Georgetown.lawreviewnetwork.com

Rational Antitrust Policy and Reverse Payment
Settlements in Hatch-Waxman Patent Litigation
This Article examines the problem of “reverse payment” settlements in patent litigation under the Hatch-Waxman Act. A reverse payment settlement involves apayment from a branded pharmaceutical company to a generic manufacturer,usually in return for the generic manufacturer’s agreement to delay marketentry. Federal appellate courts, regulatory agencies, and commentators aredivided about the legality of such agreements under antitrust law. This Articleargues that the importance of product market definition has been overlookedin existing treatments of the issue. The Article develops an empirically based“Settlement Competition Index” that could be used by courts and regulatoryagencies to evaluate reverse payment settlements. A formula to calculate theSettlement Competition Index is provided and tested with hypothetical andreal-world examples. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
REVERSE SETTLEMENT PAYMENTS IN HATCH-WAXMAN LITIGATION . .
REGULATORY FRAMEWORK . . . . . . . . . . . . . . . . . . . . . . . . . . .
BRIEF OVERVIEW OF “REVERSE PAYMENT” SETTLEMENTS . . . . . . .
CASE LAW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second Circuit: Presumptively Lawful with SettlementPolicy Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Federal Circuit: Presumptively Lawful with PatentExclusionary Zone Rationale . . . . . . . . . . . . . . . . . . . .
Eleventh Circuit: Exclusionary Zone Test with MixedSignals About Presumptive Legality . . . . . . . . . . . . . . .
Sixth Circuit: Per Se Unlawful . . . . . . . . . . . . . . . . . . .
Summary of Case Law . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
* Associate Professor of Law, Seton Hall University School of Law. 2010, David W. Opderbeck. I would like to thank Charles Sullivan, Erik Lillquist, Thomas Healy, Jon Romberg, Mark Lemley,Michael Carrier, Scott Hemphill, and Robin Feldman for comments on an earlier draft of this Article.
The author also gratefully acknowledges the excellent research assistance of Kerri Lee.
PENDING LEGISLATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
EFFORTS TO THEORIZE THE REVERSE PAYMENT PROBLEM . . . . . . . . .
CURRENT PROPOSALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
TOWARDS A SETTLEMENT COMPETITION INDEX? . . . . . . . . . . . . .
Under and Overdeterrence . . . . . . . . . . . . . . . . . . . . . .
Settlement Amount as a Proxy for Patent Strength . . . . .
A PROPOSED “SETTLEMENT COMPETITION INDEX” . . . . . . . . . . . . . .
THE IMPORTANCE OF ASSESSING PRODUCT MARKET CONCENTRATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ASSESSING THE PROBABILITY OF PATENT ENFORCEMENT CALCULATING THE SETTLEMENT COMPETITION INDEX . . . . . . . . .
Hypothetical Examples . . . . . . . . . . . . . . . . . . . . . . . .
A Real-World Example . . . . . . . . . . . . . . . . . . . . . . . .
Establishing Safety Zones and Zones of Illegality . . . . .
HEIGHTENED SCRUTINY: WEIGHING REASONABLENESS . . . . . . . . .
CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
“Reverse payment” settlements present important problems for debates over intellectual property policy concerning pharmaceutical products. A reverse pay-ment settlement involves a payment from a branded pharmaceutical company toa generic manufacturer, usually in return for the generic manufacturer’s agree-ment to delay market entry.1 The context of these settlements is patent infringe- 1. See infra section I.B. Reverse payment settlements are also called “exclusion settlements” and “pay-to-delay settlements.” See also Daniel A. Crane, Ease over Accuracy in Assessing PatentSettlements, 88 MINN. L. REV. 698, 698–99 (2004) (“I have argued that such ‘reverse payments’ (alsoknown as ‘exit payments’ or ‘exclusion payments’) should not be accorded per se treatment under theantitrust laws and should be approved so long as the patentee has a strong ex ante likelihood ofsucceeding on the merits of its infringement claim and thereby excluding the infringing use from themarket.”); John B. Reiss et al., Your Business in Court: 2008–2009, 64 FOOD & DRUG L.J. 755, 774(2009) (“On June 3, 2009, in a sixteen-to-ten vote, the Commerce, Trade, and Consumer ProtectionSubcommittee cleared the bill for a ‘full markup.’ The chairman of the FTC argued that, by passing abill banning pay-to-delay settlements, consumers would save $35 billion over ten years.”) (footnoteomitted).
ment litigation under the Hatch-Waxman Act.2 The Hatch-Waxman Act allows ageneric manufacturer to challenge a branded pharmaceutical company’s patentclaims without incurring the risk of damages.3 This provision was designed tohelp weed out weak patents and facilitate competition with generics. Reversepayment settlements arguably frustrate this design by enabling the brandedmanufacturer to buy off the generic company’s challenge.
The Federal Trade Commission (FTC), in the wake of an aggressive anti- trust enforcement program, recently suggested that this issue is “a matter ofpressing national concern.”4 Four federal circuit courts of appeals have weighedin with dramatically conflicting results.5 The Sixth Circuit, for example, hasruled that most reverse payment settlements are per se unlawful,6 while theFederal Circuit, in a somewhat terse opinion, recently ruled that reverse pay-ment settlements simply represent a valid exercise of the patent owner’s exclu-sionary rights and, therefore, present no competitive concerns.7 The Departmentof Justice (DOJ) sat on the sidelines during the Bush administration but recentlyoffered a confusing and highly nuanced proposal.8 Congress is consideringsevere legislation that has attracted major lobbying attention from the brandedand generic pharmaceutical industries.9 In addition, most academic commenta-tors have argued in favor of significant restrictions on reverse payments.10 This Article argues that existing proposals have overlooked the importance of product market definition. As a result, existing proposals threaten to over-deter potentially beneficial settlements, underdeter deleterious settlements, orfail to supply meaningful guidance about which agreements might be permis-sible. This Article offers a new approach that is consistent with how courts andregulatory agencies currently assess analogous agreements between competi-tors, including exclusive intellectual property licenses and mergers. This ap-proach results in an empirically based “Settlement Competition Index,” whichcan be used to establish antitrust safety zones, zones of per se illegality, andzones in which a complete rule of reason analysis should be applied.11 2. Drug Price Competition and Patent Term Restoration (Hatch-Waxman) Act of 1984, Pub. L. No.
98-417, 98 Stat. 1585 (codified as amended at 15 U.S.C. §§ 68b–68c, 70b (2006); 21 U.S.C. §§ 301note, 355, 360cc (2006); 28 U.S.C. § 2201 (2006); 35 U.S.C. §§ 156, 271, 282 (2006) [hereinafterHatch-Waxman Act]).
3. See infra section I.A.
4. Anticompetitive Pay-for-Delay Settlements in the Pharmaceutical Industry: Why Consumers and the Federal Government Are Paying Too Much for Prescription Drugs: Hearing on H.R. 1706 Beforethe Subcomm. on Courts and Competition Policy of the H. Comm. on the Judiciary, 110th Cong. 1(2009) (prepared statement of Richard A. Feinstein, Director, Bureau of Competition, Federal TradeCommission), available at http://www.ftc.gov/os/2009/06/P859910payfordelay.pdf.
5. See infra section I.C.
6. See infra section I.C.4.
7. See infra section I.C.2.
8. See infra section I.D.
9. See infra section I.E.
10. See infra Part II.
11. Under a rule of reason analysis, “antitrust plaintiffs must demonstrate that a particular contract or combination is in fact unreasonable and anticompetitive before it will be found unlawful.” Texaco Inc.
Part I of this Article describes the regulatory context; the nature of reverse payment settlements; and existing judicial, regulatory, and legislative efforts toaddress the problem. Part II describes a range of important academic commen-tary that seeks to theorize and offer proposals for regulating these agreements.
Part III develops and defends the “Settlement Competition Index,” using bothhypothetical and real-world market examples.
I. REVERSE SETTLEMENT PAYMENTS IN HATCH-WAXMAN LITIGATION Under the Federal Food, Drug, and Cosmetic Act, no prescription drug can be marketed prior to gaining approval from the Food and Drug Administration.12For a new drug, the applicant must submit a New Drug Application (“NDA”),13which requires multiple phases of clinical trial testing for safety and efficacy.14The total cost of developing a new drug is in the hundreds of millions of dollars.15The clinical review and approval process ordinarily takes at least five to ten years.16As a result of this lengthy review and approval process, a patented drug usually doesnot reach the market until a significant portion of its patent term has expired.17 At the same time, however, potential generic competitors could not, absent a statutory exception to the Patent Act, develop a competing product that would beready to market when the patent expires.18 Moreover, without the prospect of patentrents, potential competitors are loathe to invest in the NDA process.19 As a result, the v. Dagher, 547 U.S. 1, 5 (2006). In contrast, “[p]er se liability is reserved for only those agreements thatare ‘so plainly anticompetitive that no elaborate study of the industry is needed to establish theirillegality.’” Id. (quoting Nat’l Soc’y of Prof’l Eng’rs v. United States, 435 U.S. 679, 692 (1978)).
12. 21 U.S.C. § 355(a) (2006).
13. See Gerald J. Mossinghoff, Overview of the Hatch-Waxman Act and Its Impact on the Drug Development Process, 54 FOOD & DRUG L.J. 187, 187 (1999).
14. See 21 U.S.C. § 355(b).
15. See OFFICE OF TECH. ASSESSMENT, U.S. CONGRESS, PHARM. R & D: COSTS, RISKS AND REWARDS, OTA-H-522, at 214 (1993), available at http://www.fas.org/ota/reports/9336.pdf (“Total estimatedpreclinical pharmaceutical R&D constituted approximately $450 million in 1988 . . . .”); Joseph A.
DiMasi et al., The Price of Innovation: New Estimates of Drug Development Costs, 22 J. HEALTH ECON.
151, 151 (2003) (estimating costs at $802 million).
16. Salomeh Keyhani, Marie Diener-West & Neil Powe, Are Development Times for Pharmaceuti- cals Increasing or Decreasing?, 25 HEALTH AFF. 461, 463 (2006) (finding that “[t]he median totalpost-IND development time for all drugs in the database was 6.4 years; it ranged from 2.6 to 17.3years . . . . The median clinical trial phase was 5.1 years and ranged from 1.4 years to 14.6 years, whilethe median regulatory review phase was 1.2 years and ranged from 0.3 years to 5.0 years.”).
17. See FED. TRADE COMM’N, GENERIC DRUG ENTRY PRIOR TO PATENT EXPIRATION: AN FTC STUDY 4 (2002), available at http://www.ftc.gov/os/2002/07/genericdrugstudy.pdf (last visited July 23, 2009);Mossinghoff, supra note 13, at 188.
18. See FED. TRADE COMM’N, supra note 17, at 4; 35 U.S.C. § 271 (infringement of patent).
19. See FED. TRADE COMM’N, supra note 17, at 5 (stating that the NDA process to develop a drug was costly and time-consuming). A brand company’s patent allows it to set high monopoly prices, whichthen provide rents for reinvestment. See Julio Nogue´s, Patents and Pharmaceutical Drugs: Understand-ing the Pressures on Developing Countries (Policy, Research & External Affairs, Int’l Trade, Int’l Econ.
Dep’t, Working Paper Series No. 502, 1990), available at http://www-wds.worldbank.org (search patent regime can foreclose competition even after a drug patent expires.
The Hatch-Waxman Act represents an effort to adjust patent policy in this regula- tory context.20 Under the Act, the term of a pharmaceutical patent subject to regula-tory delay can be extended for up to five years, for a total period of market exclusivityof up to fourteen years.21 A potential generic competitor, meanwhile, is exemptedfrom patent liability for activities relating to regulatory approval.22 Moreover, ageneric competitor can “piggyback” on the original manufacturer’s NDA safety andefficacy data by filing an Abbreviated New Drug Application (“ANDA”) showingthat its generic version is “bioequivalent” to the patented drug.23 As a result, genericcompetition can commence upon patent expiration, with a far lower cost of regulatoryapproval to the generic competitor.
The Hatch-Waxman framework also provides incentives for potential generic competitors to challenge drug patents before they expire. An ANDA applicant mayfile a “Paragraph IV” certification, under which the applicant certifies that the chal-lenged patent is invalid or will not be infringed by the generic version.24 The first filerof a Paragraph IV certification receives a 180-day period of generic market exclusiv-ity.25 Upon receipt of the Paragraph IV certification, the patent owner has 45 days inwhich to file a patent infringement suit against the ANDA filer.26 The Paragraph IV filer is exempt from damages for infringement so long as it has not begun to market the product.27 Thus, the Paragraph IV process changesthe ordinary risk calculus for patent litigation. The patent owner risks losingits patent, but the alleged infringer does not risk a damage award. Moreover,challenges to the patent are affirmatively encouraged by the prospect of the180-day exclusivity period.
B. BRIEF OVERVIEW OF “REVERSE PAYMENT” SETTLEMENTS In Hatch-Waxman cases, the typical reverse payment settlement involves an agreement by the generic manufacturer to refrain from marketing a generic “WPS502”). The competing non-patent holders, generic companies, would not be able to reap thiseconomic benefit, which is a deterrent to market entry.
20. Hatch-Waxman Act, supra note 2.
21. 35 U.S.C. § 156(c)(3), (g)(6)(A) (2006).
22. See id. § 271(e)(1); Merck KGaA v. Integra Lifesciences I, Ltd., 545 U.S. 193, 202 (2005).
23. See 21 U.S.C. § 355(j)(2)(A)(iv) (2006).
24. See id. § 355(j)(2)(A)(vii)(IV). This is referred to as a “Paragraph IV” certification because it falls under the fourth paragraph of the relevant statutory section. The ANDA filer also can elect to fileunder paragraphs I–III, which entail a certification that the branded manufacturer failed to file therequired “Orange Book” listing of the patent, the patent has expired, or approval is being soughteffective on a date after patent expiration. See id. § 355(j)(2)(A)(vii)(I)–(III). In fact, most ANDA filersdo not elect to file under Paragraph IV. See C. Scott Hemphill, An Aggregate Approach to Antitrust:Using New Data and Rulemaking To Preserve Drug Competition, 109 COLUM. L. REV. 629, 634 & n.13(2009).
25. 21 U.S.C. § 355(j)(5)(B)(iv); see Hemphill, supra note 24, at 634.
26. 35 U.S.C. § 271(e)(5) (2006).
27. 35 U.S.C. § 271(e)(4).
version in exchange for a monetary payment from the patent owner.28 Thelength of the marketing restriction can vary. In early examples, it ran for theentire remaining patent term, but in more recent settlements, the parties agreedto some division of the remaining patent term.29 Many settlements also involvelicenses from the generic company to the patent owner of other technology, andother ancillary provisions.30 Settlements also differ concerning whether a firstParagraph IV filer retains its 180-day market exclusivity.31 Thus, most of thesesettlements address the following four fundamental points: (1) amount ofreverse payment; (2) length of generic marketing restriction; (3) retention ofgeneric market exclusivity; and (4) ancillary licenses.32 Reverse payment settlements have been the subject of antitrust litigation in various federal courts, resulting in conflicting opinions from four differentcircuit courts of appeals. Some of these opinions stem from enforcement actionsbrought by the FTC. Others result from consolidated multidistrict privateantitrust actions brought by third-party payors, employee unions, and otherinterested parties, often following on the heels of FTC enforcement action. Asdiscussed in detail below, the Second Circuit has adopted a deferential generalpolicy in favor of settlement; the Federal Circuit has held that reverse paymentsettlements are presumptively lawful because of the legitimate exclusionarypower of patents; the Eleventh Circuit, in a series of cases, has developed a testintended to reflect the legitimate exclusionary zone of the patent; and the SixthCircuit has held reverse payments per se unlawful.
1. Second Circuit: Presumptively Lawful with Settlement Policy Rationale The Second Circuit’s current view is expressed in In re Tamoxifen Citrate Antitrust Litigation, which involved the patent covering tamoxifen, a block-buster cancer drug.33 In the infringement litigation following the generic manu- 28. See Matthew Avery, Note, Continuing Abuse of the Hatch-Waxman Act by Pharmaceutical Patent Holders and the Failure of the 2003 Amendments, 60 HASTINGS L.J. 171, 191–94 (2008)(describing varieties of settlements); Hemphill, supra note 24, at 647–56 (describing typology ofsettlements); Christopher M. Holman, Do Reverse Payment Settlements Violate the Antitrust Laws?,23 SANTA CLARA COMPUTER & HIGH TECH. L.J. 489, 494–500 (2007) (describing nature and typology ofreverse payment settlements).
29. See Holman, supra note 28, at 495–97.
30. See id. at 498–500.
31. See Avery, supra note 28, at 191–94; Hemphill, supra note 24, at 651–55.
32. See Avery, supra note 28, at 191–94; Hemphill, supra note 24, at 647–56; Holman, supra note 28, at 494–500. The Medicare Prescription Drug, Improvement, and Modernization Act of 2003(“MMA”) requires that, as of 2004, all patent litigation settlement agreements between brand-namedrug manufacturers and generic drug applicants be filed with the FTC and DOJ. Pub. L. No. 108-173,§ 1112, 117 Stat. 2066, 2461–63. Recent filings are available on the FTC’s website at http://ftc.gov/bc/healthcare/drug/index.htm (last visited June 30, 2009).
33. 466 F.3d 187, 193 (2d Cir. 2006). At the time of the settlement, the patent was owned by Zeneca.
facturer Barr’s Paragraph IV filing, the district court found the patent invalidbased on fraud when the patent holder withheld testing information from thepatent office.34 While an appeal of the ruling was pending, the parties agreed toa settlement, in which Barr would receive a $21 million payment and a non-exclusive license to sell an off-brand version of the patented drug, and Barr’ssupplier would receive payments of over $45 million.35 In return, Barr agreed todelay generic entry until after the expiration of the patent, unless the patent wassubsequently declared invalid in litigation brought by another challenger.36 Theparties also agreed to move to vacate the district court’s judgment.37 The settlement was challenged by various consumers, medical benefit provid- ers, and advocacy groups on antitrust and other grounds in multiple lawsuits,which were eventually consolidated.38 The plaintiffs alleged that, in addition tothe formal settlement terms, Barr and Zeneca had reached an informal “under-standing” that, if any other generic manufacturer attempted to market a genericversion of tamoxifen, Barr would revert to a Paragraph IV certification and seekto invoke the 180-day exclusivity period.39 Indeed, Barr claimed the exclusivityperiod as the first Paragraph IV filer against subsequent filers.40 Regardless, theother generic manufacturers were not able to enter the market because theywere held to have infringed Zeneca’s patent.41 The district court rejected the challenges to the settlement and granted a motion to dismiss the claims under Federal Rule of Civil Procedure 12(b)(6).42The Second Circuit affirmed.43 The Second Circuit noted long-standing policyin favor of settlement, including in patent and other intellectual propertycases.44 In fact, the court stated, settlement promotes the goals of patent law byfacilitating certainty, which encourages innovation.45 The court rejected the 34. Id.
35. Id. at 193–94.
36. Id.
37. Id. at 194. The settlement, in fact, was contingent on the court’s granting of this motion, which the court eventually granted. Id. In three subsequent Paragraph IV cases involving tamoxifen and othergeneric manufacturers, different courts held that the judgment of invalidity in the Barr litigation was notissue preclusive because it had been vacated. Id. at 194–95. In an unrelated case, the practice ofvacating judgments pursuant to settlement subsequently was invalidated in almost all cases by theSupreme Court, but that holding was not retroactive. See id. at 194 & n.8 (citing U.S. BancorpMortgage Co. v. Bonner Mall P’ship, 513 U.S. 18, 27–29 (1994)).
38. Id. at 196; In re Tamoxifen Citrate Antitrust Litig. (Tamoxifen II), 277 F. Supp. 2d 121, 123 39. Tamoxifen II, 277 F. Supp. 2d at 125.
40. In re Tamoxifen Citrate Antitrust Litig., 466 F.3d at 195. Barr was able to make this claim because the FDA’s a priori “successful defense” rule, under which the 180-day exclusivity period wasonly available to a filer that successfully defended an infringement lawsuit, was held invalid by severalcourts, and subsequently was modified by the FDA. See id. A labyrinthine procedural history followed,until the patent eventually expired. See id. at 196.
41. See id. at 196.
42. Id. at 197.
43. Id. at 198–99.
44. Id. at 202–03.
45. Id. at 203.
argument that the settlement allowed a “weak” patent to remain in forcebecause that argument requires an impermissible, and impossible, post hocreview of how an inherently risky and uncertain litigation process would haveconcluded.46 The court further rejected the argument that reverse payment settlements are inherently anticompetitive.47 Instead, the court concluded that reverse paymentsmake sense in the Hatch-Waxman Paragraph IV context because most of therisk in the infringement litigation is borne by the patent holder rather than thealleged infringer.48 Moreover, the court rejected the claim that reverse paymentsautomatically cross the threshold of legality if the value of the reverse paymentfar exceeds what the generic manufacturer could earn by selling a genericversion of the product.49 The patent holder presumably is making a paymentbased on the value of the patent grant.50 Thus, “the patent holder is seeking toarrive at a settlement in order to protect that to which it is presumably entitled: alawful monopoly over the manufacture and distribution of the patented prod-uct.”51 Although such settlements might protect weak patents, a truly weakpatent is likely to face challenges from multiple would-be generic entrants, andthe patent holder is unlikely to buy out all potential challengers.52 Finally, thecourt held that the terms of the settlement did not unlawfully exceed the scopeof the tamoxifen patent.53 The settlement, the court noted, permitted othermanufacturers to challenge the patent and did not restrict access to unrelated ornon-infringing products.54 The Second Circuit’s Tamoxifen opinion represents one of the most pro- settlement approaches to the reverse payment problem. There are signs, how-ever, that the Second Circuit might reconsider its views. Another reversepayment appeal is currently pending before the Second Circuit, which involveschallenges by labor unions and pharmacy chains to the $350 million settlementbetween Bayer AG and various generic manufacturers concerning the anti-bacterial drug ciprofloxacin hydrochloride (“Cipro”).55 On April 6, 2009, theSecond Circuit invited the Department of Justice to submit an amicus brief 46. See id. at 204, 212.
47. Id. at 206.
48. See id. at 207.
49. Id. at 208–09.
50. Id. at 209.
51. Id. at 208–09.
52. Id. at 212. The court noted that “[t]he point will come when there are simply no monopoly profits with which to pay the new generic challengers.” Id. 53. Id. at 213.
54. Id. at 213–14. In a dissenting opinion, Judge Pooler argued that the plaintiffs had adequately alleged a claim that the settlement agreement should have been subject to a full rule of reason analysis.
Id. at 221, 228 (Pooler, J., dissenting).
55. Ark. Carpenters Health & Welfare Fund v. Bayer, AG, No. 05-2851-cv(L) (2d Cir. July 6, 2009), summary available at Posting of Lyle Denniston to SCOTUSblog, http://www.scotusblog.com/2009/07/us-strong-stand-on-pay-for-delay-on-drugs/ (July 6, 2009, 19:54 EST).
detailing the Department’s views on the case.56 This move seems to signal thatthe Second Circuit might reconsider the position it took in Tamoxifen.
2. Federal Circuit: Presumptively Lawful with Patent Exclusionary ZoneRationale The most recent appellate decision is the Federal Circuit’s ruling in the Ciprofloxacin litigation.57 In the Ciprofloxacin settlement, the generic manufac-turer, Barr, agreed to delay generic entry until six months prior to patentexpiration in exchange for a $49.1 million payment.58 In addition, the patentowner, Bayer, agreed to supply Barr with Cipro for resale or to make quarterlypayments to Barr over a seven-year period, for a total of $349 million inquarterly payments.59 In subsequent Paragraph IV litigation against other ge-neric manufacturers, Bayer successfully defended the ’444 Patent.60 The settle-ment was challenged on antitrust and other grounds by Cipro purchasers andadvocacy groups.61 The district court granted summary judgment against theplaintiffs, and the Federal Circuit affirmed.62 The Federal Circuit held that the settlement agreements did not exceed the “exclusionary zone” of the ’444 Patent.63 According to the Federal Circuit, “theessence of the Agreements was to exclude the defendants from profiting fromthe patented invention. This is well within Bayer’s rights as the patentee.”64 Inaddition, the court cited public policy in favor of settlement and the commonpractice of settlements in patent litigation whereby the alleged infringer agreesnot to challenge the patent’s validity.65 The Federal Circuit distinguished theCiprofloxacin settlement from the settlement invalidated by the Sixth Circuit inIn re Cardizem66 because the Cardizem settlement entailed restraints on thegeneric manufacturer that exceeded the patent’s exclusionary zone.67 56. See Brief for the United States in Response to the Court’s Invitation at 1, Ark. Carpenters Health & Welfare Fund, No. 05-2851-cv(L), available at http://www.justice.gov/atr/cases/f247700/247708.pdf.
On October 15, 2008, the Federal Circuit issued an opinion upholding the reverse payment settlementin one of the Cipro cases that had been severed from the litigation in the Second Circuit. In reCiprofloxacin Hydrochloride Antitrust Litig., 544 F.3d 1323, 1327 (Fed. Cir. 2008), discussed infrasection I.C.2.
57. In re Ciprofloxacin, 544 F.3d at 1327–28. The litigation concerned U.S. Patent No. 4,670,444 (the “’444 Patent”) owned by Bayer AG and Bayer Corp. (“Bayer”), which is directed to ciprofloxacinhydrochloride (“Cipro”), an antibacterial compound. Id. 58. Id. at 1328–32 & n.5.
59. See id. at 1329 & n.5.
60. Id. at 1329. The ’444 Patent had been amended after reexamination in the Patent Office. Id.
61. Id. at 1329–30.
62. Id. at 1330, 1341.
63. Id. at 1332–33.
64. Id. at 1333.
65. Id. at 1333–34.
66. In re Cardizem CD Antitrust Litig., 332 F.3d 896 (6th Cir. 2003).
67. In re Ciprofloxacin, 544 F.3d at 1335. These included the generic manufacturer’s agreement not to relinquish its statutory 180-day exclusivity period, which would have delayed other generic entrants, 3. Eleventh Circuit: Exclusionary Zone Test with Mixed Signals AboutPresumptive Legality The Eleventh Circuit has visited the question of reverse payment settlements several times, with somewhat different results. In the first case, Valley Drug Co.
v. Geneva Pharmaceuticals, Inc., the court reviewed settlements between brandedmanufacturer Abbott Laboratories and generic manufacturers Geneva Pharmaceu-ticals and Zenith Goldline Pharmaceuticals concerning the hypertension andprostate drug Hytrin.68 Both Geneva and Zenith had filed Paragraph IV certifica-tions with respect to various Hytrin-related patents held by Abbott; Geneva wasthe first filer.69 Abbott sued Geneva for infringement.70 In the Abbott–Zenith agreement, Zenith admitted validity and infringement, and agreed not to produce an infringing drug, in return for $6 million inguaranteed payments and further payments of up to $6 million each quarter,with various contingencies.71 In the Abbott–Geneva agreement, Geneva agreed not to sell any product containing the patented compound until the patent expired, another party intro-duced a generic version, or Geneva prevailed in infringement litigation.72Geneva further agreed not to sell or transfer its 180-day exclusivity period.73 Inreturn, Abbott agreed to pay Geneva $4.5 million per month until another partymarketed a generic version or Geneva prevailed in infringement litigation.74Abbott’s patent for the crystalline form of the drug was invalidated by thedistrict court under the on sale bar.75 This decision was affirmed by the FederalCircuit.76 While Abbott’s petition for certiorari to the Supreme Court was pending, the FTC issued an order following an investigation of the Abbott–Genevasettlement.77 Pursuant to a consent order with Abbott, the FTC prohibited anysettlement provision under which the ANDA first filer would relinquish its180-day exclusivity period or that would prohibit research on a drug that is not and its agreement not to manufacture competing non-infringing versions of the drug covered by thepatent. Id. 68. 344 F.3d 1294, 1296, 1298 (11th Cir. 2003).
69. Id. at 1298–99.
70. Id. However, Abbott failed to file an infringement suit with respect to an alternate formulation for which Geneva had also filed a Paragraph IV certification. Id. at 1299. Zenith instituted court actionagainst Abbott seeking delisting of the challenged patents, and Abbott counterclaimed for infringement.
Id. 71. Id. at 1300.
72. Id.
73. Id.
74. Id. at 1300.
75. See id. at 1301. Under the “on sale” bar, that patent is invalid if the patented product was on sale in the United States more than one year prior to the date of the patent application. 35 U.S.C. § 102(b).
76. Id.
77. In re Abbott Labs., No. C-3945 (Fed. Trade Comm’n May 22, 2000) (decision and order), available at http://www.ftc.gov/os/2000/05/c3945.do.htm.
the subject of a patent infringement action.78 The FTC further prohibited anysettlement agreement entered pendente lite, unless accompanied by a stipulatedpreliminary injunction that included a provision for FTC review.79 Following the FTC’s Order, a variety of plaintiffs filed private antitrust actions, which were consolidated before a multidistrict litigation panel.80 Thedistrict court granted summary judgment in plaintiffs’ favor, finding that thesettlements constituted unlawful market allocation agreements.81 The EleventhCircuit reversed.82 According to the Eleventh Circuit, an agreement to refrain from infringing a patent cannot constitute an illegal market allocation because such an agreementwould be coextensive with the patent’s lawful exclusionary zone.83 The fact thata patent conceivably could be held invalid does not nullify this principle.84 Thecourt noted that “[p]atent litigation is too complex and the results too uncertainfor parties to accurately forecast” the outcome with any precision.85 Even alarge reverse payment does not necessarily suggest a weak patent because,“[g]iven the asymmetries of risk and large profits at stake, even a patenteeconfident in the validity of its patent might pay a potential infringer a substan-tial sum in settlement.”86 However, the court noted that if the agreement goesbeyond the legitimate exclusionary potential of the patent—for example, byrestricting sales of compounds not subject to the patent claims—those provi-sions might raise antitrust concerns.87 The case was remanded for considerationof this question.88 The Eleventh Circuit next faced the issue in Schering-Plough Corp. v. Federal Trade Commission.89 Procedurally, this case involved a challenge bythe settling parties to an FTC order invalidating settlements in two situations.90In the first case, the generic manufacturer, Upsher, agreed to delay generic entryof the challenged patent for blood pressure medication.91 Under the settlement,Schering-Plough received an exclusive license to market a cholesterol-reducingproduct and other products owned by Upsher for royalty fees of $60 millionplus milestone and percentage payments.92 The record contained evidence that,prior to the settlement, Schering-Plough had investigated and obtained substan- 78. Id. pt. II.
79. Id. pt. III.
80. Valley Drug Co., 344 F.3d at 1295–96.
81. Id. at 1301.
82. Id. at 1306.
83. Id. at 1305–06.
84. See id. at 1306–07.
85. Id. at 1308.
86. Id. at 1310.
87. See id. at 1311–12.
88. Id. at 1313.
89. 402 F.3d 1056 (11th Cir. 2005).
90. Id. at 1058.
91. Id. at 1059.
92. Id. at 1059–60.
tial valuations of the Upsher cholesterol product.93 However, the Upsher prod-uct proved to be unprofitable.94 In the second case, Schering-Plough agreed to split the remaining patent life with another generic manufacturer, ESI, which facilitated generic entry threeyears before patent expiration.95 Schering-Plough further agreed to pay ESI$5 million for legal fees, another $10 million if ESI obtained FDA approval tomarket the product by a certain date, and $15 million for licenses of unrelatedproducts owned by ESI.96 These agreements were challenged by the FTC several years after they were concluded.97 The FTC ruled that settlements in which “the generic receivesanything of value and agrees to defer its own research, development, productionor sales activities” are unlawful restraints of trade, unless the payment is$2 million or less, used solely for litigation costs, and the FTC is notified of thesettlement.98 The Eleventh Circuit rejected the FTC’s approach.99 According to the court, neither rule of reason nor per se analysis is appropriate in antitrust casesinvolving patents because, “[b]y their nature, patents create an environment ofexclusion, and consequently, cripple competition. The anticompetitive effect isalready present.”100 Instead of traditional antitrust analysis, the court created athree-part test, which “requires an examination of: (1) the scope of the exclusion-ary potential of the patent; (2) the extent to which the agreements exceed thatscope; and (3) the resulting anticompetitive effects.”101 Applied to the settle-ments at issue, the court noted that Schering-Plough’s patent appeared to bestrong, the license payments by Schering-Plough for unrelated products werenot sham “reverse payments,” and the settlement agreements had no improperanticompetitive effects.102 Concerning anticompetitive effects, the court rejected the FTC’s antipathy towards reverse payments in Hatch-Waxman cases. The court reasoned thatreverse payments should be expected in Paragraph IV cases—indeed, that“[r]everse payments are a natural by-product of the Hatch-Waxman process”— 93. Id. at 1059–60.
94. Id. at 1060.
95. Id.
96. See id. at 1060–61 & n.8.
97. Id. at 1061.
98. Id. at 1062. The FTC’s decision reversed an earlier opinion by an administrative law judge (ALJ) that the settlements were lawful. Id. at 1061–62.
99. Id. at 1065–66.
100. Id.
101. Id. at 1066 (citing Valley Drug Co. v. Geneva Pharms., Inc., 344 F.3d 1294, 1312 (11th Cir.
102. Id. at 1072–76. Concerning the Upsher settlement, the court cited the extensive record developed before the ALJ, which suggested the licenses from Upsher represented bona fide businessdecisions. Id. at 1068–71. Concerning the ESI settlement, the court was persuaded by the general policyin favor of settlements, the district court’s mediation, and subsequent approval of the settlement. Id. at1076.
because the statute permits generic manufacturers to challenge patent validitywithout incurring any significant risk.103 In fact, the court suggested that “Hatch-Waxman essentially redistributes the relative risk assessments and explains theflow of settlement funds and their magnitude.”104 Under these circum-stances, settlements usually are procompetitive, and pro-innovation, becausethey facilitate certainty.105 Later that same year, however, a different Eleventh Circuit panel applied the Schering-Plough test to permit a challenge to a reverse payment settlement. InAndrx Pharmaceuticals v. Elan Corp., the Eleventh Circuit considered a chal-lenge by generic manufacturer Andrx to a settlement entered into by the patentholder and a different generic company, SkyePharma.106 SkyePharma was thefirst Paragraph IV filer with respect to naproxen, an anti-inflammatory medica-tion, which was covered by a patent owned by Elan.107 In the settlementagreement, SkyePharma admitted infringement and received a license to manu-facture controlled release naproxen.108 Andrx claimed that SkyePharma did notintend to market the product, in which case the 180-day exclusivity periodwould not begin to run, thereby preventing generic competition prior to patentexpiration.109 The Eleventh Circuit held that these facts supplied an adequatebasis for an antitrust claim at the pleading stage.110 According to the court, if thefacts as pled were true, “this dynamic would exceed the scope of [the] exclusionintended by the . . . patent.”111 The Sixth Circuit addressed reverse settlements in In re Cardizem CD Anti- trust Litigation.112 This case involved an agreement under which the genericmanufacturer, Andrx, refrained from marketing a generic version of the drug inquestion in return for interim payments of $10 million per quarter while theunderlying patent infringement litigation was pending.113 The agreement furtherprovided that Andrx would receive a final payment of $100 million, less anyinterim payments, if the litigation terminated without a finding of infringe-ment.114 The litigation finally settled nearly two years after the companies 103. Id. at 1074 (quoting In re Ciprofloxacin Hydrochloride Antitrust Litig., 261 F. Supp. 2d 188, 104. Id.
105. See id. at 1075.
106. 421 F.3d 1227, 1230–31 (11th Cir. 2005).
107. Id. at 1231.
108. Id.
109. Id.
110. Id. at 1234–36.
111. Id. at 1235. However, the court affirmed the dismissal of Andrx’s claims that Elan had engaged in sham litigation by filing patent infringement suits against Andrx. See id. at 1234.
112. 332 F.3d 896 (6th Cir. 2003).
113. Id. at 902.
114. Id. at 903.
reached this agreement.115 Pursuant to the final settlement, Andrx received a$50.7 million payment (for a total of $89.83 million in interim and finalpayments) and began to market a generic version, with the benefit of the 180-day exclusivity period under the Hatch-Waxman Act.116 In short, Andrx waspaid to delay generic entry while the litigation was pending, without sacrificingthe exclusivity period once the litigation terminated.
The agreements were challenged on antitrust and other grounds in various consolidated cases by indirect purchasers and other putative class representa-tives.117 The case reached the Sixth Circuit on interlocutory appeal from thedistrict court’s grant of summary judgment finding that the interim agreementwas a per se illegal horizontal restraint of trade.118 The Sixth Circuit emphati-cally agreed: “There is simply no escaping the conclusion that the Agreement . . . was, at its core, a horizontal agreement to eliminate competition in the market for Cardizem CD throughout the entire United States, a classic exampleof a per se illegal restraint of trade.”119 According to the court, “it is one thingto take advantage of a monopoly that naturally arises from a patent, but anotherthing altogether to bolster the patent’s effectiveness in inhibiting competitors bypaying the only potential competitor $40 million per year to stay out of themarket.”120 The common thread in the case law is that any evaluation of reverse payment settlements must account for the exclusionary power of the patent. Theseopinions represent a significant split among circuit courts, however, about howto weigh patent rights against the public policy interest in generic drug competi-tion. Many law professors and economists are concerned about this split,particularly insofar as some of the case law suggests that reverse paymentsettlements are shielded from scrutiny by the patent laws, as evidenced by anamicus brief filed in support of the plaintiffs’ petition for certiorari in theFederal Circuit Ciprofloxacin case.121 The Supreme Court thus far, however, hasnot taken up the question.122 115. Id.
116. Id.
117. Id. at 903 n.7.
118. Id. at 905.
119. Id. at 908.
120. Id. (footnote omitted). The Sixth Circuit also rejected defendants’ arguments concerning antitrust injury. Id. at 912. Accordingly, the court affirmed the district court’s grant of summaryjudgment in plaintiffs’ favor. Id. at 915.
121. See Brief Amici Curiae of 54 Intellectual Prop. Law, Antitrust Law, Econ., and Bus. Professors et al. in Support of Petitioner at 2–3, Ark. Carpenters Health & Welfare Fund v. Bayer AG, 129 S. Ct.
2828 (2009) (mem.) (No. 08-1194).
122. Most recently, the Court declined to grant certiorari in the Ciprofloxacin litigation before the Federal Circuit. See Ark. Carpenters Health & Welfare Fund, 129 S. Ct. 2828.
The key antitrust enforcement agencies, the Federal Trade Commission and Department of Justice, also have taken somewhat contradictory positions onreverse payment settlements.
As reflected in the Schering-Plough case discussed above, the FTC believes reverse payment settlements in the Hatch-Waxman context are presumptivelyanticompetitive.123 Indeed, the FTC considers the social costs imposed by suchsettlements “a matter of pressing national concern.”124 The FTC claims that itsenforcement actions had deterred “pay for delay” settlements until the appellatecourt rulings from the Second, Eleventh, and Federal Circuits once againopened the door to this practice.125 The FTC, therefore, supports federal legisla-tion that would generally prohibit settlements where the generic manufacturerreceives value in exchange for an agreement to refrain from selling the genericproduct.126 The Department of Justice initially took a less dogmatic stance than the FTC.
In an amicus brief relating to a certiorari petition filed in Schering-Plough, theDOJ stated that “[i]n the patent context . . . a settlement involving restrictionson the sale of the products in question is not necessarily impermissible.”127 TheDOJ noted that reverse payments in Hatch-Waxman cases implicate competingpolicy concerns: the statutory right to exclude afforded by a patent and theHatch-Waxman policy facilitating challenges to weak patents.128 The DOJ didnot articulate a precise standard under which these competing concerns could bebalanced. Rather, it argued that the FTC’s proposed standard was unclear andthat the complex procedural history of Schering-Plough rendered it an inappro-priate vehicle for deciding broad policy questions.129 Recently, however, perhaps reflecting a change in philosophy under the Obama administration, the DOJ has taken a more critical stance that wouldrequire a complete rule of reason analysis of most reverse payment settlements.
In response to an invitation from the Second Circuit in Arkansas CarpentersHealth & Welfare Fund v. Bayer AG, the DOJ submitted a brief in which itasserted that “[s]ettlements [i]nvolving [a] [p]ayment [i]n [e]xchange for [a]n[a]greement [t]o [w]ithdraw [a] [v]alidity [c]hallenge [a]nd [l]imit [c]ompetition 123. See Appeal Brief of Counsel Supporting the Complaint, In re Schering-Plough Corp., No. 9297 (Fed. Trade Comm’n Aug. 9, 2002), available at http://www.ftc.gov/os/2002/08/scheringtrialbrief.pdf.
124. Anticompetitive Pay-for-Delay Settlements in the Pharmaceutical Industry, supra note 4, at 1 (prepared statement of Richard A. Feinstein, Director, Bureau of Competition, Federal Trade Commis-sion).
125. Id. at 9.
126. Id. at 19.
127. Brief for the United States as Amicus Curiae at 8, Fed. Trade Comm’n v. Schering-Plough Corp., 548 U.S. 919 (2006) (No. 05-273), available at http://www.usdoj.gov/atr/cases/f216300/216358.pdf.
128. Id. at 10–11.
129. Id. at 14–16.
[a]re [p]resumptively [u]nlawful.”130 Nevertheless, the DOJ noted that thispresumption could be rebutted if “there is no reason to find that the settlementdoes not provide a degree of competition reasonably consistent with the parties’contemporaneous evaluations of their prospects of litigation success.”131 The DOJ argued that it is “neither necessary nor appropriate” for the court to assess the likelihood of success in the underlying patent litigation because that“determination would be based on information available to the parties whenthey entered into the settlement.”132 Nevertheless, the DOJ suggested that“[l]iability properly turns on whether, in avoiding the prospect of invalidationthat accompanies infringement litigation, the parties have by contract obtainedmore exclusion than warranted in light of that prospect.”133 The presumption of illegality would be rebutted if the payment to the alleged infringer does not exceed litigation costs.134 If the payment is “greatly in excessof avoided litigation costs,” the focus would turn to the “nature and extent of thegeneric competition permitted.”135 According to the DOJ, settlements that pre-clude generic entry prior to patent expiration would necessarily fail to carry thisburden.136 If the settlement provides for generic entry prior to patent expiration, the presumption could be rebutted if “the settlement preserved a degree of competi-tion reasonably consistent with what had been expected if the infringementlitigation went to judgment.”137 The parties would be required to show that thesettlement terms “reasonably reflected their contemporaneous evaluations of thelikelihood that a judgment in the patent litigation would have resulted in genericcompetition before patent expiration.”138 However, even if the parties werehighly certain that the patent would be upheld, “a reverse payment settlementpermitting significantly less generic competition than would be consistent withthat likelihood would be an unreasonable restraint on competition.”139 Bills pending in the Senate and House would prohibit or limit most reverse payment settlements.140 The House bill and the original Senate bill wouldamend the Federal Trade Commission Act by prohibiting an ANDA filer from 130. Brief for the United States in Response to Court’s Invitation, supra note 56, at 21.
131. Id. at 10.
132. Id. at 24 & n.8.
133. Id. at 25. The DOJ acknowledged that this view was at odds with its prior views. Id. at 26 n.9.
134. Id. at 28.
135. Id. at 29.
136. Id.
137. Id. at 30.
138. Id. at 30–31.
139. Id. at 31.
140. See Preserve Access to Affordable Generics Act, S. 369, 111th Cong. § 2(b) (as reported and amended by S. Comm. on the Judiciary, Oct. 15, 2009); Protecting Consumer Access to Generic DrugsAct of 2009, H.R. 1706, 111th Cong. § 2(a) (2009).
receiving anything of value in exchange for an agreement “not to research,develop, manufacture, market, or sell [the ANDA product] . . . for any period oftime.”141 The Senate bill would also add the penalty of forfeiture of the 180-dayHatch-Waxman exclusivity period for violation of this provision.142 The Housebill contains a similar prohibition and penalty provision.143 The revised Senate bill includes a finding that reverse payment settlements “have unduly delayed the marketing of low-cost generic drugs contrary to freecompetition, the interests of consumers, and the principles underlying antitrustlaw.”144 The revised bill, however, provides only a rebuttable presumption ofillegality where the ANDA filer receives value or agrees to limit its activitieswith respect to the ANDA product.145 The presumption can be rebutted “if theparties to such agreement demonstrate by clear and convincing evidence thatthe procompetitive benefits of the agreement outweigh the anticompetitiveeffects of the agreement.”146 The bill includes a list of factors that can beconsidered under this rule of reason analysis.147 In addition, the bill wouldprovide a safe harbor for settlements in which the ANDA filer is permitted earlymarket entry and is reimbursed for “reasonable litigation expenses not to exceed$7,500,000.”148 Finally, the bill would permit the FTC to impose civil penaltiesfor violations of up to three times the value received in the settlement.149 141. H.R. 1706 § 2(a)(2); S. 369 § 3. Section 28 of the Federal Trade Commission Act, as proposed, will allow the Federal Trade Commission to initiate enforcement proceedings against parties to a patentsettlement agreement. S. 369 § 3. In addition, the proposed section 28 allows for the agreement to bepresumed anticompetitive if the “ANDA filer receives anything of value; and . . . the[ ] filer agrees tolimit or forego research, development, manufacturing, marketing, or sales of the ANDA product for anyperiod of time.” S. 369 § 3.
142. S. 369 § 5.
143. H.R. 1706 §§ 2(a), 4.
144. S. 369 § 2(a)(6)(B).
145. Id. § 3(a).
146. Id.
147. Id. The factors are: (1) the length of time remaining until the end of the life of the relevant patent, compared with the agreed upon entry date for the ANDA product; (2) the value to consumers of the competition from the ANDA product allowed under the (3) the form and amount of consideration received by the ANDA filer in the agreement resolving or settling the patent infringement claim; (4) the revenue the ANDA filer would have received by winning the patent litigation;(5) the reduction in the NDA holder’s revenues if it had lost the patent litigation;(6) the time period between the date of the agreement conveying value to the ANDA filer and the date of the settlement of the patent infringement claim; and (7) any other factor that the fact finder, in its discretion, deems relevant to its determination of competitive effects under this subsection.
II. EFFORTS TO THEORIZE THE REVERSE PAYMENT PROBLEM Reverse payment settlements have attracted attention from a number of prominent intellectual property and competition law scholars. As discussedbelow, the various proposals these scholars have offered focus on efforts tobalance the legitimate exclusionary power of patents against the benefits ofgeneric competition, while at the same time affording some certainty to liti-gants. Herbert Hovenkamp, Mark Janis, and Mark Lemley, for example, suggestthat cases in which the antitrust analysis is unclear “should be decided on IPgrounds.”150 By way of comparison, cases in the “clear” category are those inwhich “(1) the agreement would be lawful under the antitrust laws even in theabsence of any IP dispute, or (2) the agreement would be unlawful under theantitrust laws even if all the IP claims that are made were fully sustained.”151The first category includes, for example, settlements by which the parties giveeach other unrestricted, non-exclusive licenses.152 The second category includescases in which there is a horizontal market division or other anticompetitivearrangement “that goes beyond the scope of the disputed patent,” such as anagreement to restrict sales of products unrelated to the patent claims.153 Theantitrust analysis would be unclear where “the settlement agreement wouldconstitute lawful use of the claimed IP right if an infringement claim was valid,but not if there were no valid IP right.”154 Hovenkamp, Janis, and Lemley suggest that rule of reason analysis is inappro- priate in disputed cases because “[t]he issue in such cases is not so much theeconomic consequences of the agreement as whether those consequences aredeemed acceptable as a matter of IP policy.”155 However, they acknowledge thatthe high costs of adjudicating the validity of an intellectual property claim might insome cases favor rule of reason analysis over a full assessment of the intellec-tual property rights.156 When it is clear that neither of the parties possessesmarket power in any relevant market, for example, the authors suggest that acourt could dispose of the case under the rule of reason instead of assessing thestrength of the intellectual property rights.157 Hovenkamp, Janis, and Lemley argue that reverse payments should be “pre- sumptively unlawful, shifting the burden of proof to the infringement plain-tiff.”158 The infringement plaintiff would then need to show “(1) that the ex ante 150. See Herbert Hovenkamp, Mark Janis & Mark A. Lemley, Anticompetitive Settlement of Intellectual Property Disputes, 87 MINN. L. REV. 1719, 1725 (2003).
151. Id.
152. Id.
153. Id. at 1726.
154. Id.
155. Id. at 1729.
156. See id. at 1732.
157. Id. at 1733.
158. Id. at 1759.
likelihood of prevailing in its infringement lawsuit is significant, and (2) that thesize of the payment is no more than the expected value of litigation and collateralcosts attending the lawsuit.”159 They suggest, however, that the inquiry into theunderlying intellectual property suit “need not be particularly searching” because“[t]he goal is merely to ensure that there is a legitimate dispute being settled.”160 Daniel Crane also argues that the question of reverse payment settlements is at heart an intellectual property issue, but is more solicitous of patent rights.161He notes that “restrictive rules regarding patent infringement settlements maycreate super-optimal uncertainty regarding patent rights leading to sub-optimalinventive activity or delays in the marketing of non-infringing substitutes.”162Crane argues that “the optimal rule would permit exit payment settlementswhen the ex ante likelihood of success of the patentee’s infringement suit ishigh and prohibit them when the ex ante probability of success is low.”163 In contrast, Thomas Cotter suggests that it is impractical to assess patent strength ex ante.164 In a commentary on Hovenkamp, Janis, and Lemley’sproposal, Cotter suggests that their three-part test is helpful, but that “requiringantitrust tribunals to scrutinize the merits of a settled IP dispute threatens tounravel the substantial private and social benefits to which the settlement givesrise, including the reduction in litigation costs that settlement generally pro-motes.”165 Cotter argues that, in some cases, settlement payments that exceedthe value of litigation costs could be procompetitive.166 Where the “amount ofthe reverse payment is higher than the saved litigation expenses but less than thedefendant’s potential loss at trial,” Cotter concludes that the agreement probablyrepresents the parties’ judgment that the patent claim likely would have suc-ceeded on the merits, even if there is less than absolute certainty about the out-come.167 If the settlement payment approaches or exceeds the litigation expenses plusthe amount the defendant could have made selling the allegedly infringing product—and, thus, the amount the plaintiff could potentially have obtained in damages in theinfringement suit—stricter antitrust scrutiny is justified.168 159. Id.
160. Id. at 1760.
161. Daniel A. Crane, Exit Payments in Settlement of Patent Infringement Lawsuits: Antitrust Rules and Economic Implications, 54 FLA. L. REV. 747, 750 (2002); see also ROBIN FELDMAN, THE ROLE OFSCIENCE IN LAW 160–69 (2009) (suggesting that courts evaluating reverse payment settlements are betterequipped to determine the likelihood of success on the merits of the patent litigation than they are toevaluate a settlement’s effect on consumer surplus).
162. Crane, supra note 161, at 749.
163. Id. at 750.
164. Thomas F. Cotter, Refining the “Presumptive Illegality” Approach to Settlements of Patent Disputes Involving Reverse Payments: A Commentary on Hovenkamp, Janis & Lemley, 87 MINN. L.
REV. 1789, 1815 (2003).
165. Id. at 1795.
166. Id. at 1802–09 (stating that “per se treatment of reverse payment settlements is inappropriate, because these agreements also have some potential to enhance rather than impede efficiency”).
167. Id. at 1814.
168. Id. at 1814–15.
Christopher M. Holman takes a somewhat different track after examining the diversity and complexity of Paragraph IV litigation settlements.169 Holmannotes that very few reverse payment settlements involve simple payments inreturn for total generic exclusion prior to patent expiration.170 He suggests thatwhere barriers to generic entry are low and there are numerous potential genericchallengers, reverse payment settlements will not pose a significant problembecause the iterative process of resolving multiple Paragraph IV challengeswould become counterproductive.171 Holman suggests that removing the 180-day exclusivity period for Paragraph IV first filers, or at least the ability to“park” exclusivity during a period of negotiated delayed entry, would solvemany of the generic entry problems that can make reverse payment settlementsproblematic.172 Scott Hemphill agrees that reverse payments are a problem because of faulty regulatory design and that the typology of reverse payment settlements is variedand complex.173 However, Hemphill is less sanguine than Holman about thecompetitive benefits of such settlements and sees much of the complexity as aneffort by the parties to hide the anticompetitive potential of such settlements inthe face of stronger FTC enforcement.174 Although strict antitrust rules couldoverdeter potentially beneficial settlements, Hemphill is far more concernedabout underdeterrence, particularly where the public interest in access to medi-cine is at stake.175 Therefore, Hemphill concludes that “a settlement should beaccorded a presumption of illegality as an unreasonable restraint of trade if thesettlement both restricts the generic firm’s ability to market a competing drugand includes compensation from the innovator to the generic firm.”176 Finally, Michael Carrier argues that reverse payment agreements should be considered presumptively unlawful in light of Hatch-Waxman’s regulatory struc-ture.177 Carrier supports his analysis with reference to the Supreme Court’sopinion in Verizon Communications Inc. v. Law Offices of Curtis V. Trinko,LLP.178 There, the Court stated that the existence of a pervasive regulatory 169. See Holman, supra note 28.
170. Id. at 494–95 (stating that “a closer look at the facts of individual cases reveals that few, if any, reverse payment settlements are as simple as that or as blatantly anti-competitive”).
171. Id. at 506. Holman argues that, if generic entry is relatively easy, “the branded company would be expected to experience a parade of subsequent third party generic companies challenging the patentand threatening to enter the market. Eventually, the cost of paying off all the potential competitorswould outstrip the profits of even the most lucrative blockbuster drug.” Id. 172. See id. at 516–19. Holman also argues that existing ANDA filing requirements remain overly costly and cumbersome. See id. at 519–23.
173. C. Scott Hemphill, Paying for Delay: Pharmaceutical Patent Settlement as a Regulatory Design Problem, 81 N.Y.U. L. REV. 1553, 1561 (2006); see Hemphill, supra note 24, at 663–70.
174. See Hemphill, supra note 24, at 685.
175. See Hemphill, supra note 173, at 1616.
176. Id. at 1561.
177. Michael A. Carrier, Unsettling Drug Patent Settlements: A Framework for Presumptive Illegal- ity, 108 MICH. L. REV. 37, 67–70 (2009).
178. Id. at 68–69 (citing Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S.
regime must be considered when evaluating conduct under the antitrust laws.179Carrier suggests that the Hatch-Waxman regime reflects a policy in favor ofpatent challenges, that the regulatory regime is relatively ineffective in achiev-ing this purpose, and that reverse payments are “uniquely concerning” becausethey allow branded and generic manufacturers to limit competition.180 B. TOWARDS A SETTLEMENT COMPETITION INDEX? The approach outlined in this Article, which is discussed in detail in Part III below, would require a more refined inquiry into the actual anticompetitiveeffects of any particular reverse payment agreement. A number of the proposalsoutlined above seek to avoid this sort of searching inquiry into market structure.
This section explains why existing proposals are inadequate and why a thoroughanalysis of market power is necessary.
Proposals that would limit settlement amounts to litigation costs, use the settlement amount as a proxy for patent strength, or hold reverse paymentsettlements presumptively unlawful, are problematic because they threaten tooverdeter potentially beneficial settlements. As Scott Hemphill notes, such arestrictive policy choice runs the risk of “false positives.”181 In Hemphill’sjudgment, “false positives” are far less damaging than “false negatives” becauseof the large deadweight losses resulting from many pharmaceutical patents.182The Hovenkamp–Lemley–Janis test, Michael Carrier’s Trinko approach, thelegislation currently pending before Congress, and the enforcement agencies’current policies likewise seem to reflect far more concern about underdeterrencethan overdeterrence.
On the surface, this stance seems appropriate given the public interest in drug price competition and the purposes of the Hatch-Waxman Act. However,an overly aggressive regulatory posture ignores another key factor in mostParagraph IV settlements: the length of the bargained-for restriction on genericcompetition.
Using settlement amounts as the sole measure of validity threatens to turn the Paragraph IV process into an all-or-nothing proposition: either the patent isupheld and there is no generic competition during the remaining patent term, orthe patent is invalidated. Most Paragraph IV settlements, however, permitgeneric entry at some point prior to patent expiration. A reverse paymentsettlement that results in a reduced patent term often will benefit the publicmore than a trial on the merits and subsequent appeals, even if the patentultimately is invalidated. As the Justice Department noted in its brief recently 179. See id.
180. See id. at 67–76.
181. See Hemphill, supra note 24, at 669–70.
182. See id. filed in Arkansas Carpenters Health & Welfare Fund (the Barr litigation), [a]t least as a general matter, a settlement dividing the remaining life of thepatent into a period of exclusion and a period of competition, based on theparties’ expectations as to the likelihood of the patent being invalidated (andtherefore their understanding of the value of a litigated outcome, on average),will adequately accommodate the public interest in freeing the market fromundeserved monopolies.183 The Justice Department’s brief, however, incorrectly assumes that any re- verse payment connected to an agreement to divide the remaining patent termmust “naturally [be] viewed as consideration for the generic’s agreement todelay entry beyond the point that would otherwise reflect the parties’ sharedview of the likelihood that the patentee would ultimately prevail in the litiga-tion.”184 If the Paragraph IV challenger cannot receive more than its litigationcosts in settlement, or if the amount of the settlement payment is the keydeterminant of antitrust exposure, the incentive to settle before trial is greatlydiminished, and this important benefit could be lost.
It is true that a settlement without any reverse payment could include other important benefits, including early entry. It is unclear, however, whether caseswould settle if early entry, without any reverse payment, is the only option onthe table. In such cases, the entry date bargained for by the generic manufac-turer presumably would be significantly earlier than an early entry date that iscoupled with a payment. At some point, the branded manufacturer will chooseto litigate rather than to sacrifice most of its remaining patent life in settlement.
For example, consider a patent with six years remaining on its term that produces rents of $1 billion per year, against which one potential generic entrantfiles a Paragraph IV certification. If the subsequent infringement litigation takesfour years from filing until a decision is issued on appeal, and the genericcompany prevails, the public will benefit from two years of early genericcompetition and up to a $2 billion increase in consumer surplus, less anyexternalities and transaction costs imposed by four years of costly litigation.185If the patent challenge is unsuccessful, there is zero increase in consumersurplus and consumers must still bear the externalities imposed by the litigation 183. Brief for the United States in Response to the Court’s Invitation, supra note 56, at 22.
184. Id.
185. These include the consumption of judicial resources and, perhaps more importantly, the diversion of time and human resources away from the core business of product innovation. SeeAlbert W. Alschuler, Mediation with a Mugger: The Shortage of Adjudicative Services and the Need fora Two-Tier Trial System in Civil Cases, 99 HARV. L. REV. 1808, 1812 (1986) (“Although litigationcommonly proves expensive to the litigating parties, these parties pay only a fraction of the cost ofoperating the courts. In that sense, governments subsidize civil litigation, and the subsidy is substan-tial.”); Thomas D. Rowe, Jr., American Law Institute Study on Paths to a “Better Way”: Litigation,Alternatives, and Accommodation: Background Paper, 1989 DUKE L.J. 824, 868–73 (discussingpositive and negative externalities and transaction costs of litigation).
process. By comparison, if the litigation settles within one year of filing, and theparties agree to split the remaining patent term, the public will benefit fromthree years of early generic competition and up to a $3 billion increase inconsumer surplus, less the amount of the reverse payment to the genericchallenger.186 It is easy to see that, under a wide variety of similar scenarios, theoverall increase in consumer surplus resulting from a reverse payment settle-ment can be greater than that resulting from a successful patent challenge. TheSettlement Competition Index avoids the potentially inefficient results ofthe Justice Department’s test by requiring a more complete evaluation of theagreement’s actual competitive effects.
2. Settlement Amount as a Proxy for Patent Strength Another significant question about my approach is whether it does more work than existing proposals that use the amount of the settlement payment as agauge of the parties’ beliefs about patent strength.187 There are at least threereasons why the amount of the settlement payment alone often is not anadequate measure.
First, many reverse payment settlements involve unrelated licenses, autho- rized generic sales, and other side deals, in addition to or in lieu of monetarypayments.188 As Scott Hemphill has noted, “[v]iewed in isolation, it is difficultto tell whether a side deal represents payment for value or disguised paymentfor delayed generic entry.”189 Hemphill concludes that the rarity of such dealsbetween branded and generic companies outside the Paragraph IV settlementcontext renders the inclusion of such deals in settlements suspicious andpresumptively unlawful.190 It is not clear, however, that such a presumption isalways warranted. In Schering-Plough, for example, the Eleventh Circuit con-cluded that a side deal, which had been explored by the branded company priorto the Paragraph IV litigation, had significant independent business merit.191 Inany event, even if Hemphill’s general suspicions are correct, the complicated,multilayered nature of the typical settlement package belies any simple correspon-dence between risk and payment size.
Second, and perhaps most importantly, unlike an ordinary patent case, the generic challenger faces no risk in the Paragraph IV infringement case beyondlitigation expenses. The pharmaceutical patent holder, in contrast, risks losing 186. Of course, if the parties were certain of the outcome and timing of the litigation, they would not rationally settle the case for any entry date prior to that which the generic could achieve through thelitigation. The primary reason for settlement, however, is that the timing of an outcome of the litigationis uncertain. The point here is that the parties’ adjustment of this uncertainty through a reverse paymentsettlement can often create a better net social welfare outcome than litigation.
187. See, e.g., Cotter, supra note 164, at 1814–15.
188. See Hemphill, supra note 24, at 663–69.
189. Id. at 668.
190. Id. at 668–69.
191. See 402 F.3d 1056, 1070 (2005).
an asset that might produce billions of dollars in rent each year.192 As theEleventh Circuit stated in Schering-Plough, “Hatch-Waxman essentially redistrib-utes the relative risk assessments and explains the flow of settlement funds andtheir magnitude.”193 In this regard, it should be noted that Hatch-Waxman is not, or at least was not designed to be, merely a simplification of existing declaratory judgmentprocedures. First, the 180-day exclusivity period for first filers provides anadditional incentive over the declaratory judgment context.194 Even without the180-day exclusivity period, however, in the context in which the Paragraph IVprocedures were originally crafted, generic manufacturers retain significantincentives to challenge patents without incurring substantial risk.195 Prior to the Supreme Court’s recent decision in MedImmune, Inc. v. Genen- tech, Inc.,196 the Federal Circuit employed the “reasonable apprehension ofsuit” test, under which a declaratory judgment plaintiff in a patent case had toshow “‘both (1) an explicit threat or other action by the patentee, which createsa reasonable apprehension on the part of the declaratory plaintiff that it will facean infringement suit, and (2) present activity which could constitute infringe-ment or concrete steps taken with the intent to conduct such activity.’”197 Underthis test, the generic manufacturer would have to incur the risk of a significantdamages suit before it could file an action for a declaratory judgment ofinvalidity or non-infringement. In fact, the Federal Circuit held in Teva Pharma-ceuticals USA, Inc. v. Pfizer, Inc. that even the filing of an ANDA with aParagraph IV certification is insufficient to confer a generic manufacturer withstanding to file a declaratory judgment action against the patent holder (al-though, in determining whether there is a “reasonable apprehension of suit,” itis a factor to be considered in the totality of the circumstances).198 Therefore,prior to MedImmune, the Hatch-Waxman Paragraph IV framework representeda significant reallocation of risk over ordinary declaratory judgment proce-dures.199 192. See Hemphill, supra note 24, at 648–49 (noting that the average U.S. annual sales of drugs implicated in reverse payment settlements was $1.3 billion, and that some had annual sales as high as$7 and $3 billion).
193. 402 F.3d at 1074.
194. See 21 U.S.C. § 355(j)(5)(B)(iv) (2006).
195. See 35 U.S.C. 271(e)(4) (2006).
196. 549 U.S. 118 (2007).
197. See Benitec Austl., Ltd. v. Nucleonics, Inc., 495 F.3d 1340, 1343–44 (Fed. Cir. 2007) (quoting B.P. Chems. Ltd. v. Union Carbide Corp., 4 F.3d 975, 978 (Fed. Cir. 1993), abrogated by Benitec Austl.,Ltd., 495 F.3d 1340).
198. Teva Pharms. USA, Inc. v. Pfizer, Inc., 395 F.3d 1324, 1333–37 (Fed. Cir. 2005) (stating that the business disadvantage to the second generic manufacturer—the plaintiff—created by the firstgeneric manufacturer’s 180-day exclusivity period is “the product of the Hatch-Waxman scheme andthe fact that [defendant] has acted in a manner permitted under that scheme. It is not the product of athreat of suit by [defendant]” and, therefore, does not amount to an actual controversy between plaintiffand defendant), abrogated by MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 132 (2007).
199. See, e.g., Glaxo Group Ltd. v. Dr. Reddy’s Labs., Ltd., 325 F. Supp. 2d 502, 508 (D.N.J. 2004) (holding that ANDA filing alone is insufficient to satisfy the “reasonable apprehension of suit” test). As In MedImmune, the Supreme Court called into question the Federal Circuit’s “reasonable apprehension of suit” test.200 However, the Federal Circuit hasgenerally interpreted MedImmune narrowly as relating primarily to cases inwhich there is an existing licensing relationship.201 According to the FederalCircuit, “[i]n the context of conduct prior to the existence of a license, declara-tory judgment jurisdiction generally will not arise merely on the basis that aparty learns of the existence of a patent owned by another or even perceivessuch a patent to pose a risk of infringement without some affirmative act by thepatentee.”202 However, the Federal Circuit demonstrated its willingness to follow MedIm- mune in Teva Pharmaceuticals USA, Inc. v. Novartis Pharmaceuticals Corp.,where the Federal Circuit recognized the abrogation of Teva PharmaceuticalsUSA, Inc. v. Pfizer Inc.203 According to the Novartis panel, the filing of anANDA entitles the filer to bring a declaratory judgment action against the patentowner.204 Because an ANDA filing is an act of infringement, which gives thepatent owner a right to sue, “[i]t logically follows that if such an action creates ajusticiable controversy for one party, the same action should create a justiciabledeclaratory judgment controversy for the opposing party.”205 It is now possible, then, for an ANDA filer to bring a declaratory judgment action if the patent owner fails to sue within the 45-day Paragraph IV window.
It remains true, however, that the Hatch-Waxman procedures significantly alterthe balance of risks in comparison to ordinary patent litigation. Under theFederal Circuit’s jurisprudence after MedImmune, unless there is an ongoinglicensing relationship, a declaratory judgment action for non-infringement orinvalidity ordinarily is only possible when the declaratory judgment plaintiff isengaging in activity that exposes it to a real risk of a claim for damages. TheANDA context is an exception because an ANDA filing is a technical act ofinfringement that, by itself, does not give rise to a claim for damages. In fact,the Federal Circuit’s decision in Novartis tips the playing field even further infavor of the generic manufacturer. Now the generic manufacturer can place apatent at risk, without any threat of a damages claim, even if the patent ownerchooses not to sue for infringement as permitted under the Paragraph IVframework.
the Glaxo court notes, the Hatch-Waxman Act was amended in 2003 to allow a Paragraph IV filer notsued for infringement by the patent holder within the 45-day window to file a declaratory judgmentaction. Id. at 507–08 (citing 35 U.S.C. § 271(e)(5) (2006)). However, the Glaxo court held that thisprovision did not modify the “reasonable apprehension of suit” test, which it believed was rooted in theconstitutional “case and controversy” requirement. See id. 200. See MedImmune, Inc., 549 U.S. at 132 n.11.
201. See, e.g., SanDisk Corp. v. STMicroelectronics, Inc., 480 F.3d 1372, 1384 (Fed. Cir. 2007).
202. See id. at 1380–81.
203. See Teva Pharms. USA, Inc. v. Novartis Pharms. Corp., 482 F.3d 1330, 1339–40 (Fed. Cir. 2007).
204. See id. at 1342.
205. Id. The court also was persuaded that 35 U.S.C. § 271(e)(5), the ANDA declaratory judgment provision, supports the generic manufacturers’ right to file a declaratory judgment action. Id. at 1342–43.
A final reason why the amount of payment for any one settlement might not provide an accurate gauge of patent strength is the possibility of multipleParagraph IV challenges by different ANDA filers. In deciding whether to settleany given challenge, the patent owner must consider (a) what signals the settle-ment will send to other existing or potential ANDA filers and (b) the costs ofsubsequent settlements.
In this regard, the ANDA settlement process bears some similarity to multidis- trict mass tort litigation, where there are multiple cases proceeding on differenttimetables.206 In such circumstances, the defendant must engage in an iterativebargaining game with various different plaintiffs. The settlement price paid toany individual plaintiff might be higher or lower than the true value of theclaim, depending on the defendant’s assessment of the value of all the relatedclaims in the aggregate and the plaintiff’s beliefs about the payoffs available tothe defendant for settling at any given point during the life of the broaderlitigation.207 In similar fashion, viewing any one Paragraph IV settlement inisolation will likely provide a distorted image of the actual value of that claim.
When the stakes are so high, and the risks so unbalanced, the patent holder likely will pay a significant settlement premium even if the risk of losing theinfringement case is very small. By statutory design, the Paragraph IV filer hasthe patent owner “over a barrel” regardless of whether the patent is strong orweak. A reverse payment may represent the patent holder’s belief that the patentis weak, but it may instead represent a risk-averse decision to buy off even aslim risk that a strong patent claim will fail in the courts. The latter decisionwould help to achieve certainty about rents that represent many multiples of thatamount or it may reflect something about where that individual settlement fallsin the process of bargaining with all potential filers. The value of an individualsettlement, therefore, cannot be used as a reliable proxy for patent strength.
III. A PROPOSED “SETTLEMENT COMPETITION INDEX” As discussed in the previous Part, existing proposals in scholarship revolve around the same central concern that has been addressed by the courts: what isthe legitimate exclusionary power of a patent? None of these authorities,however, have focused on how patents function in particularized productmarkets. In this Part, I suggest what is, in effect, a significant elaboration of theHovenkamp–Janis–Lemley test. This new test can be developed into a “safeharbor” framework that focuses on patent strength in connection with productmarket structure. Under this test, the court or regulatory agency first wouldconsider the following criteria: 206. For a description of the multidistrict litigation and settlement process, see generally Howard M.
Erichson, A Typology of Aggregate Settlements, 80 NOTRE DAME L. REV. 1769 (2005).
207. See generally id. at 1786–95 (describing how aggregate settlements are distributed).
(1) The difference in product market concentration that would likely result (2) The probability that the patent will be held to be valid and infringed.
These criteria would be used to create a Settlement Competition Index (“SCI”), which provides a rough empirical gauge of the potential anticompetitive effectsof the settlement. As discussed in detail in section III.C below, the SCI formulais the market concentration (derived from the Herfindahl–Hirschman Index(“HHI”)) prior to generic entry, less the likely market concentration aftergeneric entry, divided by the probability of enforcement, or otherwise repre-sented as: Settlements at the lower and upper ranges of this Index would be presump- tively valid or invalid, respectively. Settlements in the middle range would befurther evaluated according to a balancing test under the rule of reason. Thisapproach would establish antitrust “safety zones” and zones of per se illegality,consistent with antitrust policy concerning intellectual property licensing andmergers generally. This would promote greater certainty and efficiency in thesettlement bargaining process.
Each element of my proposal is unpacked in the next section. Some possible objections to my approach are discussed in the following sections. The finalsection provides details about the calculation and application of the SCI.
A. THE IMPORTANCE OF ASSESSING PRODUCT MARKET CONCENTRATION The first prong of my proposed test involves assessment of market concentra- tion before and after the agreement. None of the courts, legal commentators, orregulatory agencies that have considered the reverse payment settlement prob-lem thus far has paid sufficient attention to the importance of product marketdefinition.208 This is surprising because a patent’s impact on market concentra-tion is a key determinant of the patent’s power.
All of the authorities agree that valid patents provide a legitimate zone of exclusion and that reverse payment settlements are problematic to the extentthey expand the patentee’s exclusionary power beyond that inherent in thepatent. The scope of a patent’s legitimate exclusionary zone is defined by the 208. As noted infra in notes 231–44 and their accompanying text, most economists who have attempted to model responses to the reverse payment settlement problem define a relevant productmarket, but oversimplify their models by assuming the market is either a monopoly or a duopoly.
scope of the claims, including the range of covered equivalents.209 The power ofa patent’s legitimate exclusionary zone, however, cannot be defined withoutreference to product market definition.210 The definition of a patent’s exclusion-ary zone must encompass both scope and power.
Scope and power are complementary, but not identical, concepts. It is true that scope and power usually involve a proportional relationship: as the scope ofa patent’s claim expands, its potential market power expands. A patent claim for“pharmaceutical compounds to treat depression” obviously would confer greatermarket power than a patent claim for the chemical formulation of the activeingredient in Prozac, fluoxetine hydrochloride.211 The doctrines that limit claimscope—novelty, non-obviousness, and the rules regarding claim constructionand the range of equivalents—therefore also serve to constrain patent power.
But these doctrines do not fully circumscribe the range of a patent’s power. Iffluoxetine hydrochloride is the only compound capable of treating depression, apatent claiming that compound confers greater market power than is the casewhen there are multiple non-infringing compounds that can be used to treatdepression with similar clinical results.
More generally, product market definition is fundamental to antitrust analysis under the rule of reason. As one leading treatise notes, “The determination ofthe relevant market is inextricably related to the question of whether thedefendant’s competitors have been or will be foreclosed from the market byvirtue of the challenged acts.”212 Only in the context of per se liability, wherethe conduct is deemed inherently anticompetitive, is the question of marketdefinition set aside.213 The authorities that seek some kind of per se ruleconcerning reverse payment settlements, however, do not explain why suchagreements are inherently anticompetitive except in a circular fashion thatinevitably points back to the need for product market definition.
A response to this view might be that any agreement that expands a patent’s scope should be considered an unlawful restraint of trade. This closely re-sembles arguments concerning tying arrangements involving patents, whichwere recently resolved by the Supreme Court in Illinois Tool Works, Inc. v.
Independent Ink, Inc.214 In Illinois Tool Works, the Court reversed the precedent rule that tying arrangements involving the tie of patented and unpatented products are per se 209. See, e.g., Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722, 732 (2002) (“The scope of a patent is not limited to its literal terms but instead embraces all equivalents to theclaims described.”).
210. See, e.g., Ill. Tool Works Inc. v. Indep. Ink, Inc., 547 U.S. 28, 36 (2006).
211. Fluoxetine hydrochloride is the generic name for a selective serotonin reuptake inhibitor originally sold under the brand name Prozac. See ELI LILLY & CO., PROZAC PRESCRIBING INFORMATION1 (2009), available at http://www.prozac.com/Pages/index.aspx.
212. 1 LOUIS ALTMAN & MALLA POLLACK, CALLMANN ON UNFAIR COMPETITION, TRADEMARKS AND MONOPOLIES § 4:31, at 4-308 (4th ed. 2009).
213. See id. § 4:31, at 4-306.
214. 547 U.S. 28 (2006).
unlawful.215 According to the Court, the conclusion that a particular tyingarrangement involving a patent is unlawful “must be supported by proof ofpower in the relevant market rather than by a mere presumption thereof.”216 TheCourt based this conclusion, in part, on Congress’s distinction between patentrights and market power in the Patent Act.217 The Illinois Tool Works Court essentially adopted Justice O’Connor’s reason- ing in her concurrence to Jefferson Parish Hospital District No. 2 v. Hyde, apatent tying case abrogated by Illinois Tool Works.218 Justice O’Connor therestated: A common misconception has been that a patent or copyright, a high marketshare, or a unique product that competitors are not able to offer suffice todemonstrate market power. While each of these three factors might help togive market power to a seller, it is also possible that a seller in these situationswill have no market power: for example, a patent holder has no market powerin any relevant sense if there are close substitutes for the patented product.
Similarly, a high market share indicates market power only if the market isproperly defined to include all reasonable substitutes for the product.219 Justice O’Connor’s reasoning is consistent with antitrust policy concerning intellectual property generally, including the Federal Trade Commission andDepartment of Justice guidelines for intellectual property licenses.220 A reversepayment settlement is not, of course, a license agreement. However, the anti-trust concern over reverse payment settlements is the same as the concern overexclusive licenses: will the agreement result in an unacceptable degree ofmarket concentration? A license agreement can allow the parties to aggregatethe legitimate exclusionary power of multiple intellectual property rights in away that illegitimately concentrates market power.221 In a transaction thatinvolves a sale or exclusive transfer of intellectual property rights, the agenciesapply a merger analysis to determine the agreement’s competitive effects.222 This approach rests on a sound theoretical foundation. Patent rights are 215. See id. at 45–46.
216. Id. at 43.
217. Id. at 41–42 (citing 35 U.S.C. § 271(d)(5) (2006)).
218. Id. at 37–38.
219. Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 37 n.7 (1984) (O’Connor, J., con- curring), abrogated by Ill. Tool Works Inc., 547 U.S. at 28.
220. See U.S. DEP’T OF JUSTICE & FED. TRADE COMM’N, ANTITRUST GUIDELINES FOR THE LICENSING OF INTELLECTUAL PROPERTY (1995), available at http://www.usdoj.gov/atr/public/guidelines/0558.htm.
221. See id. § 2.1 (“As with other forms of private property, certain types of conduct with respect to intellectual property may have anticompetitive effects against which the antitrust laws can and doprotect. Intellectual property is thus neither particularly free from scrutiny under the antitrust laws,nor particularly suspect under them.”); id. § 3.1 (“[A]ntitrust concerns may arise when a licensingarrangement harms competition among entities that would have been actual or likely potential competi-tors . . . .”).
probabilistic, not certain, because validity and infringement are always decidedafter the alleged infringement through litigation.223 This means that it is impos-sible to determine patent scope with certainty ex ante. As noted in section III.Cbelow, it is possible to assess with some degree of confidence the probabilitythat an infringement claim will succeed. This alone, however, does not answerthe question whether the parties’ adjustment of this probability through settle-ment should be held unlawful as an impermissible restraint of trade. Thatquestion must circle back to the likely effects of the restraint in the particularproduct market in which this probabilistic right is being asserted. Both theproduct market and the probability of enforcement must be considered to assessthe reasonableness of the restraint under all the circumstances.
Most economists that have attempted to model the effects of reverse payment settlements have recognized the need for objective measurements of productmarket concentration. For example, in a paper that provides detailed economicmodels of reverse payment settlements, Carl Shapiro developed a ratio he callsthe “‘patent competition index’ (PCI).”224 The PCI provides an index of the lossof consumer surplus when firms settle a patent challenge by merging andidentifies the amount of efficiencies the merger must generate in order topromote consumer welfare.225 Shapiro suggests that the PCI could be integratedwith the DOJ–FTC Horizontal Merger Guidelines because “[t]he safe-harborprovisions in merger enforcement can reasonably be viewed as indicating themagnitude of efficiencies that are credited to merging parties as a matter ofcourse.”226 He notes that “a somewhat greater increase in concentration wouldbe permitted if the acquired firm is operating under a patent cloud.”227 My proposal suggests some expansion of, complications to, and simplifica- tions of economic models such as Shapiro’s. The expansion is that Shapiro’sbasic insight about merger analysis should apply even when the settlementwill not result in a merger. This is consistent with the way in which enforce-ment agencies currently evaluate exclusive horizontal intellectual propertylicenses.228 An exclusive horizontal license functions like a merger in that it 223. See Mark A. Lemley & Carl Shapiro, Probabilistic Patents, J. ECON. PERSP., Spring 2005, at 75, 75 (“[E]conomists have increasingly recognized that a patent does not confer upon its owner the right to excludebut rather a right to try to exclude by asserting the patent in court.” (citation omitted)); Carl Shapiro, AntitrustLimits to Patent Settlements, 34 RAND J. ECON. 391, 395 (2003) (“[A] patent is best viewed as a probabilisticproperty right. What the patent grant actually gives the patentholder is the right to sue to prevent others frominfringing the patent. Nothing in the patent grant guarantees that the patent will be declared valid, or that thedefendant in the patent suit will be found to have infringed. In other words, all real patents are less strong thanthe idealized patent grant usually imagined in economic theory.”).
224. Shapiro, supra note 223, at 403.
225. Id.
226. See id. (implicating the DOJ–FTC Horizontal Merger Guidelines by stating that “[a]n extension to this article would be to integrate this analysis with traditional structural merger analysis based onmeasures of market concentration”).
227. Id.
228. As the joint Department of Justice–Federal Trade Commission Antitrust Guidelines for the Licensing of Intellectual Property note, the FTC and DOJ “will apply a merger analysis to an outright allows potential competitors to aggregate existing or potential market shares ina given product market. The licensee in such a transaction often contributesnothing other than monetary consideration and an agreement not to challengethe licensed intellectual property rights.
A reverse payment settlement, in turn, is similar to an exclusive license in that competition that might have resulted from a challenge to the intellectualproperty right is limited by agreement.229 Of course, the potential efficiencyjustifications differ because, in an ordinary exclusive license, the monetaryconsideration paid to the licensor presumably reflects an efficient allocation ofresources and the exchange of technology may involve spillover benefits result-ing from the acquisition of ancillary knowledge by the licensee.230 The poten-tial efficiency justifications differ when one party to the transaction is beingpaid not to develop a technology. Nevertheless, the initial question of anti-competitive effects remains similar: will the merger, exclusive license, orreverse payment settlement result in a harmful aggregation of market power?Thus, the merger analysis model can also apply to “ordinary” reverse paymentsettlements.
A way in which I propose to complicate Shapiro’s model (and other similar economic models) concerns the structure of the product market after consumma-tion of the settlement. Most existing economic models assume only duopolisticcompetition if the settlement is not consummated and the patent is held validbut not infringed.231 This assumption, however, is unrealistic in many pharma-ceutical product markets. My proposal includes an assessment of the productmarket as it actually exists for the drug subject to the settlement. Again, this isconsistent with existing DOJ–FTC intellectual property licensing and mergerguidelines and other sources of antitrust policy.
In antitrust cases, the relevant market “is composed of products that have reasonable interchangeability for the purposes for which they are produced— sale by an intellectual property owner of all of its rights to that intellectual property and to a transactionin which a person obtains through grant, sale, or other transfer an exclusive license for intellectualproperty (i.e., a license that precludes all other persons, including the licensor, from using the licensedintellectual property).” U.S. DEP’T OF JUSTICE & FED. TRADE COMM’N, supra note 220, § 5.7.
229. An exclusive license is different in that it does not exclude the licensed party from the market.
However, the exclusive license does, by definition, exclude other potential licensees, and may precludeor limit challenges to the patent by the licensee. Cf. MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118,126–37 (2007) (discussing circumstances under which a patent licensee can seek a declaratoryjudgment of patent invalidity).
230. See, e.g., U.S. DEP’T OF JUSTICE & FED. TRADE COMM’N, supra note 220, § 2.3 (noting procompetitive benefits of licensing).
231. See, e.g., id. Other modelers have made similar simplifying assumptions. See Jeremy Bulow, The Gaming of Pharmaceutical Patents, in 4 NAT’L BUREAU OF ECON. RESEARCH, INNOVATION POLICY AND THE ECONOMY 145, 159–60 (Adam B. Jaffe, Josh Lerner & Scott Stern eds., 2004) (modeling “monopolydays, triopoly days, and duopoly days” resulting from Paragraph IV litigation settlements); Robert D.
Willig & John P. Bigelow, Antitrust Policy Toward Agreements that Settle Patent Litigation, 49ANTITRUST BULL. 655, 656 n.2 (2004) (recognizing that the patent holder might not have monopolypower in real markets but adopting assumption of market power to simplify modeling).
price, use and qualities considered.”232 For pharmaceutical products, “the onlylogical place from which to determine the relevant product market is from thearray of therapeutically substitutable choices available to the doctor.”233 Experttestimony from physicians, pharmacists, third-party payors, and other sourcescan help establish which products are therapeutic substitutes.234 It is alsopossible to quantify the cross-price elasticity of demand for branded and genericproducts in the same therapeutic class.235 Often, a significant variety of competing patented and generic products exist in the same class to treat the same condition. In fact, there has been significantpolicy debate over whether patents perversely incentivize second generation“me too” drugs in a blockbuster class that yields little or no marginal increase intherapeutic value.236 For example, at least five variations of selective serotoninreuptake inhibitors (SSRIs) have been approved by the FDA to treat depres-sion.237 These are sold under at least seven brand names by three differentbranded manufacturers.238 Four of the patents relating to these compounds haveexpired and generic versions of these drugs are sold by generic manufactur-ers.239 In such a product market, the loss of generic competition with respect toone compound would not result in a true monopoly, nor would the presence ofgeneric competition with respect to one patent result in a duopoly. In otherwords, the scope of the patent claims for any one compound does not define theboundaries of the relevant product market.

Source: http://georgetown.lawreviewnetwork.com/files/pdf/98-5/Opderbeck.pdf