arora, and Cohen examined whether thebiotechnology industry, Walsh, Arora, and Cohen noted existence of multiple research tool patentsthat informal reliance on this defense by members of the associated with a new product or processresearch community has helped to prevent an anticommonsor lack of access to existing patents from stifling follow-on poses anticommons concerns.148 Theyinnovation. See Walsh et al., Research Tool at 333-34. concluded that such concerns have “not been especially problematic,” because The Federal Circuit has stated a narrow scope ofthis exemption in an opinion in October 2002: Madey v. mechanisms are being used, such as relyingDuke University, 307 F.3d 1351, 1362 (Fed. Cir. 2002), on a research exemption, obtaining a license,cert denied, 123 S. Ct. 2639 (2003). Some believe that this or inventing around patents, to prevent harmdecision will chill university research, because researcherswill no longer be able to rely on the exemption to overcomeanticommons or access issues. See Cohen 10/30 at 149-52,161-62. “uncertain”), at 142 http://www.ftc.gov/os/comments/intelpropertycomments/ho See Michael A. Heller & Rebecca S. rtonthomasjarticle.pdf.Eisenberg, Can Patents Deter Innovation? TheAnticommons in Biomedical Research, SCIENCE MAG., 144May 1, 1998, available at See Kirschner 2/26 at 241.http://www.sciencemag.org/cgi/content/full/280/5364/698 145and Box 3-5. For further discussion of anticommons and See id. at 310-11.related issues deriving from the presence of multiple 146patents, see supra Ch. 2(III)(C). See id. at 241. He went on to note that one of those companies no longer receives royalties because its 143 See Kirschner 2/26 at 241, 310-11; Caulfield patent expired.3/19 at 163-64; McGarey 11/6 at 153-54. See also Tom 147Horton, Patenting Our Lives and Our Genes: Where Does See id. at 310. Reach-through royalties areCongress Stand in the Coming Clash? 7-8 (noting the discussed below.development of practical problems from the proliferation of 148biotechnology patents but finding the effect on research See Walsh et al., Research Tool at 286-89. 24
  • 134. Bo x 3-5 . Can Patents Deter Innovation? The Anticomm ons in Biomedical Research Michael A. Heller and Rebecca S. Eisenberg, Science 1998 May 1; 280: 698-701. The tragedy of the anticommons refers to a problem that might arise when multiple owners each have a right to exclude others from a scarce resource, and no one has an effective privilege of use. There are two mechanism s by which a go vernm ent might inadvertently create an antico mmons: (1) by creating too many con current fragments of intellectual property rights in potential future prod ucts; (2) by perm itting too many existing patent owners to stack licenses on to p of the future discoveries of users. The authors theorize that patenting of gene fragments and of receptors useful for screening potential pharmac eutical p roducts are two situations in wh ich too many c oncurrent fragments may result in an anticommons. If a tragedy of the anticommons were to emerge, it might endure because of the transaction costs of rearranging entitleme nts, hetero geneous interests of o wners, and cognitive biases amo ng researche rs, the authors suggest. The authors suggest that policy-makers should seek to ensure coherent boundaries of existing patents and to minimize restrictive licensing practices that interfere with product development. Otherwise, they conclude that more patent rights may lead paradoxically to fewer useful products for improving human health.to innovation from occurring.149 Another in this field are sophisticated enough tofactor that mitigated anticommons concerns, understand that,” he argued.153the authors noted, is the very high number oftechnological opportunities in the b. Access to Existing Technologiesbiotechnology industry, which enables firms Needed for Follow-On Innovationto redirect their research efforts to areas lessencumbered by patent claims to avoid There is a debate among scholars aspossible infringement issues.150 to the optimal balance of incentives to innovate between parties engaged in initial Some panelists expressed views research and parties engaged in follow-onsimilar to these findings.151 One panelist research. Some contend that broad patentscommented, for example, that licensors tend maximize innovation by enabling the initialto be “fairly sensitive” to the implications of inventor to coordinate future follow-onroyalty-stacking for product R&D.154 Others contend that restrictedcommercialization.152 “If the licensor . . . is access to patents - especially broad patents -about to propose a royalty that’s going to kill on discoveries such as research tools canthe product, [the licensor] is not going tomake any money. And most of the players 149 See id. at 331, 333-34 (Although thesemechanisms may prevent projects from being stopped, 153 Id. But cf. Janice M. Mueller, No “Dilettantethese scholars cautioned that they impose social costs, such Affair:” Rethinking the Experimental Use Exception toas time delays and distraction from research.). Patent Infringement for Biomedical Research Tools, 76 WASH. L. REV. 1, 7, 57 (2001) (arguing that “[t]he royalty 150 Id. at 304, 331-32. stacking problem in biotechnology . . . has escalated in severity”). 151 See Blackburn 2/26 at 314-15; Beier 2/26 at312-13; Seide 3/19 at 189; Dreyfuss 7/10 at 62 154 See, e.g., Edmund W. Kitch, The Nature and Function of the Patent System, 20 J. L. & ECON. 265 152 See Blackburn 2/26 at 315. (1977). 25
  • 135. harm follow-on innovation.155 existing legal exemption,” inventing around patents, using the technology offshore, or In their business survey of the seeking to invalidate the patent, butbiotechnology industry, Professors Walsh, cautioned that many of these mechanismsArora, and Cohen evaluated whether the can impose social costs.160later possibility has arisen. They concludedthat there is no evidence that biotechnology E. Licensing Practices forresearch has been significantly impeded. Biotechnology ResearchNevertheless, “the prospect exists and Toolsongoing scrutiny is warranted.”156 Theynoted that access restrictions that harm The panelists discussed two licensinginnovation are most likely to occur when a arrangements that have been used in theresearch tool will be used primarily to biotechnology industry to provide firms withdevelop innovations that will compete with access to research tools: reach-throughone another in the marketplace, and the license agreements and patent pools. Theyresearch tool is potentially key to progress in also offered some observations on the meritsone or more therapeutic areas.157 In such of exclusive licensing of research tools.circumstances, the patent holder may seekeither to develop the technology itself or 1. Reach-Through Licenseexclusively license it to another.158 Given Agreementsthat multiple technologies may require theuse of such a research tool to foster further Reach-through license agreementsinnovation, the authors saw such a (RTLAs) are a form of licensing agreementdevelopment as likely to retard used by patent owners that hold rights on ainnovation.159 These scholars also observed biotechnology research tool, or otherthat mechanisms to mitigate such harm to upstream areas of research, to share in theinnovation exist, such as “invoking a value of the discoveries by licensees.‘research exemption’ that is broader than the Typically, RTLAs establish royalty obligations measured as a percentage of 155 See, e.g., Merges & Nelson, 90 COLUM. L. sales of the licensee’s product. Usually,REV. 839; Frederick M. Scherer, The Economics of Human however, the licensee of the research toolGene Patents, 77 ACADEMIC MEDICINE 1348 (2002). For does not need access to the research tool tofurther discussion of these issues, see supra Ch. 2(III). make or sell its product. Rather, the licensee 156 Walsh et al., Research Tool at 331. uses the research tool only to identify and develop the product.161 By letting eventual 157 Id. at 333. The authors cite stem cell market results determine the amount oftechnology as an example of a technology to which a patentholder might prefer to restrict access. Id. See also Cohen royalties paid, RTLAs potentially are a10/30 at 94-95 (discussing Geron’s incentives to limit means to overcome some of theaccess to embryonic stem cell technology). uncertainties and valuation disputes that may 158 Walsh et al., Research Tool at 333. 160 159 Id. at 290-91, 333 (arguing that “no one firm Id. at 324, 334-35.can even conceive of all the different ways that the 161discovery might be exploited. . . .”). See supra Ch. 3(III)(D)(1)(c). 26
  • 136. impede efficient licensing, as discussed risk” of creating an anticommons bysupra in Chapter 2. 162 fostering royalty stacking.168 Another panelist expressed concern that, by One panelist identified two ways in “demanding royalties on the sale of awhich reach-through license agreements for product that is not covered by their patent,”research tools can promote competition and a licensing company could be violating theinnovation. First, they can facilitate access patent misuse and antitrust laws.169 Thisto a wide range of research tools by reducing panelist stated that it is unclear how antitrustthe up-front licensing costs.163 This access is would weigh the competitive effects of theseparticularly important in the context of the types of arrangements and suggested thatbiotechnology industry, which includes additional guidance by the Agencies may bemany small and yet-to-be-profitable firms.164 necessary to provide certainty surroundingSecond, RTLAs may facilitate risk-sharing the use of RTLAs.170between the tool owner and the licensee.165One panelist suggested that RTLAs might 2. Patent Poolsplace too much risk on the licensor, becausethe research tool may prove useful in the Patent pools involve “patents [frominitial stages of R&D, but the potential multiple patentees being] licensed in aproduct ultimately might fail in the clinical package, either by one of the patent holderstrial phase, thereby denying the tool owner or by a new entity established for thislicensing fees.166 Such risk-allocation purpose, usually to anyone willing to pay theissues, however, might be resolved through associated royalties.”171 A biotechnologyadjustments to the pricing levels in trade association stated that voluntary patentRTLAs.167 pools are “one of the important potential solutions to concerns regarding overlapping Other panelists identified potential patents.”172 Indeed, this participant notedways in which RTLAs might harm approvingly the paper released by the PTOcompetition and innovation, and noted entitled “Patent Pools: A Solution to theuncertainty surrounding the antitrust analysis Problem of Access in Biotechnologyof these agreements. One panelist Patents?,” which discusses the use of patentcontended that RTLAs present a “severe pools as a means of fostering access to 162 According to one panelist, RTLAs tend to beused “in more unique tool technology” rather than 168 See Kirschner 2/26 at 311.“fungible research tools.” See Blackburn 2/26 at 315. 169 Earp 2/26 at 270. 163 See id. at 275. 170 See id. at 272-73, 327-28. For further 164 See Beier 2/26 at 265; Blackburn 2/26 at 275- discussion of RTLAs under the antitrust laws, see Second76; BIO (stmt) 2. Report (forthcoming). 165 See Blackburn 2/26 at 275. 171 Shapiro, Navigating the Patent Thicket at 127. For antitrust treatment of patent pools, see Second 166 See Oehler 2/26 at 278. Report (forthcoming). 167 172 See Blackburn 2/26 at 279. BIO (stmt) 12. 27
  • 137. patented research tools.173 rather than exclusive, licenses. One panelist explained that firms prefer to grant non- The OECD, however, has questioned exclusive licenses on their research tools,whether industry participants can solve the because it is impossible to know in advancetransaction cost problems that arise in whether any particular licensee will succeedmarkets for genetic inventions by forming in bringing a product to market.175 Hepatent pools.174 It noted that these suggested that when the patentee can profittechnologies are fundamentally different from the exploitation of a research tool, thefrom the electronics sector, in which patent incentives exist to drive the broadpools are used more frequently because of dissemination of the particular tool.176 Hethe importance of standards and did, however, note that there “are probablyinteroperability. examples of tools that maybe are appropriately exclusively licensed” and3. Non-Exclusive Licensing of suggested that the market for potential Patented Research Tools genomic cancer targets might be such a market.177 Two of the panelists observed thatowners of patented research tools generally Another panelist cited an example tohave the incentive to grant non-exclusive, demonstrate the potentially adverse implications for a business of exclusive 173 licensing: in a market with two competitors United States Patent and Trademark Office,Patent Pools: A Solution to the Problem of Access in over the provision of genomic databaseBiotechnology Patents? 3 (2000), at information, one of the companies gave anhttp://www.uspto.gov/web/offices/pac/dapp/opla/patentpoo exclusive license to its database to a largel.pdf. pharmaceutical company. The direct 174 Organisation for Economic Co-operation and consequence of this exclusive license was toDevelopment, Genetic Inventions, Intellectual Property force the other large pharmaceuticalRights and Licensing Practices: Evidence and Policies 67(2002) (“It is true that there is a growing interdependence companies to seek nonexclusive access toamong patents, that the claims of many patents are the rival firm’s database.178 This panelistnarrower, and that patents are held by multiple owners. noted that the economics of licensingLicensing transaction costs are burdensome and freedom ofoperation is restricted, thus increasing the potential for databases or research tools dictate thatconflict among researchers. However, the pharmaceutical companies license on a nonexclusive basis,biotechnology industry may be fundamentally different because it is not possible to build a businessfrom the electronics sector. It is not an industry in whichdefining standards is important, and assuringinteroperability of technologies is not very important,especially not in the development of therapeutics. Acompany’s worth is tightly tied to its intellectual property 175and fosters a ‘bunker mentality.’ There are likely to be See Blackburn 2/26 at 264.disagreements among partners over the value of the 176different patents in a pool, and dominant players may not See id. at 265.have a strong incentive to join the pool. If a limited field of 177application and essential patents can be defined, the patent See id. at 264 (noting that his company haspool model is worthy of consideration in biotechnology identified so many potential genomic cancer targets that(Marks et al., 2001). The suitability of the patent pool for supply exceeds demand, and licensees can insist onbiotechnology patents certainly requires further study, as exclusive licenses).does the role of government in promoting them.”), at 178http://www.oecd.org/dataoecd/42/21/2491084.pdf. See Bendekgey 2/26 at 268-69. 28
  • 138. around a single customer.179 discrete. Biotechnology patents might harm follow-on innovation through the creation ofF. Conclusion an anticommons and by restricting access to inventions. A few panelists suggested that Biotechnology innovation is heavily these problems can be mitigated bydependent on the patent rights that have mechanisms such as reach-through royaltybeen available for biotechnology inventions agreements, cross-licensing, and patentsince 1980. Patents help firms to recover pools. It is also possible that recenthigh, fixed R&D costs and are particularly uncertainty about the scope of the researchuseful in enabling biotechnology companies, exemption may hinder non-commercialwhich are generally small in size, to attract research.capital investment and to contract with otherfirms for commercial development of theirinventions. This capital is critical forongoing R&D, because productcommercialization in the biotechnologyindustry is particularly time-consuming andexpensive. Patent disclosures assist theinnovation process by encouraginginformation dissemination and enabling thepublication of discoveries in the scientificliterature. Competition also encouragesinnovation, although panelists typically gavegreater stress to the role of patents. Poor quality biotechnology patentsalso have the potential to harm innovationby causing companies to avoid the field ofinquiry covered by such patents, rather thanto seek to invalidate them. Panelists statedthat litigation is too expensive and time-consuming for small biotechnologycompanies. Views varied on whether patentquality in the biotechnology field differedfrom that in other industries. Biotechnology, with its heavyinvestment in basic research and researchtools, poses more issues of cumulativeinnovation than pharmaceutical drugs, forwhich much of the innovation process was 179 See id. at 269. 29
  • 139. IV. THE COMPUTER Technology, Inc; Desi Rhoden representing HARDWARE INDUSTRIES, Advanced Memory International, Inc.; Frederick J. Telecky, Jr. representing Texas INCLUDING Instruments; Richard L. Thurston SEMICONDUCTORS representing Taiwan Semiconductor Manufacturing Company, Ltd.; Harry WolinA. Introduction representing Advanced Micro Devices, Inc.; and Gary Zanfagna representing Honeywell In the computer hardware industries, International. Two scholars, Bronwyn H.panelists reported that firms’ attitudes Hall, from the University of California,toward the role of competition and patent Berkeley, and Rosemarie Ham Ziedonis,protection in furthering innovation depends from the University of Pennsylvania, alsoon the nature of the firm. Panelists stressed participated in business perspective panelsthe importance of competition and trade on the computer hardware industry.secrecy as drivers of innovation forintegrated design and manufacturing firms B. Industry Descriptionand foundries; for specialized design firms,panelists gave greater emphasis to patents. In general terms, the computerDiscussion frequently highlighted the special hardware industries produce the physicalissues that arise in industries characterized components for computers,by incremental, cumulative innovation and telecommunications, and other informationby products requiring a great many, technology devices, such as the computerseparately held patents. Commentators, for itself, monitors, servers, routers, andexample, extensively discussed the problems scanners.180 The semiconductor industrythat patent thickets pose for innovation and produces one particular type of hardware:the licensing arrangements that firms use to the integrated circuits and discrete devicesmaneuver through such thickets to achieve that process binary data through the controlproduct commercialization. Commentators of electrical signals. Integrated circuits arealso expressed concern that patents may more commonly referred to as ‘chips’ ordeter innovation in the computer hardware ‘processors.’industries as a result of hold-up strategies byfirms unconstrained by litigation concerns. The panelists discussed various types of firms that drive innovation in these The panelists who represented industries: specialized design firms,computer hardware firms at the Hearings integrated firms, and semiconductorwere Robert Barr representing Cisco foundries.181 Both specialized design firmsSystems, Inc; George B. Brunt representing and integrated firms engage in R&D, butAlcatel USA; Peter N. Detkin representingIntel Corporation; Stephen P. Foxrepresenting Hewlett-Packard Company; Les 180 “Hardware” is a general term thatHart representing Harris Corporation; Julie distinguishes the physical aspects of computers and relatedMar-Spinola representing Atmel devices from “software,” which is the intangible aspect that controls hardware through programs.Corporation; Daniel McCurdy representingThinkFire; Joel Poppen representing Micron 181 See Ziedonis 3/20 at 11, 16. 30
  • 140. they differ in terms of the ownership of advanced manufacturing facilities canmanufacturing facilities. Specialized design become obsolete in less than five years, andfirms, which emerged in the 1980s,182 that less advanced facilities become obsoletecontract with semiconductor foundries183 to even more quickly.186have their products manufactured; integratedfirms own their manufacturing facilities.184 C. The Role of Competition inOne panelist observed that the emergence of Spurring Computerindependent semiconductor foundries (or Hardware Innovation“contract manufacturers”) “enabled thecreation and proliferation of a new Panelists representing integratedgeneration of semiconductor companies - the firms, foundries, and hardware companiesfabless semiconductor company.”185 observed that competition drivesPanelists reported that manufacturing innovation.187 Similarly, the business surveyfacilities cost at least two billion dollars to of Cohen, Nelson, and Walsh shows thatconstruct, and the construction of the most obtaining lead-time over rivals, which is aadvanced facilities can cost in excess of four function of the competitive process, is onebillion dollars. They also stated that more of the two key mechanisms for ensuring appropriability of returns on R&D 182 See Jeffrey T. Macher et al., Semiconductors, investments in the semiconductor industry.in U.S. INDUSTRY IN 2000: STUDIES IN COMPETITIVE The other mechanism is trade secretPERFORMANCE 247 (1999), at protection.188http://www.nap.edu/books/0309061792/html/245.html.Hall and Ziedonis, in their business survey of the effects ofstrengthening patent rights on firms in the semiconductorindustry, attribute the emergence of specialized design 186 See Poppen 2/28 at 683; Thurston 3/20 at 29;firms to the strengthening of patent rights in the 1980s. Ziedonis 3/20 at 16, 83; Hall & Ziedonis, 32 RAND J.Bronwyn H. Hall & Rosemarie Ham Ziedonis, The Patent ECON. at 110.Paradox Revisited: An Empirical Study of Patenting in theU.S. Semiconductor Industry, 1979-1995, 32 RAND J. 187 See Detkin 2/28 at 751 (stating that “the clearECON. 101,104 (2001) and Box 3-6. A similar version of driving force behind innovation is competition”); Poppenthis study is available in draft under the name The Effects 2/28 at 750; Fox 2/28 at 757; Barr 2/28 at 674-77; Bruntof Strengthening Patent Rights on Firms Engaged in 3/20 at 91; Thurston (stmt) 9. For discussion of theCumulative Innovation: Insights from the Semiconductor changing nature of competition in the semiconductorIndustry, June 2001, at industry, see Peter C. Grindley & David J. Teece,http://emlab.berkeley.edu/users/bhhall/papers/HallZiedonis Managing Intellectual Capital: Licensing and Cross-01%20libecap.pdf. Licensing in Semiconductors and Electronics, 39 CAL. MGMT. REV. 8, 27-29 (1997). 183 Foundries are referred to as wafer fabricationfacilities, or “fabs” for short, in the semiconductor industry. 188 See W. M. COHEN ET AL., PROTECTING THEIR INTELLECTUAL ASSETS: APPROPRIABILITY CONDITIONS AND 184 See Ziedonis 3/20 at 17. WHY U.S. MANUFACTURING FIRMS PATENT (OR NOT) (National Bureau of Econ. Research Working Paper No. 185 Richard L. Thurston, Opening Statement of 7552, 2000), at http://papersdev.nber.org/papers/w7552Dr. Richard L. Thurston, Vice President and General (hereinafter COHEN ET AL., PROTECTING THEIRCounsel, Taiwan Semiconductor Manufacturing Company INTELLECTUAL ASSETS); Rosemarie Ziedonis, The Role of(3/20/02) 3, at Patents in Semiconductors: Insights from Two Recenthttp://www.ftc.gov/opp/intellect/020320richardthurstonstat Studies (3/20/02) (slides) at 2, atement.pdf (hereinafter Thurston (stmt)); see also Thurston http://www.ftc.gov/opp/intellect/020320rosemarieziedonis.3/20 at 10 (noting that Taiwan Semiconductor pdf. Trade secrecy is discussed below. Patents wereManufacturing Company has contracted with over 175 considered relatively unimportant for securing returns tofabless companies). innovation in the hardware industry. 31
  • 141. The representative of one hardware example, that trade secrecy is useful in thecompany stated that between 1984 and 1993, early stages of innovation.195the first 10 years of the company’sexistence, it filed only one patent, which Other panelists discussed how theyissued in 1992.189 Yet by 1994, “the choose between the use of trade secretcompany had grown to over a billion dollars protection and patents as means to protectin annual revenue. This growth was their inventions. They stated that firmsobviously not fueled by patents, it was consider whether they could detect patentfueled by competition and by open, infringement.196 Disclosure of an inventionnonproprietary interfaces.”190 Another due to patent requirements may simplypanelist stated that “competition is what enable rival firms to copy the inventiondrives . . . innovation; patents have almost without the patentee being able to detect andnothing to do with innovation.”191 sue for patent infringement.197 BecauseSimilarly, a third panelist noted that manufacturing processes cannot easily be“innovation is driven by competition in all observed by rivals, trade secrecy isof our markets.”192 particularly important for foundries and the manufacturing facilities of integratedD. Alternative Means of firms.198 Panelists observed that holders of Fostering Innovation trade secrets risk losing access to their technologies, however. Should a rival The panelists representing integrated company obtain a patent on an invention forfirms and foundries identified trade secrecy which a company had used trade secretas an important mechanism for protecting a protection, the patentee could successfullycompany’s investment in innovation.193 sue the company that used trade secretSome panelists expressed the view that trade protection for patent infringement, despitesecret protection is a supplement to patent its having discovered the invention earlier.199protection in the sense that the two are usedin different factual contexts, rather than as One panelist noted that reliance onsubstitutes to be used in the same trade secrecy could harm competition andcontexts.194 One panelist suggested, for innovation by stifling the flow of 189 This panelist represented Cisco Systems. 190 195 Barr 2/28 at 673-74. See Brunt 3/20 at 47. 191 196 Rhoden 2/28 at 754. See, e.g., McCurdy 3/20 at 49-50, 53; Thurston 3/20 at 30, 47-48; Detkin 2/28 at 665. 192 Zanfagna 3/20 at 90. 197 See McCurdy 3/20 at 49-50. 193 See Thurston 3/20 at 29-30, 47-8; Wolin 3/20 198at 51; Ziedonis 3/20 at 52; McCurdy 3/20 at 53; Brunt 3/20 See Thurston 3/20 at 30, 47-48.at 26, 46-47; Detkin 2/28 at 666; Barr 2/28 at 756 and10/30 at 79-80. 199 See id. at 47; McCurdy 3/20 at 49; MERGES & DUFFY , PATENT LAW AND POLICY: CASES AND MATERIALS 194 See Ziedonis 3/20 at 52; McCurdy 3/20 at 53; at 463 (explaining that trade secrets do not serve as priorBrunt 3/20 at 47. art). 32
  • 142. information to the public domain.200 includes the use of patents as bargainingAnother panelist, however, questioned chips in cross-licensing negotiations.205 Itwhether patents yield significantly better thereby reflects the strategy identified byresults, asserting that the disclosure of economic analysts of using the prospect ofinformation through patents is seldom mutually assured destruction to achievesufficient for a rival to replicate the detente, as discussed supra in Chapter 2.innovation.201 That panelist viewed thefrequent inclusion of trade secret 1. The Role of Patents in Spurringinformation in modern patent licenses to Innovationfacilitate the licensee’s harnessing of thetechnology as evidence of the uninformative A number of representatives ofnature of patent disclosures.202 integrated firms, foundries, and hardware companies testified that patents areE. The Implications of Patent necessary for innovation, and thus they Protection for Innovation obtain patents for offensive reasons.206 One panelist stated, for example, that the The panelists differed on how patents prevention of free riding is their primaryaffect innovation; differences depended on motivation for obtaining patents; three otherwhether patents fulfilled offensive or reasons are to negotiate cross-licenses, todefensive purposes.203 Although the terms obtain freedom to operate, and to generatedo not have a precise definition, “offensive revenue through licensing.207 Anotherpatenting” generally means obtaining patents panelist contended that, although patents areto appropriate returns in R&D; it can require necessary to prevent free riding, the numberthe patent to be enforced through of patents in the semiconductor industry farlitigation.204 In this sense, the term is exceeds any requirement for that purpose.208synonymous with the traditional economic He pointed to the pharmaceutical industry asjustification for the patent system. an example of one in which only a few“Defensive patenting” is primarily motivated patents cover each product, yet heby a desire to ensure freedom to operate and considered free riding to be successfully 200 See Brunt 3/20 at 46. 205 Cross-licensing is discussed below in the 201 See McCurdy 3/20 at 53; see also Barr 2/28 at context of patent thickets. Obtaining freedom to operate755-56 (“it’s been my experience in my practice, not just and patent mining are discussed below in the context ofwith Cisco, that I’ve actually never met an engineer that hold-up.learned anything from a patent”). But see Telecky 2/28 at754 (finding patent disclosures “a source of ideas”). 206 See Thurston (stmt) 5; Fox 2/28 at 753; Barr 2/28 at 678, 755; Brunt 3/20 at 23-24. 202 See McCurdy 3/20 at 38, 53. 207 See Fox 2/28 at 753. 203 See e.g. Detkin 2/28 at 751. 208 See Barr 2/28 at 678 (stating that, in an ideal 204 See Teece 2/27 at 507; David J. Teece, IP, world, to prevent copying in the semiconductor industryCompetition Policy, and Enforcement Issues (2/27/02) “we’d need probably one or two or three for each product(slides) at 8, at on the key features, and that’s what I think you’ll find inhttp://www.ftc.gov/opp/intellect/020227davidjteece.pdf. [the pharmaceutical and medical devices] industries.”). 33
  • 143. prevented.209 2. The Potential for Patents to Impede Innovation Specialized design firms typicallyobtain patents for offensive purposes. a. Patent Thickets in the ComputerAccording to Professor Ziedonis, patents are Hardware Industriescritical business assets for design firms, andare used in a manner consistent with how the None of the panelists disputed thepatent system was intended to operate.210 existence of densely overlapping patentSuch firms seek “very strong, solid patent rights (i.e., a patent thicket) in the computerprotection” for two reasons: to raise venture hardware industries. One panelist stated thatcapital and to stake out proprietary positions more than “90,000 patents generally relatedprimarily against other niche market rivals, to microprocessors are held by more thanbut also against integrated firms.211 10,000 parties.”214 Likewise, he reported, there are approximately 420,000 Professor Ziedonis noted two semiconductor and systems patents held bydifferences about the patenting behavior of more than 40,000 parties.215 This panelistspecialized design firms when compared to observed that the number of patents onthat of integrated firms, foundries, and semiconductor-related inventions hashardware companies. First, the rate at which increased to the point where there is anspecialized design firms are enforcing their “unavoidable overlap” of intellectualpatent rights is high. Four out of every property. 216hundred patents issued to specialized designfirms are enforced through a court action, Panelists discussed three reasons forwhich is a “very, very high number relative the emergence of patent thickets in theto other industries and within the computer hardware industries: (1)semiconductor industry.”212 Second, as the incremental innovation due to the nature ofrevenue of specialized design firmsincreases and the companies mature,attitudes toward patenting shift, so that suchfirms begin to patent more defensively andto increase their patent portfolio size, shenoted.213 214 Detkin 2/28 at 667-68 and Peter N. Detkin, A Semiconductor Patent Survey (2/28/02) (slides) at 5, at http://www.ftc.gov/opp/intellect/020228peterndetkin.pdf (hereinafter Detkin Presentation). 209 See Barr 2/28 at 678. 215 Detkin 2/28 at 667-68 and Detkin 210 Presentation at 5. See Ziedonis 3/20 at 19. 216 211 Detkin 2/28 at 668 (“there’s an unavoidable Id. at 17-18. overlap of IP. . . people are tripping over each other’s patents right and left”); see also Barr 2/28 at 677; Macher 212 Id. at 18 (observing, however, that specialized et al., Semiconductors at 281. Commentators havebiotechnology firms exhibit a similar high rate of patent described the computer hardware industries as primeenforcement). examples of “complex product industries,” in which relatively numerous patents protect individual 213 See id. commercializable products. See, e.g., Cohen 2/20 at 30. 34
  • 144. Bo x 3-6 . The Patent Paradox Revisited: An Em pirical Study of Patenting in the US Semiconductor Industry, 1979-1995 Bronwyn H. Hall and Rosemarie Ham Ziedonis, RAND Journal of Economics, Vol. 32, No. 1, Spring 2001, pp 101 -128 . Bro nwyn H . Hall and R osem arie H am Ziedonis conducted an emp irical study of patenting practices in the semiconductor industry in order to explain a paradox in the economic literature: the patenting rate per R&D dollar doubled in the semiconductor industry since the mid-1980s, while other economic studies indicated that industry participants did not regard patents as an important means for recouping investments in innovation. The study was based on a com bination of qualitative and quantitative research methods. The qualitative analysis involved the authors conducting interviews with intellectual property managers and executives from several U.S. semiconductor firms. The quantitative analysis involved the authors compiling a database of the paten t portfo lios of 100 p ublicly traded U.S. semico nduc tor firms w hose R& D expenditure s were prima rily focused on semiconductor-related areas from 1975 to 1998. They matched these data with financial and other variab les to formulate estimates of the patent propensities of individual firms d uring the perio d of the study. The authors concluded that the significant increase in patenting per R&D dollar was attributable to the strengthening of patent rights in the United States, which spurred “patent portfolio races” among capital-intensive firms. Firms were engaged in these races to reduce concerns about “being held up by external patent owners and at nego tiating acc ess to external techno logies o n more favo rable terms.”the underlying technology; (2) the rise of which a patent thicket has developed.218defensive patenting; and (3) the ease ofobtaining patents at the PTO. (ii). The Rise of Defensive Patenting(i). Incremental Innovation and the As discussed above, firms in the Nature of Hardware and computer hardware industries have been Semiconductor Technology obtaining patents at rapidly increasing rates largely for defensive purposes. The Four industry representatives likelihood of firms holding overlappingtestified that the technology developed by intellectual property increases as morethe hardware and semiconductor industries patents issue over semiconductor andis susceptible to the creation of patent hardware innovations. In this way, thethickets, because hardware and problem is self-perpetuating. As onesemiconductors contain an incredibly large panelist acknowledged, “the only practicalnumber of incremental innovations.217 The response to this problem of unintentionalcomplex nature of computer hardware and sometimes unavoidable patenttechnology is one factor that contributes to infringement is to file hundreds of patentsthe existence of a technology thicket over each year ourselves.”219 In their research, Professors Hall and 217 See Detkin 2/28 at 669-70, 710-11; Poppen2/28 at 684, 712; Barr 2/28 at 713-14; Fox 2/28 at 714.Their testimony offered confirmation of similar 218observations by academic panelists. See, e.g., R. Levin See Teece 2/27 at 500.(stmt); Lemley 2/25 at 37 (noting the cumulative nature of 219semiconductor innovation). Barr 2/28 at 677; see also Hart 4/9 at 42-42. 35
  • 145. Ziedonis identified a “pro-patent” shift in Texas Instruments’ business – promptedthe US legal environment in the 1980s as the other firms also to commence patent miningstimulus for the rise of defensive programs.224patenting. 220 The authors believe that thisshift resulted from a series of congressional (iii). Ease of Obtaining Patentsreforms in the early 1980s, including thecreation of the Court of Appeals for the Professor Ziedonis contended thatFederal Circuit, which “put in place a the ease of obtaining patents at the PTO,number of procedural and substantive rules although not the sole cause of the thicket, isthat collectively strengthened the rights of a contributing factor.225 She cited interviewsUS patent owners.”221 conducted with participants in the semiconductor industry in which the Professors Hall and Ziedonis also participants stated that the standard foridentified two events that arose out of the obviousness should be increased so as to“pro-patent” shift and signaled the prevent “very trivial inventions” beingimportance of the new patent regime to patented by the PTO.226firms in the semiconductor industry. First,Polaroid’s successful patent infringement b. The Potential for Patent Thickets tosuits against Kodak resulted in Polaroid Harm Innovationbeing “awarded almost $1 billion indamages and Kodak . . . [being] barred from The panelists discussed several wayscompeting in the instant-film camera in which patent thickets can harmbusiness.”222 This case created a fear among innovation.227 First, the need of integratedfirms that owned manufacturing facilities firms and hardware companies to developthat the “courts were willing to take an extensive patent portfolios for defensiveaggressive stance against infringement by purposes diverts funding from R&D into thehalting – either temporarily or permanently – obtaining of patents. As one panelistproduction utilizing infringedtechnologies.”223 Second, the revenueobtained by Texas Instruments from mining 224 Hall & Ziedonis, 32 RAND J. ECON. at 109.its patents – that is, seeking patent royalties Panelists reported that some companies have sought tofrom firms that operate outside the range of license their patents to companies that operate outside the market of the patent holder, because a higher royalty can be extracted due to an imbalance in bargaining positions. See Brunt 3/20 at 25; Poppen 2/28 at 684; Thurston 3/20 at 34. In this situation, one panelist contended, the management 220 Hall & Ziedonis, 32 RAND J. ECON. at 105. of a company treats patents as an asset that must generate a return, instead of as a means to exclude parties from a 221 Hall & Ziedonis, The Effects of Strengthening particular invention. See Wolin 3/20 at 81. See also infraPatent Rights on Firms Engaged in Cumulative Ch. 3(IV)(E)(2)(c)(i).Innovation: Insights from the Semiconductor Industry at 22512. See Ziedonis 3/20 at 15-16. 222 226 Hall & Ziedonis, 32 RAND J. ECON. at 109. Id. 223 227 Id. A number of panelists discussed the threat Another potential harm, resulting from theof an injunction. See, e.g., Poppen 2/28 at 686, 691, 725; strategic use of patents in licensing negotiations, isDetkin 2/28 at 722-23; Barr 2/28 at 723. addressed in the next section. 36
  • 146. observed, “the time and money we spend on c. The Strategic Use of Patents inpatent filings, prosecution, maintenance, Licensing Negotiationslitigation and licensing could . . . be muchbetter spent on product development and Panelists discussed the strategic useresearch leading to more innovation.”228 of patents in licensing negotiations, and in particular one type of strategic use, generally Patent thickets can reduce follow-on known as “hold-up.”233 They discussedinnovation by requiring an innovator to seek hold-up as enabled by sunk costs that a firmlicenses from multiple patentees.229 In these already has invested in product developmentindustries, one panelist reported, “hundreds, or manufacturing, before learning of thethousands of patents cover a single patent, which in turn enable the patentee toproduct.”230 As discussed supra in Chapter demand royalties higher than it could have2, the transaction costs and potential for sought before the firm sunk its costs; with soroyalty stacking involved in obtaining very many patents at issue, panelistsmultiple licenses from numerous patent suggested, infringing someone’s patent mayholders may pose obstacles to the be inevitable, but there may be nodevelopment of follow-on technologies.231 economically feasible way, prior to making sunk investments, to identify and obtain Patent thickets also can harm rights to all the relevant patentedinnovation by creating uncertainty, which technologies.234 Some commentators argueaffects investment decisions. One panelist that hold-up in this sense harms competitionstated that the proliferation of patents and and innovation.235 Others suggest that suchpatent-related litigation has created behavior constitutes a legitimate exercise of“pervasive uncertainty about legal rights . . . a patentee’s right to exclude.236[that] heightens risks surrounding innovationinvestment decisions . . . [and] is withoutdoubt a serious drag on the technological 233 For discussion of hold-up for antitrustand scientific progress that the patent system enforcement purposes, see Timothy J. Muris, The FTC andwas designed to promote.”232 the Law of Monopolization, 67 ANTITRUST L.J. 693, 704 (2000); Benjamin Klein, Market Power in Franchise Cases in the Wake of Kodak: Applying Post-Contract Hold-Up Analysis to Vertical Relationships, 67 ANTITRUST L.J. 283 (1999). 228 Barr 2/28 at 677-78. Similarly, anotherpanelist contended that “patents are assets that suck money 234 See, e.g., Barr 2/28 at 677; Detkin 2/28 atout of the system.” Brunt 3/20 at 25. 764. 229 See Shapiro, Navigating the Patent Thicket at 235 See Shapiro, Navigating the Patent Thicket at120-121. 124-26; see also supra Ch. 2(III)(C)(2) and infra Ch. 3(IV)(E)(2)(c)(iii). 230 Poppen 2/28 at 684. 236 See generally Frederick J. Telecky, Statement 231 See Detkin 2/28 at 764 (noting the presence of Frederick J. Telecky, Jr., Senior Vice President andof “half a million patents owned by 40,000 parties . . . and General Patent Counsel, Texas Instruments: FTC/DOJwe have to worry about how we’re going to negotiate with Hearings on “Competition and Intellectual Property Lawthem”); Poppen 2/28 at 690 (raising royalty stacking and Policy in the Knowledge-Based Economy” (2/28/02) 5concerns). (“refusal to license is at the heart of the patent system”), at http://www.ftc.gov/opp/intellect/020228telecky.pdf 232 Fox 2/28 at 696; see also Barr 2/28 at 675-76. (hereinafter Telecky (stmt)). 37
  • 147. (i). The Rise of Non-Practicing Entities In their business survey, ProfessorsHall and Ziedonis concluded that The potential for hold-up to result insemiconductor firms with large sunk costs in mutually assured destruction means firmscomplex manufacturing facilities started to actively participating in the industry – patentpatent defensively in the 1980s to reduce, practicing entities (PPEs) – are unlikely toamong other things, “concerns about being employ this strategy against each other.241held up by external patent owners.”237 These Panelists, however, identified firms referredconcerns stemmed in part from Polaroid’s to as non-practicing entities (NPEs) that cansuccessful patent infringement suit against successfully employ a hold-up strategyKodak.238 One industry participant without fear of retaliation.242 NPEs obtaininterviewed by Professors Ziedonis and Hall and enforce patents against other firms, butstated, “a preliminary injunction would be either have no product or do not create ordetrimental to a firm if it means shutting sell a product that is vulnerable todown a high-volume manufacturing facility; infringement countersuit by the companyloss of one week’s production alone can cost against which the patent is being enforced.millions of dollars.”239 Firms in the As discussed supra in Chapter 2, MADcomputer hardware industries responded to strategies to mitigate hold-up will not workthe possibility of having their production against NPEs, who are not susceptible to theenjoined by accumulating large patent threat of a countersuit shutting down theirportfolios. If a rival company sought to production.243 In contrast, NPEs canemploy a hold-up strategy against them, they threaten PPEs with patent infringement andwould draw on their portfolio to assert an injunction, which, if granted, could inflictpatent infringement counterclaims against substantial losses.244that rival, resulting in what panelistsdescribed as “mutually assured destruction” Panelists identified three types ofor “MAD.”240 NPEs in the computer hardware industry: (1) non-practicing design firms, which patent their inventions but do not make or sell patented products to consumers; (2) 237 Hall & Ziedonis, 32 RAND J. ECON. at 104;see also Box 3-6; Rosemarie Ziedonis, When the Giants’Shoulders are Crowded: Fragmented Rights and Patent 241Strategies in Semiconductors 4 (July 2002) in draft at See Poppen 2/28 at 684-86.http://www.isnie.org/ISNIE02/Papers02/ziedonis.pdf. 242 See Rhoden 2/28 at 723-24; Carl Shapiro, 238 See Hall & Ziedonis, 32 RAND J. ECON. at Technology Cross-Licensing Practices: FTC v. Intel109. (1999), in 4 THE ANTITRUST REVOLUTION: ECONOMICS, COMPETITION AND POLICY 350, 356 (John E Kwoka, Jr. & 239 Id.; see generally John R. Boyce & Aidan Lawrence J. White eds. 2004).Hollis, Innovation, Imitation & Preliminary Injunctions in 243Patents (Public Comment), at See Poppen 2/28 at 685-89; Detkin 2/28 athttp://www.ftc.gov/os/comments/intelpropertycomments/02 671-72.05xxhollis.pdf. 244 See Poppen 2/28 at 685-89; Detkin 2/28 at 240 See Hall 2/28 at 662; Detkin 2/28 at 669; 671-72; Hall & Ziedonis, 32 RAND J. ECON. at 109. ForPoppen 2/28 at 684-85; Barr 2/28 at 713; see also Hall & additional discussion of issues raised by NPE conduct, seeZiedonis, 32 RAND J. ECON. at 109. supra Ch. 2(III)(C)(2) and Second Report (forthcoming). 38
  • 148. “professional” patent assertion companies lawsuits filed by a group of 136 companies,that buy patents from other companies, for example, involved patents not inventedparticularly those that are bankrupt, and then by the company.247 Two panelists confirmedassert them against practicing entities; and that an increasing number of companies are(3) “patent miners,” which are companies seeking to buy and sell the patent portfoliosthat assert their patent portfolios against of failing companies to assert against otherfirms outside of their business.245 firms.248 In their business analysis of licensing practices in the semiconductor and Professor Ziedonis noted that the electronics industry, Professors Grindley andnumber of cases filed by NPEs has increased Teece observe that “occasionally, firms cansince the mid-1980s, and that the sale of purchase a portfolio of patents with which topatents by failing companies has increased establish cross-licensing relationships; butsince the 1990s.246 One third of the patent quality patents often are not available in this fashion.”249 245 See Poppen 2/28 at 685-88; Detkin 2/28 at672. (ii). Hold-Up and Patent Thickets The panelists discussed two reasons for the In industries such as the computeremergence of “patent mining” by companies. First, theneed to patent defensively has forced many firms to hardware industries, where innovation isdevelop extensive patent portfolios, at considerable cost. cumulative, panelists noted that hold-up isOne business representative stated that it costs about more likely to occur, because the presence of$200,000 to maintain a patent worldwide over a period of20 years. See Brunt 3/20 at 25. Panelists reported that a patent thicket makes patent infringementsome companies have sought to offset these costs by very difficult to avoid.250 As Professorseeking to license their patents to other companies, Shapiro observed, participants in theparticularly companies that operate outside the market ofthe patent holder, because a higher royalty can be extracted semiconductor industry receive “thousandsdue to an imbalance in bargaining positions. See id.; of patents . . . each year and manufacturersPoppen 2/28 at 684; Thurston 3/20 at 34. can potentially infringe on hundreds of Second, panelists contended that business patents with a single product.”251 Anotherattitudes towards patents have changed since the 1980s. panelist stated that “the large number ofThe management of some companies, some asserted, havebegun to treat patents as an asset that must generate areturn, instead of as a means to exclude parties from aparticular invention. See Wolin 3/20 at 81. Panelists cited 247two examples to support this change in attitude. First, a See Ziedonis 3/20 at 73-74.number of panelists mentioned Texas Instruments, which 248successfully instigated a patent mining program in the late See Thurston 3/20 at 75; Wolin 3/20 at 76.1980s to save the company from bankruptcy, and thereby 249became an example to other companies of how to mine Grindley and Teece, 39 CAL. MGMT. REV. attheir patents. See Thurston 3/20 at 28-29; Wolin 3/20 at 31; see also Shapiro 11/6 at 176 (observing that “I’ve even81; Ziedonis 3/20 at 83; Telecky 2/28 at 653; Macher et al., seen a situation where a portfolio was split up and someSemiconductors at 281; Grindley & Teece, 39 CAL. MGMT. patents split off to a third party who had no otherREV. at 20. Second, a widely read book in business circles commercial interests, so they could assert it mostentitled Rembrandts in the Attic encourages manager