concluding that "Plaintiffs' accounting of the state variations [was] not comprehensive and glosses over important differences."Summary of this case from In re Niaspan Antitrust Litig.
CIVIL ACTION No. 2:06-cv-1833
Goldberg, J. MEMORANDUM OPINION
Presently before me is the End Payor Class Plaintiffs' motion for class certification filed in the consolidated antitrust lawsuit known as the In re Modafinil Litigation. The prospective class of End Payor Plaintiffs includes consumers and Third-Party Payors ("TPPs"), such as health insurance plans, which purchased the brand-name pharmaceutical, Provigil, or its generic equivalent for either their own use, their families' use, or their beneficiaries' use between June 1, 2006 and September 30, 2013.
The other cases consolidated within the In re Modafinil Litigation are: King Drug Co. of Florence, Inc. v. Cephalon, Inc. (Dkt. No. 06-1797); Apotex, Inc. v. Cephalon, Inc. (Dkt. No. 06-2768); and Federal Trade Commission v. Cephalon, Inc. (Dkt. No. 08-2141) (settlement agreement reached, pending court approval). Only the End Payors' case is implicated by the instant motion for class certification. Therefore, I will refer to End Payors as Plaintiffs.
Plaintiffs have brought this antitrust lawsuit against the manufacturer of Provigil, Cephalon, Inc., as well as four generic pharmaceutical companies: Teva Pharmaceutical Industries, Ltd. and Teva Pharmaceuticals USA, Inc. ("Teva"); Ranbaxy Laboratories, Ltd. and Ranbaxy Pharmaceuticals, Inc. ("Ranbaxy"); Mylan Pharmaceuticals, Inc. and Mylan Laboratories, Inc. ("Mylan"), and Barr Laboratories, Inc. ("Barr") (collectively "Generic Defendants"). At the center of this case are four Hatch-Waxman reverse-payment settlements, executed in 2005 and 2006 between Cephalon and each of the Generic Defendants, which are alleged to be anticompetitive for delaying the market entry of generic Provigil. Cephalon is also accused of maintaining an illegal monopoly by enforcing its patent on Provigil, which was allegedly obtained by committing fraud on the Patent and Trademark Office ("PTO").
The End Payor Plaintiffs and Mylan have reached an agreement in principle for settlement.
Plaintiffs seek certification of two classes: (1) a class of End Payors bringing claims under the state antitrust and consumer protection laws of twenty-three states and the District of Columbia; and (2) an unjust enrichment class, bringing claims under the laws of twenty-five states and the District of Columbia. Plaintiffs have also articulated numerous class member exclusions. Plaintiffs seek class certification under Federal Rule of Civil Procedure 23(b)(3), and assert that all of the requirements of Rule 23 have been satisfied. Defendants vigorously oppose certification and urge that Plaintiffs have failed to demonstrate the requirements of ascertainability, predominance and superiority.
For the reasons that follow, I find that certification of the End Payor class is not appropriate because Plaintiffs have failed to satisfy the Rule 23 requirements of ascertainability, predominance and superiority by a preponderance of the evidence. Accordingly, Plaintiffs' motion is denied.
I. FACTUAL AND PROCEDURAL HISTORY
A. Overview of the In re Modafinil Litigation
In April 1997, the PTO issued U.S. Patent No. 5,618,845 ("the '845 patent") to Cephalon, which patented a specific formulation of modafinil known as Provigil, a wakefulness-promoting drug. In 2002, Cephalon was granted a reissue patent on Provigil, U.S. Patent No. RE 37,516 ("the RE '516 patent"), which was scheduled to expire October 6, 2014. However, as a result of studying the drug's effects on children, Cephalon also received an additional six months of pediatric exclusivity on Provigil, extending Cephalon's exclusivity period through April 6, 2015.
On December 24, 2002, all four Generic Defendants filed Abbreviated New Drug Applications ("ANDAs") for generic Provigil, each certifying that Cephalon's patent was either invalid or would not be infringed by their generic modafinil product. As first-filers, all of the Generic Defendants were entitled to share in 180 days of exclusive marketing upon FDA approval, a characteristic of the Hatch-Waxman Act, Pub. L. No. 98-417. As a result of the Generic Defendants' ANDA filings, Cephalon sued the Generic Defendants for patent infringement on March 28, 2003.
All of the litigation between Cephalon and the Generic Defendants was settled between December 2005 and February 2006, while motions for summary judgment were pending. The settlements each permitted the Generic Defendants to launch their generic Provigil product on a "date certain" prior to the expiration of the RE '516 patent—April 6, 2012. The agreements further contained "contingent-launch provisions," which permitted each Generic Defendant to market generic Provigil prior to the date certain if any other company marketed generic Provigil, whether through a license or at-risk, or if the RE '516 patent was declared invalid, unenforceable, or not infringed by generic Provigil. Each of these settlement agreements contained provisions for and/or were signed alongside licenses for intellectual property, active pharmaceutical ingredient supply agreements, and pharmaceutical development agreements. Cephalon agreed to pay a total of approximately $300 million to the Generic Defendants as a result of these agreements. Plaintiffs allege that but-for these payments the Generic Defendants would have launched generic Provigil at risk, and thus lower-cost generic competition would have been brought to the prospective class members by June 2006.
Launching "at risk" means that a company has chosen to market its generic product, despite the fact that it is actively being accused of patent infringement and the court has not yet determined whether the patent is valid or has been infringed. Under the Hatch-Waxman Act, when a patent holder files an infringement lawsuit within forty-five days of an ANDA containing a certification that the patent is invalid or not infringed, the FDA may not approve the ANDA for thirty months. If the case is resolved during the thirty-month stay, the FDA will take action on the ANDA consistent with the court's judgment. However, if the case is still ongoing at the end of the thirty-month stay, the FDA may approve the ANDA, at which point the generic company may choose to launch at risk. If the infringement lawsuit is eventually resolved in favor of the patent holder, the generic company may owe damages for its at-risk launch. King Drug Co. of Florence, Inc., 2015 WL 356913, at *2 (citing 21 U.S.C. § 355(j)(5)(B)(iii); Federal Trade Commission v. Actavis, Inc., 133 S. Ct. 2223, 2228 (2013)).
Additional details regarding these settlement agreements and the Hatch-Waxman administrative framework may be found at this Court's Memorandum Opinion addressing Defendants' motions for summary judgment on Plaintiffs' Actavis claims. See King Drug Co. of Florence, Inc., 2015 WL 356913, at *1-5.
Each of these settlement transactions have been characterized by Plaintiffs as anticompetitive reverse-payment settlement agreements that violate the antitrust laws. See Federal Trade Commission v. Actavis, Inc., 133 S. Ct. 2223 (2013). Furthermore, Cephalon is alleged to have violated the antitrust laws by procuring its Provigil patent by fraud on the PTO, and then enforcing said patent to keep competitors off of the market. See Walker Process Equip., Inc. v. Food Mach. & Chem. Corp., 382 U.S. 172 (1965).
B. Facts Pertinent to Class Certification
In support of class certification, Plaintiffs presented the expert testimony of Dr. Raymond S. Hartman. Plaintiffs posit that Dr. Hartman's methodology of measuring antitrust impact and aggregate damages demonstrates that the elements of Plaintiffs' claims can be satisfied through proof at trial that is common to the class. Dr. Hartman's methodology will be explored in greater detail herein, but to summarize, his methodology considered the amounts charged to Plaintiffs for branded and generic Provigil in the real world, and compared it to the amounts Plaintiffs would have been charged in the but-for world—that is, the world absent the allegedly anticompetitive conduct. Dr. Hartman opined that but-for the settlement agreements, the Generic Defendants would have launched their generic Provigil products at-risk in June 2006, which would have brought significant savings to TPPs and consumers. (Hartman Damages Exp. Rep., Apr. 26, 2011, ¶¶ 25-26.) According to Dr. Hartman, these overcharges constitute the relevant anticompetitive harm. (Id. at ¶¶ 42-44.) Dr. Hartman also considered the profits gained by Defendants during this time period, and compared them to the profits Defendants would have realized in the but-for world. According to Dr. Hartman, the difference between these two figures is an accurate measurement of Defendants' unjust enrichment. (Id. at ¶¶ 47-48.)
At the class certification hearing, the parties jointly agreed that, although various aspects of both Dr. Hartman's and Defendants' expert, Dr. Hughes' testimony were challenged by way of Daubert motions, those challenges would not be pressed for purposes of class certification. (Hrg. Tr., Mar. 24, 2015, pp. 5-9.)
To arrive at these figures, Dr. Hartman used yardsticks—data compiled from the generic launches of similar drugs—to calculate the rates of generic substitution and pricing of generic Provigil in the but-for world, and to demonstrate that consumers and TPPs would have paid less for their prescriptions if generic Provigil had entered the market. (Hrg. Tr., Mar. 24, 2015, pp. 76-87.) Dr. Hartman also considered data derived from the real-world launch of generic Provigil, which occurred in April 2012, to support and update his calculations. (Hartman Supp. Exp. Rep., Dec. 20, 2013, ¶¶ 6-7.)
Dr. Hartman explained that the use of yardsticks "is accepted everywhere in the industry, government research, academic research and litigation." (Hrg. Tr., Mar. 24, 2015, p. 77.) In fact, many of the yardstick drugs examined by Dr. Hartman were actually used by Cephalon to predict the impact of generic launch at or around the time of the settlement agreements. (Hartman Damages Exp. Rep., Apr. 26, 2011, ¶¶ 36-38.) Defendants do not challenge the use of yardsticks as a general matter.
Defendants presented competing testimony from Dr. James W. Hughes, who opined that significant variations throughout the pharmaceutical industry prevent Plaintiffs from being able to identify class members or prove antitrust impact and damages without substantial individualized inquiry. (Hughes Exp. Rep., June 10, 2011, ¶ 3.) Regarding TPPs, Dr. Hughes stated that establishing injury and the amount of damages will depend upon the particular contractual relationships each TPP has with its insureds, pharmacies, drug manufacturers, and pharmacy benefit managers ("PBMs"), all of which may vary over time. (Id. at ¶ 8.) Dr. Hughes also identified several categories of potentially uninjured persons who might otherwise fall within the class definition: (1) brand loyalists, or persons who choose to purchase the brand despite the availability of a generic; (2) consumers with the same copay for branded and generic drugs; (3) consumers who have not paid out-of-pocket for their prescriptions due to meeting an out-of-pocket maximum or an employer-funded health reimbursement account; (4) patients who would not have been prescribed Provigil in the but-for world due to decreased promotion; and (5) consumers whose insurers would place generic Provigil on a non-preferred tier. (Id. at ¶¶ 15-19.) As will be explored below, while some of these categories of uninjured persons have been excluded from the class, some remain.
Dr. Hartman described PBMs as entities that largely act as middlemen in managing pharmacy benefits. PBMs "organize, negotiate, manage contracts, [and] govern reimbursement" between and among TPPs and retail pharmacies. They may also, in rare instances, act as an insurer through a subsidiary. Where a PBM acts as a TPP, Dr. Hartman opined it would fall within the class definition. (Hrg. Tr., Mar. 24, 2015, pp. 72-75.)
During the class certification hearing, Dr. Hughes explained that simply excluding uninjured persons from the class definition would not prevent numerous individualized inquiries that would make class treatment inappropriate:
You're not going to be able to determine on any sort of average basis who the consumers are with flat copays. It's in the contract. You have to go to the contract to see. Certain brand loyal consumers get the question of . . . had the generic been available, who would have purchased the brand, who would have purchased the generic? Again, you have to look individually to see what individual consumers would have done.(Hrg. Tr., Mar. 25, 2015, pp. 38-39.) Dr. Hughes further opined that, without a means of identifying class members, and particularly uninjured persons within the class, Plaintiffs were unable to satisfy the ascertainability and predominance requirements under Rule 23. (Id. at pp. 43-44.)
C. Proposed Class Definitions
Plaintiffs seek certification of two groups of End Payors: (1) a class of End Payors asserting claims under state antitrust and consumer protection laws; and (2) a class of End Payors asserting claims for unjust enrichment under state law. The proposed class definitions are as follows: State Antitrust/Consumer Protection Class
In their reply memorandum in support of their motion for class certification, Plaintiffs withdrew their claim under the Tennessee Consumer Protection Act and the Wisconsin Deceptive Trade Practices Act. (Pls.' Rep., p. 17 n.48.) Additionally, Plaintiffs state that they are not presenting consumer protection claims in the District of Columbia, Iowa, or Mississippi or antitrust claims in Florida, Massachusetts or Nebraska. (Pls.' Mot. for Class Cert., p. 12 n.25.)
All persons or entities in Arizona, California, District of Columbia, Florida, Hawaii, Iowa, Kansas, Maine, Massachusetts, Michigan, Minnesota, Mississippi, Nebraska, Nevada, New Mexico, New York, North Carolina, North Dakota, South Dakota, Tennessee, Utah, Vermont, West Virginia, and Wisconsin who purchased Provigil and/or its generic equivalent intended for consumption by themselves, their families or their members, employees, plan participants, beneficiaries or insureds from June 1, 2006 through September 30, 2013.
State Unjust Enrichment Class
All persons or entities in Arizona, California, District of Columbia, Florida, Hawaii, Iowa, Kansas, Kentucky, Louisiana, Maine, Massachusetts, Michigan, Minnesota, Mississippi, Nebraska, Nevada, New Mexico, New York, North Carolina, North Dakota, South Dakota, Tennessee, Utah, Vermont, West Virginia, and Wisconsin who purchased Provigil and/or its generic equivalent modafinil, intended for consumption by themselves, their families or their members, employees, plan participants, beneficiaries or insureds from June 1, 2006 through September 30, 2013.
Plaintiffs do not include Kentucky's unjust enrichment law in their proposed jury instructions or state law charts. However, Plaintiffs do not state that they are abandoning their unjust enrichment claims under Kentucky law. Therefore, it appears that the omission is an oversight and I will include Kentucky's unjust enrichment law in my consideration.
The class definitions exclude the following persons and/or entities: (1) Defendants and their respective subsidiaries, affiliates and employees; (2) all governmental entities (except for government funded employee benefit plans); (3) all persons or entities who purchased modafinil, including Provigil, for purposes of resale or directly from Defendants to the extent and solely to the extent of such purpose for resale or as a direct purchaser; (4) insured individuals covered by plans imposing a flat dollar copay that was the same dollar amount for generic as for brand purchases; (5) individuals who bought only branded Provigil after generic modafinil became available ("brand loyalists"); (6) insured individuals who purchased only generic modafinil (not branded Provigil) pursuant to a fixed copay applicable to generic drugs; (7) fully insured health plans, i.e., plans that purchased insurance from another third-party payor covering 100% of the plan's reimbursement obligations to its members; and (8) all PBMs without capitation agreements. (Pls.' Br., pp. 12-13.)
II. STANDARD OF REVIEW
"The class action is an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only." Wal-Mart Stores v. Dukes, 131 S. Ct. 2541, 2550 (2011) (quoting Califano v. Yamasaki, 442 U.S. 682, 700-01 (1979)) (quotation marks omitted). In order to certify a class action, the plaintiffs bear the burden of proving by a preponderance of the evidence that the putative class satisfies all of the prerequisites identified in Federal Rule of Civil Procedure 23(a) and one of the subcategories of Rule 23(b). Fed. R. Civ. P. 23; In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 320 (3d Cir. 2008). "[P]roper analysis under Rule 23 requires rigorous consideration of all the evidence and arguments offered by the parties." Hydrogen Peroxide, 552 F.3d at 321. "[T]he court must resolve all factual or legal disputes relevant to class certification, even if they overlap with the merits—including disputes touching on elements of the cause of action." Id. at 307. "Weighing conflicting expert testimony at the certification stage is not only permissible; it may be integral to the rigorous analysis Rule 23 demands." Id. at 323 (citations omitted).
Subsection (a) of Rule 23 contains four prerequisites for any class action:
(1) the class is so numerous that joinder of all members is impracticable;Fed. R. Civ. P. 23(a).
(2) there are questions of law or fact common to the class;
(3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and
(4) the representative parties will fairly and adequately protect the interests of the class.
For certification under Rule 23(b)(3), the movant must also demonstrate "that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy." Fed. R. Civ. P. 23(b)(3). These requirements are known as predominance and superiority. In re Flonase Antitrust Litig., 284 F.R.D. 207, 215 (E.D. Pa. 2012).
In addition to these requirements, there are two "essential prerequisite[s]" to class certification under Rule 23(b)(3): (1) a "clearly defined class and set of claims, issues, or defenses to be given class treatment"; and (2) "the class must be currently and readily ascertainable based on objective criteria." Marcus v. BMW of N. Am., LLC, 687 F.3d 583, 592-93 (3d Cir. 2012) (citations omitted).
III. CLASS DEFINITION AND ASCERTAINABILITY
A. Clearly-Defined Class
"An order that certifies a class action must define the class and the class claims, issues, or defenses." Fed. R. Civ. P. 23(c)(1)(B). "[T]he text of the order or an incorporated opinion must include (1) a readily discernible, clear, and precise statement of the parameters defining the class or classes to be certified, and (2) a readily discernible, clear, and complete list of the claims, issues, or defenses to be treated on a class basis." Marcus, 687 F.3d at 591 (quoting Wachtel v. Guardian Life Ins. Co., 453 F.3d 179, 187 (3d Cir. 2006)). Clearly defining the contours of the class ensures that parties have clarity, and that class members understand their rights and make informed opt-out decisions. Id. at 591-92.
The parties have not raised any concerns about the class definitions, and my own review reveals that the classes are clearly defined. Plaintiffs have consistently alleged that, with limited exceptions, every person who purchased Provigil or its generic equivalent during the relevant time period and in the relevant jurisdictions suffered an overcharge, in violation of the state antitrust and consumer protection laws, and are due compensation based upon Defendants' unjust enrichment. As noted above, Plaintiffs have also excluded numerous categories of uninjured class members. Because the class definitions clearly identify the claims at issue, I find that the proposed class definitions satisfy the requirements of Rule 23(c)(1)(B).
The United States Court of Appeals for the Third Circuit has recently identified ascertainability as an important prerequisite for class treatment. "The ascertainability inquiry is two-fold, requiring a plaintiff to show that: (1) the class is 'defined with reference to objective criteria'; and (2) there is 'a reliable and administratively feasible mechanism for determining whether putative class members fall within the class definition.'" Byrd v. Aaron's Inc., 784 F.3d 154, 163 (3d Cir. 2015) (quoting Hayes v. Wal-Mart Stores, Inc., 725 F.3d 349, 355 (3d Cir. 2013)). "If class members are impossible to identify without extensive and individualized fact-finding or 'mini-trials,' then a class action is inappropriate." Marcus, 687 F.3d at 593. Ascertainability serves several goals: (1) "it eliminates serious administrative burdens that are incongruous with the efficiencies expected in a class action by insisting on the easy identification of class members"; (2) "it protects absent class members by facilitating the best notice practicable under Rule 23(c)(2) in a Rule 23(b)(3) action"; and (3) "it protects defendants by ensuring that those persons who will be bound by the final judgment are clearly identifiable." Id. (citations and quotation marks omitted).
The same rigorous analysis that a district court is required to undertake for Rule 23 requirements must be conducted with respect to ascertainability, and the plaintiff bears the burden of proving this element by a preponderance of the evidence. Carrera v. Bayer Corp., 727 F.3d 300, 306 (3d Cir. 2013) (quoting Marcus, 687 F.3d at 593). "A plaintiff may not merely propose a method of ascertaining a class without any evidentiary support that the method will be successful[,]" and a party's assurance that it will be able to establish ascertainability at some point in the future is insufficient. Id. at 306; see also Shelton v. Bledsoe, 775 F.3d 554, 559 (3d Cir. 2015) (class members must be "identifiable at the moment of certification").
The plaintiffs' method for identifying class members must be "administratively feasible," meaning "that identifying class members is a manageable process that does not require much, if any individual factual inquiry." Id. at 307-08 (citations omitted). "Ascertainability provides due process by requiring that a defendant be able to test the reliability of the evidence submitted to prove class membership." Id. at 307. It is insufficient to rely solely on a potential class members' "say so" that they belong within the class through affidavits or declarations. Marcus, 687 F.3d at 594.
The Third Circuit has recently issued several opinions on the ascertainability requirement, and last year a district court in Tennessee was confronted with class certification issues in a Hatch-Waxman reverse-payment settlement case very similar to the case before me. An examination of this precedent is instructive to analyzing the complex class certification concerns at issue here.
1. Recent Precedent - Ascertainability
The Third Circuit first provided a detailed analysis of the ascertainability requirement in Marcus v. BMW of North America, LLC, 687 F.3d 583 (3d Cir. 2012), wherein the named plaintiff had brought fraud, breach of warranty, and breach of contract claims against a tire manufacturer and a car company. The plaintiff alleged that the manufacturer's run-flat tires ("RFTs") that had been placed on his leased BMW were defective because the tires were highly susceptible to flats, were unable to be repaired, and were exorbitantly priced. Id. at 588. The plaintiff sought to certify a class under Rule 23(b)(3) on behalf of all purchasers and lessees of certain model-year BMWs equipped with the RFTs sold or leased in New Jersey with tires that had gone flat or been replaced. Id.
The Third Circuit determined that the plaintiff had failed to demonstrate that the proposed class was ascertainable because there was no database or manifest indicating which BMWs had been sold and driven off of the lot with RFT tires. Id. at 593-94. In denying the certification motion the Court found that, even if the cars with the tires at issue could be identified, an individualized inquiry would need to be undertaken to determine whether that consumer's RFTs had gone flat and been replaced, as was required by the class definition. Id. at 594. The court further found that simply submitting affidavits from potential class members stating that they belong within the class was not a sufficiently reliable methodology. Id.
The Third Circuit expanded upon the ascertainability requirement in Carrera v. Bayer Corp., 727 F.3d 300 (2013). There, an indirect purchaser of Bayer's One-A-Day WeightSmart product brought claims against Bayer under the Florida Deceptive and Unfair Trade Practices Act and sought class certification on behalf of all persons who purchased WeightSmart in Florida. Bayer argued that it did not maintain a list of class members because the class consisted of indirect purchasers, and potential class members were unlikely to have documentary proof of their purchase. Id. at 304. The plaintiff responded that he could identify class members through retailer records of online sales and sales made with store loyalty/rewards cards, or through class member affidavits. Id. The Third Circuit rejected this proposal and determined that class certification was inappropriate for lack of ascertainability because the plaintiff had failed to present evidence that the retailers had records of WeightSmart purchases during the relevant period, or that those records would be able to identify purchasers of WeightSmart. Id. at 308-09. The Carrera court reaffirmed its pronouncement in Marcus that affidavits from potential class members were insufficient. Id. at 309.
Most recently, in Byrd v. Aaron's Inc., 784 F.3d 154 (3d Cir. 2015), the Third Circuit sought to clarify the ascertainability requirement, citing some confusion among district courts and litigants. In Byrd, the plaintiffs, lessees of a laptop computer from the defendant, Aaron's Inc., alleged that spyware had been placed on their computer and that the defendants had activated that spyware to obtain pictures from the laptop's camera, as well as screenshots, without the plaintiffs' knowledge or consent. Citing violations of the Electronic Communications Privacy Act, 18 U.S.C. § 2511, the plaintiffs sought to certify a class containing all persons who leased or purchased a computer from the defendant, as well as their household members, where the computer contained spyware that was activated without the person's consent. Id. at 159-60. The district court denied certification for lack of ascertainability, finding that the class was underinclusive, overly broad, and did not adequately define "household member." Id. at 160-61.
The Third Circuit reversed, holding that the plaintiffs had met their burden through defendant's records, which provided the identities and addresses of 895 class members. The court clarified that a plaintiff does not need to actually identify all class members at the time of class certification to demonstrate ascertainability, "a plaintiff need only show that 'class members can be identified.'" Id. at 163, 166-67 (quoting Carrera, 727 F.3d at 308 n.2) (emphasis in original). "Accordingly, there is no records requirement," although a plaintiff must still provide evidentiary support that the proposed method of ascertaining the identities of class members will be successful and administratively feasible. Id. at 164.
The Byrd opinion also found that the district court's concern with underinclusiveness was not an appropriate ascertainability inquiry, and that the class potentially being overbroad was more appropriately considered as a predominance issue. Id. at 166-69. Further, the court disagreed with the district court's determination that the "household members" included in the class were not ascertainable, noting that the plaintiffs had presented government documents that could be used to identify those individuals. Id. at 169-70. The court remarked that:
[t]here will always be some level of inquiry required to verify that a person is a member of the class; for example, a person's statement that she owned or leased an Aspen Way computer would eventually require anyone charged with administering the fund resulting from a successful class action to ensure that person is actually among the 895 customers identified by the Byrds. Such a process of identification does not require a "mini-trial," nor does it amount to "individualized fact-finding." . . . "[T]he size of a potential class and the need to review individual files to identify its members are not reasons to deny class certification."Id. at 170-71 (quoting Carrera, 727 F.3d at 307; Young v. Nationwide Mut. Ins. Co., 693 F.3d 532, 539-40 (6th Cir. 2012)).
In re Skelaxin (Metaxalone) Antitrust Litigation, 299 F.R.D. 555 (E.D. Tenn. 2014), closely resembles the facts before me, as it involved antitrust allegations brought by a number of end payors who purchased a branded drug and alleged that the brand manufacturer illegally delayed market entry of the generic drug. Relying in part on the Third Circuit precedent discussed above, the district court noted that "[i]f class members are impossible to identify without extensive and individualized fact-finding or mini-trials then a class action is inappropriate." Id. at 567 (quoting Marcus, 687 F.3d at 593). The court found that "[g]iven the discrepancy between End Payors' expert's testimony and the class definition, the Court cannot determine which entities or individuals are members of the class and which are not." Id. at 560. To make this determination, the court found, would require a transaction-by-transaction inquiry that was incompatible with a Rule 23(b)(3) class action. Id.
The court described the "end payor" class as persons who purchased the brand drug "for consumption and other than for resale." In other words, end payors were persons or entities who purchased the brand drug "for their own use and by logical extension were the final consumers who absorbed the overcharge[.]" Id. at 562.
In reaching these conclusions, the district court noted the various links in the pharmaceutical supply chain, and that numerous entities could contribute all or part of the cost of any particular prescription. Thus, the court found it impossible to determine whether an end payor belonged within the class without considering "the individual contractual relationships underlying each transaction." Id. at 569. Indeed, much like the record before me, the plaintiffs' expert in Skelaxin had opined that identifying who was injured in any particular transaction was a claims administration issue. The district court rejected this testimony, finding that "the issue with End Payors' class is not whether a purchaser was damaged in each individual transaction; the issue is whether a purchaser constitutes a class member." Id. at 570. "Until proceeding through each transaction and resolving factual disputes about who 'bears the burden' of the price in that transaction, the Court [could not] say who [was] a member of the class." Id. at 571. In short, the district court concluded that ascertaining class members would entail "individual inquiry into contracts covering millions of purchases." Id. at 570.
2. Ascertainability of the End Payor Class
As set forth previously, the proposed class definition essentially includes persons from numerous states who purchased Provigil (or its generic equivalent) from June 1, 2006 through September 30, 2013. Eight categories of persons and entities are specifically excluded from the proposed class.
Defendants raise two primary arguments in support of their position that the proposed class is not ascertainable. The first argument is that Plaintiffs have failed to present any databases or other records that would identify who belongs within the class, let alone who is excluded. Defendants point out that they do not have such records, as the class is comprised of persons who purchased Provigil from other entities in the distribution chain, such as retailers. Defendants stress that Plaintiffs' class certification expert, Dr. Hartman, has clearly acknowledged that he has no methodology for identifying class members. Defendants also assert that even if Plaintiffs had presented a methodology of identifying class members, that process would be overwhelmed by individualized inquiries and would not be administratively feasible.
Defendants note that Plaintiffs failed to address the prerequisite of ascertainability in their opening brief in support of class certification, and urge that this should amount to a waiver. (Defs.' Br., p. 13 n.3.) While I agree with Defendants that Plaintiffs did not sufficiently address ascertainability in their opening brief, I will not resolve this motion on the basis of a waiver and will carefully consider the arguments Plaintiffs raise in their reply brief, supplemental briefing, and at oral argument.
According to Defendants, the difficulties in identifying class members are further amplified by the multiple entities that may be involved in each pharmaceutical transaction who may have paid for all or a portion of the drug. These possible End Payors could include TPPs, health plan sponsors, PBMs, and the consumers themselves. Defendants stress that many End Payors, particularly consumers, will not have receipts or other ways of proving a purchase of Provigil during the relevant time period.
Plaintiffs respond that they have demonstrated ascertainability by proffering "a well-defined class with certain clearly articulated exclusions." (Pls.' Reply, p. 6.) They maintain that other courts within this district have certified classes of end payors without expressing concerns about ascertainability, and the result here should be the same. Plaintiffs attempt to distinguish Marcus and Carrera by pointing out that those cases arose in the "products liability context" where a complete lack of reliable and objective records existed to identify class members. According to Plaintiffs, that is not the case here, where "comprehensive records documenting any purchase of Provigil can be easily obtained from any number of reliable, objective sources" such as "IMS-NPA data, insurance companies, healthcare plans, health and welfare funds, pharmacies, as well as consumers." (Pls.' Reply, p. 7, n.18.)
Despite Plaintiffs' assurances, a careful review of the record before me demonstrates that Plaintiffs have not met their burden of demonstrating ascertainability.
a. Reliable Methodology for Identifying Class Members
In support of their position that records are readily available, Plaintiffs point to the customer history of named Plaintiff Shirley Panebianco, which was obtained from Bridge Pharmacy. This document lists the various prescriptions that Panebianco had filled during the relevant time period, as well as the amount paid out of pocket and the amount covered by her health insurance plan. (See Pls.' Reply, Meltzer Decl., Ex. 27.) Plaintiffs also point to a short chart of claims data that was attached to their reply brief. This document does not identify any individuals, but rather lists patients by number and identifies the state, pharmacy, and/or city in which the prescription was filled, the submitted cost of the prescription and the copayment paid by the consumer. (See id. at Ex. 28.)
This chart was also submitted to rebut Defendants' assertion that Plaintiffs only have standing to sue in New York and Pennsylvania. However, Plaintiffs only provided evidence of standing through purchases and reimbursements in fourteen out of the twenty-six states, which may raise some standing concerns. In any event, because I find that class certification should be denied for a variety of other reasons, I need not address this issue further.
Aside from these two exhibits—one which lists the prescriptions of one consumer, and the other, which identifies consumers by number as opposed to including their identities—Plaintiffs have not presented any evidence that these records can be utilized to identify class members. While one consumer's prescription history dating back to 2006 has been presented, the record before me contains no evidence as to whether other pharmacies kept reliable records of this same type of patient data over that time period. Plaintiffs' counsel assures the Court that such records were available. Carrera mandates, however, that it is insufficient to simply make assurances that records are readily available without providing evidence that "retailer records in this case can be used to identify class members." Carrera, 727 F.3d at 308. And the record before me is certainly different from Byrd, where the defendant maintained a list of the 895 persons that belonged within the class.
Dr. Hartman does not cure these deficiencies, and indeed seemed to dispute Plaintiffs' counsel's assurance that readily available records could be used to ascertain the identities of the class members. When asked if a database or other collection of records existed from which the identities of class members could be derived, Dr. Hartman testified that "[w]ithout them coming forward on their own[,]" he was unaware of any such records. (Hrg. Tr., Mar. 24, 2015, p. 182.) Dr. Hartman stated numerous times that "at this stage of [his] analysis, [he] ha[d]n't been asked to" identify the class members and that he could simply verify that they belonged in the class "when class members come forward with their claims." (See id. at pp. 149-51; see also id. at p. 162 ("The Court: Do you have a methodology to ascertain the members of the class presently? . . . [Dr. Hartman]: No, I haven't been asked to do that yet").)
Dr. Hartman's testimony on this issue was as follows:
Q. Sir, are you aware of any records from which the identity of the members of the class can be derived?(Hrg. Tr., Mar. 24, 2015, p. 182.)
A. Without them coming forward on their own?
Q. Without them coming forward on their own.
A. I would have to investigate that. None that could be easily linked to mine.
Q. So the IMS system doesn't - -
A. The IMS system does not - - it does it by type of payer. I can break out copay, co-insurance, but individual plans, it does not identify those that I know of. There may be product lines that have that, but I have not used them.
Q. As you sit here now, you're not aware of those records?
A. As I sit here now, they may exist. IMS is always improving its product lines. But I've never used them if they do exist.
Although Plaintiffs continue to stress that they need not present a methodology at this time, and that such an analysis would be more appropriate during a damages allocation, this argument is contradicted by clear pronouncements from the Third Circuit. Plaintiffs must, at the time of class certification, present a methodology to identify class members, and prove by a preponderance of the evidence that such methodology will be effective and will not require extensive individualized inquiry and mini-trials. See Carrera, 727 F.3d at 306 (the class must be "currently and readily ascertainable based on objective criteria") (emphasis added). Plaintiffs' argument that other courts within this district and elsewhere have certified similar classes of End Payors without being concerned with ascertainability is diminished by the fact that these cases were decided before Marcus, Carrera and Byrd. Indeed, the two cases cited by Plaintiffs, In re Flonase Antitrust Litigation, 284 F.R.D. 207 (E.D. Pa. 2012) and In re Wellbutrin XL Antitrust Litigation, 282 F.R.D. 126 (E.D. Pa. 2011), never even addressed the issue of ascertainability.
Plaintiffs' reliance on In re Nexium Antitrust Litigation, 777 F.3d 9 (1st Cir. 2015), does not assist their position. It is accurate that the United States Court of Appeals for the First Circuit held that the plaintiffs' failure to identify a methodology for distinguishing between injured and uninjured class members did not preclude class certification. However, in reaching this conclusion, the court found that "[w]hile it is true that a proper mechanism for exclusion of [uninjured] consumers has not yet been proposed, plaintiffs' expert made no concession that such a mechanism could not be developed." Id. at 20.
My understanding of Third Circuit precedent, particularly Hydrogen Peroxide, and subsequently Carrera, is that much more is needed in this Circuit than an expert who did not concede that an ascertainability mechanism could not be developed. See Hydrogen Peroxide, 552 F.3d at 318 (Rigorous analysis required, and assurances that a Plaintiff "intends or plans to meet the requirements is insufficient"); Carrera, 727 F.3d at 306 (same). Put another way, plans to create a methodology at a later date do not satisfy the rigorous analysis insisted upon by the Third Circuit and I do not read Byrd to alter these requirements. Moreover, the First Circuit in In re Nexium was satisfied that class member testimony in the form of an affidavit or declaration would meet the ascertainability requirement. As noted above, this method has been squarely rejected in this Circuit. Compare In re Nexium, 777 F.3d at 20 (categorizing affidavits and declarations as an acceptable means of identification), with Marcus, 687 F.3d at 594 ("Forcing [defendants] to accept as true absent persons' declarations that they are members of the class, without further indicia of reliability, would have serious due process implications").
For all of the above reasons, I agree with Defendants that Plaintiffs have failed to present evidence of a reliable "mechanism for determining whether putative class members fall within the class definition." Byrd, 784 F.3d at 163.
b. Administratively Feasible Methodology for Identifying Class Members
Plaintiffs have also not met their burden of establishing that any methodology for identifying class members would be administratively feasible. Dr. Hughes credibly testified that he was unaware of any administratively feasible approach that would allow Plaintiffs to distinguish class members from persons that fell within an exclusion, and that Dr. Hartman's yardstick methodology did not address this concern. (Hrg. Tr., Mar. 25, 2015, pp. 43-46, 80-81.) I find Dr. Hughes' analysis on this issue convincing.
I have carefully considered the guidance provided in Byrd that overbreadth of a class raises questions of predominance as opposed to ascertainability. Byrd, 784 F.3d at 168; see also Grandalski v. Quest Diagnostics Inc., 767 F.3d 175, 184-85 (3d Cir. 2014) (by focusing on the individualized inquiry required to establish harm in its ascertainability analysis, the district court conflated predominance with ascertainability). However, by choosing to define its class with eight specific exclusions, Plaintiffs have created the need for a structured, multi-stepped, individualized fact-finding process in order to determine which individuals would fall within the class definition and which would fall within one of the eight exclusions. See In re Skelaxin, 299 F.R.D. at 570 (similar considerations raised questions about "whether a purchaser constitutes a class member") (emphasis in original).
At oral argument, Plaintiffs' counsel acknowledged that, in the absence of a database or comprehensive list of class members, sending notice to potential consumer class members would necessitate numerous steps. According to Plaintiffs' counsel, this process would first require a plan to be developed by a consulting company, wherein notices would be sent to TPPs. While Plaintiffs' counsel assured me that a list of TPP class members had been compiled, this list was not offered into evidence and thus its reliability could not be challenged by Defendants or examined by the Court. Plaintiffs' counsel next described setting up a website which would allow the consumer class members to make themselves known to Plaintiffs' counsel. Thereafter, individual records would need to be compiled to determine whether that person actually belongs within the class, or if they fall within an exclusion. (Oral Arg. Tr., May 6, 2015, pp. 7-12.) Dr. Hartman acknowledged that in order to identify consumers and TPPs that fell within the class, he would need to conduct a detailed analysis of the contracts between various entities and also examine the purchasing history of the individual. (See Hrg. Tr., Mar. 24, 2015, pp. 175-76, 195-96.)
To the extent that this multi-step notice process acts as a means of identifying class members, it entails "extensive and individualized fact-finding or 'mini-trials'" that Byrd stated would make class certification inappropriate. Byrd, 784 F.3d at 163 (citation omitted). Indeed, Plaintiffs' ascertainability problems are compounded by the complex nature of the pharmaceutical and insurance industries. Many individualized questions must be answered in order to determine whether an individual falls within the class definition, such as: Was the individual a brand loyalist?; What was the individual's copay for branded as opposed to generic Provigil?; Was the individual a member of a group plan that provided full or partial reimbursement?; What, if any, rebates factored into the consumer payment?
These very same hurdles also caused concern to the district court in In re Skelaxin. Under very similar circumstances, the court denied class certification, as it required "consideration of the individual contractual relationships underlying each transaction." In re Skelaxin, 299 F.R.D. at 569. I share those concerns and agree with Judge Collier when he stated:
[U]ntil proceeding through each transaction and resolving factual disputes about who "bears the burden" of the price in that transaction, the Court cannot say who is a member of the class, that is, who has paid or reimbursed a portion of the purchase price. Although the class is circumscribed to only those entities who paid for their own consumption or the consumption of their constituents, End Payors' expert testified that, depending on the circumstances of each transaction, an end payor may be one or more entities or individuals sharing the burden.Id. at 571.
Plaintiffs argue that Defendants are confusing claims administration with ascertainability, and insist that the court in In re Skelaxin made this same error. They point to an antitrust case involving allegations of delayed generic entry where class counsel successfully provided settlement checks to 816,000 consumers and 2,500 TPPs during the claims administration of a settlement class. That claims administration was conducted by gathering and analyzing records from insurers and pharmacies to determine who was owed a portion of the settlement. See In re Tricor Indirect Purchaser Antitrust Litig., C.A. No. 05-360 (D. Del.). However, this procedure was used in the context of a settlement, which raises different certification issues. See Carrera, 727 F.3d at 308 n.4 (questioning whether methods used to identify purchasers in a settlement class would be relevant to resolving ascertainability in a litigation class); see also In re Skelaxin, 299 F.R.D. at 571-72 ("The proposed 'claims administration' procedure is wholly post-hoc whittling of the class in the context of settlement. This methodology does nothing for the individual fact-finding required if this case were put to a jury") (emphasis in original).
I also recognize and have considered Byrd's conclusion that verification of a person's membership in the class for purposes of fund administration does not require a "mini-trial." See Byrd, 784 F.3d at 170. But again, fund administration and the rigorous analysis required by the Third Circuit for establishing ascertainability are separate and distinct. My concerns about ascertainability focus on whether Plaintiffs can reliably identify class members at the outset. By contrast, the fund administration process would occur at the conclusion of litigation, and simply verify that any particular consumer or TPP is indeed one of the previously-identified members of the class. See id. (a person's statement that she belongs within the class would need to be verified at the conclusion of successful litigation to ensure she is actually among the previously-identified class members).
In summary, Plaintiffs have failed to present any evidence that they have developed a methodology for ascertaining the identities of class members, aside from simply assuring the court that records of Provigil prescriptions exist. Nor have Plaintiffs presented any evidence to demonstrate that it is possible to ascertain class members in an administratively feasible manner without highly individualized inquiry. Accordingly, Plaintiffs have failed to meet their burden of satisfying the ascertainability requirement.
IV. RULE 23(A) REQUIREMENTS
A. Rule 23(a)(1) - Numerosity
Defendants do not dispute Plaintiffs' ability to meet their burden as to the numerosity requirement, where Plaintiffs must establish that "the class is so numerous that joinder of all members is impracticable." Fed. R. Civ. P. 23(a)(1). Whether joinder is impracticable requires consideration of the number of class members, expediency, and the inconvenience of trying individual cases. Jackson v. Se. Pa. Transp. Auth., 260 F.R.D. 168, 186 (E.D. Pa. 2009) (citations omitted). While there is no precise number that will meet this requirement, classes in excess of forty members tend to satisfy numerosity. Id. Dr. Hartman has identified in excess of five million total Provigil prescriptions filled in the relevant jurisdictions from 2006 through January 2011. (Hartman Damages Exp. Rep., Apr. 26, 2011, Attachment E.1.e.) Therefore, I find that Plaintiffs have satisfied the numerosity requirement. See In re Terazosin Hydrochloride Antitrust Litig., 220 F.R.D. 672, 685 n.21 (S.D. Fla. 2004) ("once the good faith estimate of the class size reaches the thousands, the joinder impracticability test is satisfied").
B. Rule 23(a)(2) - Commonality
The commonality requirement is met where the members of the class's claims "depend upon a common contention" that is "capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke." Dukes, 131 S. Ct. at 2551. "Because the requirement may be satisfied by a single common issue, it is easily met." Baby Neal v. Casey, 43 F.3d 48, 56 (3d Cir. 1994).
As in other similar antitrust cases, "[e]ach class member's claims depend on whether or not the defendants unlawfully engaged in anticompetitive behavior to limit the entry of generic competitors," which will require evidence common to the class. In re Wellbutrin XL, 282 F.R.D. at 137 (citing In re Warfarin Sodium Antitrust Litig., 391 F.3d 516, 528 (3d Cir. 2004)). Defendants do not dispute Plaintiffs' ability to establish commonality, and I find that it has been satisfied.
C. Rule 23(a)(3) - Typicality
"The typicality inquiry is intended to assess whether the action can be efficiently maintained as a class and whether the named plaintiffs have incentives that align with those of absent class members so as to assure that the absentees' interests will be fairly represented." Baby Neal, 43 F.3d at 57 (citations omitted). Typicality exists "[i]f the representative's claims and those of the absent class members arise from the same course of conduct and are based on the same legal theories . . . regardless of factual differences underlying the individual claims." In re Wellbutrin XL, 282 F.R.D. at 138 (citing Baby Neal, 43 F.3d at 57-58). The court must consider "whether the named plaintiff's individual circumstances are markedly different or . . . the legal theory upon which the claims are based differs from that upon which the claims of other class members will perforce be based." Hassine v. Jeffes, 846 F.2d 169, 177 (3d Cir. 1998) (quotation marks and citations omitted). "The typicality requirement is intended to preclude certification of those cases where the legal theories of the named plaintiffs potentially conflict with those of the absentees." Georgine v. Amchem Prods., Inc., 83 F.3d 610, 631 (3d Cir. 1996) (citations omitted).
Plaintiffs argue that typicality has been established because both the named and absent class members maintain the same claims and legal theories—that the allegedly anticompetitive conduct of Cephalon and the Generic Defendants constituted a violation of state antitrust, consumer protection and unjust enrichment laws. Plaintiffs allege that the Defendants' conduct injured both the class representatives and the absent class members through overcharges and unjust enrichment. Defendants respond that a conflict of interest exists between class members that may share an overcharge, and that this conflict defeats typicality. I do not agree. See In re Cardizem CD Antitrust Litig., 200 F.R.D. 326, 337 (E.D. Mich. 2001) (quoting In re S. Cent. States Bakery Prods. Antitrust Litig., 86 F.R.D. 407, 418 (M.D. La. 1980)) ("A naked allegation of antagonism cannot defeat class certification; there must be an actual showing of a real probability of a potential conflict which goes to the subject matter of the suit"); (see also Section IV. D. infra). Plaintiffs' and the absent class members' claims are based on largely identical legal theories and focus heavily on Defendants' course of conduct. See In re Flonase, 284 F.R.D. at 218 (finding typicality had been established where the indirect purchasers all alleged "that the same unlawful conduct injured both the class representatives and the absent class members"). Therefore, I find that Plaintiffs have satisfied the typicality requirement.
Conflicts of interest between and among class members are often addressed in the context of adequacy of representation. See Dewey v. Volkswagen Aktiengesellschaft, 681 F.3d 170, 183-84 (3d Cir. 2012) ("Certain intra-class conflicts may cause the interests of the representative plaintiffs to diverge from those of the unnamed class members. The adequacy requirement is designed to ferret out such conflicts of interest.") (citation and quotation marks omitted). Accordingly, I will address Defendants' conflict arguments in more detail in section IV. D. infra.
D. Rule 23(a)(4) - Adequacy of Representation
Adequacy of representation has two prongs: (1) "the plaintiff's attorney must be qualified, experienced, and generally able to conduct the proposed litigation"; and (2) "the plaintiff must not have interests antagonistic to those of the class." Wetzel v. Liberty Mut. Ins. Co., 508 F.2d 239, 247 (3d Cir. 1975) (citation omitted). The adequacy requirement necessitates that the court consider whether conflicts of interest exist between named parties and those they represent. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 625 (1997).
Defendants do not challenge the adequacy of Plaintiffs' counsel, all of whom have extensive experience handling complex class action litigation, including cases in the antitrust context. (See Pls.' Mot., Meltzer Decl., Exs. 15-17.) Defendants do, however, argue that conflicts exist between and among the named Plaintiffs and absent class members that defeat adequacy of representation. First, Defendants argue that conflicts exist between certain class members who actually benefitted from the absence of generic Provigil and those who were allegedly harmed by the absence of generic Provigil. For instance, Dr. Hughes opined that brand loyalists often benefit from a delay in generic entry because their copayment may increase when a generic product becomes available. (Hughes Supp. Exp. Rep., Dec. 20, 2013, ¶¶ 10, 49-51.) Plaintiffs respond, and I agree, that brand loyalists have been excluded from the class definition, and therefore no such conflict exists.
Next, Defendants argue that conflicts exist between the various End Payors who may bear the cost of any particular Provigil prescription through cost-sharing. For example, a consumer may pay a portion of the cost of Provigil through a copay, while his insurer, a TPP, would pay the remainder. Defendants assert that this framework, where alleged overcharges are split among class members, would result in class members battling against each other for recoveries, resulting in a conflict of interest. Plaintiffs respond that shared portions of an overcharge does not constitute a conflict, but is instead an integral part of the damages allocation process.
The court in In re Cardizem CD Antitrust Litigation, 200 F.R.D. 326 (E.D. Mich. 2001), considered a similar argument, and found that a hypothetical conflict regarding apportionment of damages was insufficient to defeat certification. "Each class member has the same interest in maximizing the aggregate amount of classwide damages." Id. at 337 (quoting In re NASDAQ Market-Makers Antitrust Litig., 169 F.R.D. 493, 512 (S.D.N.Y. 1996)) (quotation marks omitted). "A naked allegation of antagonism cannot defeat class certification; there must be an actual showing of a real probability of a potential conflict which goes to the subject matter of the suit." Id. (quoting In re S. Cent. States Bakery Prods. Antitrust Litig., 86 F.R.D. 407, 418 (M.D. La. 1980)); see also Kohen v. Pacific Inv. Mgmt. Co. LLC, 571 F.3d 672, 680 (7th Cir. 2009) ("To deny class certification now, because of a potential conflict of interest that may not become actual, would be premature").
Defendants have not presented evidence to establish a real probability of a conflict of interest among class members. All class members have a common interest in maximizing aggregate classwide damages. As in In re Cardizem, I find that the risk of conflicts during a damages allocation is speculative, and in any event, if conflicts were to arise at that stage of the litigation, those conflicts could be alleviated through the creation of subclasses of consumers and insurers. Therefore, I find that Plaintiffs have met the adequacy of representation requirement.
V. ANTITRUST CLAIMS - PREDOMINANCE
The predominance requirement considers "whether proposed classes are sufficiently cohesive to warrant adjudication by representation." Amchem, 521 U.S. at 623. In order to certify a class under Rule 23(b)(3), "questions of law or fact common to class members [must] predominate over any questions affecting only individual class members." Fed. R. Civ. P. 23(b)(3). While commonality and predominance present similar considerations, the predominance standard is "far more demanding." Hydrogen Peroxide, 552 F.3d at 311 (citations omitted).
"Rule 23(b)(3) requires a showing that questions common to the class predominate, not that those questions will be answered, on the merits, in favor of the class." Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S. Ct. 1184, 1191 (2013) (emphasis in original). Individual questions need not be absent, so long as common questions predominate. Id. at 1196.
When conducting a predominance inquiry, the court must consider the elements of the underlying cause of action. In re Flonase, 284 F.R.D. at 219 (quoting John Fund, Inc. v. Halliburton Co., 131 S. Ct. 2179, 2184 (2011)). In order to prevail on their state law antitrust claims, Plaintiffs must prove: (1) a violation of the state antitrust laws; (2) individual injury resulting from the violation, also known as antitrust impact; and (3) measurable damages. See Hydrogen Peroxide, 552 F.3d at 311.
As noted below, Plaintiffs failed to address the question of which states' antitrust laws should govern. However, under any potentially applicable law, Plaintiffs must demonstrate antitrust impact. See Hydrogen Peroxide, 552 F.3d at 311. As I have concluded that Plaintiffs have not demonstrated that antitrust impact is capable of proof through common, class-wide evidence, I need not reach the choice of law question.
Defendants largely do not dispute that the evidence Plaintiffs would present at trial to establish a violation of the state antitrust laws would be common to the class. Instead, Defendants argue that Plaintiffs are not able to demonstrate antitrust impact or damages using class-wide evidence because determining whether a particular purchaser was harmed and the extent of that harm will necessitate individualized inquiries.
I agree with Plaintiffs that evidence of Defendants' alleged violations of the state antitrust laws would be common to the class. See In re Wellbutrin XL, 282 F.R.D. at 140 ("If each class member pursued its claims individually, the class member would have to prove the same antitrust and consumer protection violations using the same documents, witnesses, and other evidence").
A. Antitrust Impact
"In antitrust cases, impact often is critically important for the purposes of evaluating Rule 23(b)(3)'s predominance requirement because it is an element of the claim that may call for individual, as opposed to common, proof." Id. (citations omitted). "[T]he task for plaintiffs at class certification is to demonstrate that the element of antitrust impact is capable of proof at trial through evidence that is common to the class rather than individual to its members." Id. at 311-12. To resolve this issue, the court must conduct a "rigorous assessment of the available evidence and the method or methods by which plaintiffs propose to use the evidence to prove impact at trial." Id. at 312 (citations omitted).
Plaintiffs largely rely upon the reports and testimony of Dr. Hartman in demonstrating that antitrust impact can be established using proof common to the class. Plaintiffs argue that Defendants' actions in unlawfully keeping generic Provigil off of the market caused universal injury to the class in several ways. First, by entering into the reverse-payment settlement agreements, Defendants allegedly kept generic Provigil off of the market through 2012, when Plaintiffs argue it would have otherwise entered the market, bringing lower prices to the class members, in 2006. (Hartman Market Def. & Impact Exp. Rep., Apr. 26, 2011, ¶¶ 128-29.) Thereafter, when generic Provigil did enter the market in 2012, Plaintiffs allege that those who switched to the generic paid more for generic Provigil than they would have paid in the but-for world—that is, if the generic had come onto the market in 2006. (Hartman Supp. Exp. Rep., Dec. 20, 2013, ¶ 7, Attach. E.3.a, b.) Plaintiffs argue that common proof of these overcharges can establish antitrust impact for the class as a whole.
Defendants respond that significant variations within the class and large groups of uninjured class members prevent Plaintiffs from proving antitrust impact on a class-wide basis. I will first address Plaintiffs' evidence of class-wide antitrust impact, and then address Defendants' arguments that individualized evidence of impact would overwhelm common evidence.
1. Plaintiffs' Class-Wide Evidence of Antitrust Impact
To meet their burden of demonstrating impact to the class, Plaintiffs must demonstrate that branded and generic Provigil prices would have been lower absent Defendants' conduct, which resulted in overcharges to the class members, and must be able to do so through common evidence. See In re Flonase, 284 F.R.D. at 221. Plaintiffs rely on the testimony of Dr. Hartman to meet these standards.
Dr. Hartman reports that by virtue of entering into the settlement agreements with the Generic Defendants, Cephalon was able to maintain a monopoly on the Provigil market and generate monopoly profits for an extended period of time. (Hartman Market Def. & Impact Exp. Rep., Apr. 26, 2011, ¶¶ 28, 129-30.) Dr. Hartman opined that absent the agreements, all of the Generic Defendants would have launched their generic Provigil products at-risk in 2006. (Id. at ¶¶ 35, 129-30.) He further explained that generic pharmaceuticals, including generic Provigil, have a lower cost to consumers and TPPs than their brand-name counterparts. (Id. at ¶¶ 62-66.) Dr. Hartman reasoned that this is why the Hatch-Waxman Act was implemented—to increase the availability and procompetitive effects of generic pharmaceuticals. (Id. at ¶¶ 55-61.)
Dr. Hartman's reports detail the significant cost savings to end payors when generic pharmaceuticals enter the market, and how those savings increase with each additional generic competitor. (Id. at ¶¶ 131-34.) He also examined the real-world effects of generic competition on the Provigil market by reviewing data derived from generic Provigil's market entry in 2012. His evaluation of the data suggests that 99% of consumers switched to the generic version of Provigil, and that they have experienced savings as a result.
According to Dr. Hartman, but-for the settlement agreements by and among Defendants, the savings to consumers who switched to generic Provigil would have been greater. (Hartman Supp. Exp. Rep., Dec. 20, 2013, ¶ 7, Attach. E.3.a, b.) This is because the price charged for a generic upon release is impacted by the price of the brand-name drug. As the price of branded Provigil increased from 2006 through 2012, the amount charged by generic manufacturers in 2012 also increased to a higher level than it would have been in 2006. Therefore, according to Dr. Hartman, even those consumers who were able to purchase generic Provigil in 2012 suffered overcharges. (Hrg. Tr., Mar. 24, 2015, pp. 66-68, 87-91.) To demonstrate the difference between the but-for cost of Provigil and the prices paid by consumers for branded and generic Provigil in the actual world, Dr. Hartman used the yardstick method described above, which has been utilized in other antitrust cases. See In re Linerboard Antitrust Litig., 305 F.3d 145, 153-55 (3d Cir. 2002) (accepting the use of the yardstick method for antitrust impact and damages on a class-wide basis); In re Flonase, 284 F.R.D. at 220 (same); In re Wellbutrin XL, 282 F.R.D. at 140-41 (same).
Plaintiffs assert that these facts are common to all class members and are sufficient to establish antitrust impact on a class-wide basis. (Hartman Market Def. & Impact Exp. Rep., Apr. 26, 2011, ¶ 135.)
2. Defendants' Challenges to Class-Wide Evidence of Antitrust Impact
Defendants argue that individualized evidence regarding antitrust impact overwhelms common evidence because the class includes numerous categories of purchasers that suffered no injury, and that there is no class-wide methodology for identifying or distinguishing between those persons and otherwise injured class members.
Dr. Hughes described several categories of uninjured consumers, which include: (1) consumers who purchased Provigil after meeting an annual out-of-pocket maximum or deductible; (2) consumers whose insurer placed generic modafinil on the same formulary tier as Provigil, and thus would have the same copay for generic and branded Provigil; and (3) brand loyal consumers who would have bought branded Provigil even if a generic had been available. (See Hughes Exp. Rep., June 10, 2011, ¶¶ 90-110, 105-110, 126-27; Hughes Supp. Exp. Rep., Dec. 20, 2013, ¶¶ 10, 11.) Dr. Hughes further identified categories of institutional payors that would otherwise fall within the class definition but who were not injured, such as: (1) institutional payors that shared risk with pharmacies through capitation agreements; and (2) institutional payors that paid more for generic Provigil due to aggressive promotion of generic substitution through their copayment structure. (Hughes Exp. Rep., June 10, 2011, ¶¶ 70, 75.) Due to these categories of uninjured class members, Dr. Hughes opined that determining whether individuals would have suffered harm, or would instead fall into one of these no-injury categories, would require individualized inquiry into the contracts that govern the relationships between entities and consumers, as well as the consumer's purchasing history. (Id. at ¶¶ 70, 75, 98, 110, 126.)
Based upon Dr. Hughes' experience and training and the reliable data he reviewed, I credit his testimony regarding the numerous categories of uninjured consumers and the extensive individualized inquiries that would follow. I thus conclude that a significant number of uninjured class members remain within the class definition, and that Plaintiffs have not identified a methodology that would identify and remove those persons on a class-wide basis. This conclusion is supported by Plaintiffs' own expert Dr. Hartman, who acknowledged that in order to identify uninjured persons or entities, whether they fall within an exclusion or not, would require a fact-intensive, individualized analysis of the contracts between various entities and the consumer, as well as the purchasing history of a particular consumer. (Hrg. Tr., Mar. 24, 2015, pp. 183-87.) While Dr. Hartman indicated that "[y]ou have to look at individualized records in that case for a few number - for a number of consumers" (id. at 186), there is no escaping that every potential class member would need to be subject to individualized inquiries. Indeed, when every class member has the potential to be a brand loyalist, a person with a flat copay or a consumer who never paid out-of-pocket for their prescriptions, and the only way to identify persons who fall within those groups is individualized inquiry, individualized inquiries would predominate.
In reaching his opinions, Dr. Hughes stated that nearly seventeen percent of consumers and nearly half of the employer groups paid the same or more for generic modafinil than they had paid for branded Provigil prior to generic entry, and therefore were not injured. (Hughes Supp. Exp. Rep., Dec. 20, 2013, ¶¶ 45, 69.) This figure was vigorously challenged by Plaintiffs. After considering this testimony, I find that Dr. Hartman convincingly refuted Dr. Hughes' seventeen percent non-impact number. Dr. Hughes obtained these percentages by comparing the amount paid for Provigil by consumers and TPPs in the year prior to generic entry—2011—to the amount paid by class members for generic Provigil in the year following generic entry—2013. (See id.) Dr. Hartman explained that in order to accurately determine whether there was an overcharge, one should take the but-for price of generic Provigil and compare it to either the price of branded Provigil, or, after generic entry, the actual price of generic Provigil, during the same time period. (Hrg. Tr., Mar. 24, 2015, pp. 136-44.) According to Dr. Hartman, when comparing the prices of these products during the same quarter, the data reflects an overcharge to virtually all class members. (Id.; Hartman Market Def. & Impact Exp. Rep., Apr. 26, 2011, ¶¶ 128-35.) Based upon this analysis, I do not accept that the number of uninjured consumer class members reaches seventeen percent, as Dr. Hughes suggested.
Additionally, despite Dr. Hartman's assurances that only a de minimis number of class members are uninjured, and that by excluding certain categories of uninjured class members, impact is capable of class-wide proof, the exclusions proposed by Plaintiffs do not resolve the predominance issue. In fact, the many exclusions proposed by Plaintiffs appear to be part of the problem.
As described above, Dr. Hughes established that Plaintiffs have not offered a class-wide methodology for identifying those persons who purchased Provigil or its generic equivalent, but who fall within an exclusion, such as brand loyalists and persons with flat copays. When the identification and exclusion of these consumers cannot be managed without considering the highly individualized purchasing history of individuals and their specific insurance plans, simply stating that they are excluded from the class definition is not sufficient to show that common issues will predominate.
Further, I find that unrebutted testimony from Dr. Hughes credibly demonstrates that more than a de minimis number of uninjured persons remain within the class, despite Plaintiffs' assurances to the contrary. Dr. Hughes has identified several categories of consumers and TPPs that have not been excluded from the class definition, but would be uninjured. One such category of institutional payors includes TPPs that are uninjured due to capitation agreements between TPPs and pharmacies. For example, TPPs may have agreements with pharmacies whereby the TPP would reimburse the pharmacy a set fee for a certain therapeutic class of drug, whether the drug was brand or generic. Therefore, pharmacies can insulate TPPs from injury. (Hughes Exp. Rep., June 10, 2011, ¶ 70.)
Another possible category of uninjured parties includes TPPs that pay more for the generic than branded Provigil because they aggressively promote generic substitution through their copayment structure. For example, when a generic is priced slightly, but not substantially, below the price of the branded drug, and the TPP requires a much higher copay for the brand-name drug than the generic, the TPP may actually pay more for the generic because they are receiving a much lesser copay contribution from the consumer. This category of TPPs could also be uninjured. (Id. at ¶ 75.)
Consumers with no out-of-pocket payment are another category of persons who could be uninjured, but have not been excluded from the class definition. Plaintiffs have not disputed Dr. Hughes' finding that, based upon employer claims data, five percent of all consumers prescribed Provigil never paid out-of-pocket for the drug from January 2005 through March 2010. (Id. at ¶ 106, Ex. 6.) According to Dr. Hughes, this result is derived from two groups of people: (1) consumers who have reached their annual out-of-pocket maximum for prescriptions prior to purchasing Provigil; and (2) consumers that are covered by an exclusively employer-funded health reimbursement arrangement or health savings account. If these consumers never paid out-of-pocket for their branded or generic Provigil prescription, they are uninjured by any delay in generic entry. (Id. at ¶¶ 105-10.)
Dr. Hughes also described consumers who received no cost-benefit from switching to the generic. For example, some health plans place non-preferred brands of generics in the highest copayment tier. If a consumer's TPP lists generic Provigil in the same or a higher copayment tier than that of branded Provigil, that consumer would not pay any less for generic Provigil, and thus would not be injured by delayed generic entry. (Id. at ¶¶ 126-27.) In fact, Dr. Hughes' supplemental expert report, which was completed following generic entry, actually demonstrates that some of the plans associated with named Plaintiff Vista Healthplan covered both branded and generic Provigil on tier three. (Hughes Supp. Exp. Rep., Dec. 20, 2013, ¶¶ 57-58.) Dr. Hughes further noted that, of the plans he reviewed, nine percent of three-tier non-Medicare plans, eighteen percent of four-tier Medicare plans, and thirty-five percent of five-tier Medicare plans placed generic Provigil on a non-preferred tier, which can result in no injury to consumers. (Id. at ¶ 59, Ex. 4.)
Dr. Hughes also opined that ten percent of three-tier non-Medicare plans, nineteen percent of Medicare four-tier plans, and forty-seven percent of Medicare five-tier plans placed generic modafinil on the same or higher formulary tier after generic entry than Provigil had occupied prior to generic entry. (Id. at ¶ 60, Ex. 5.) I reject Dr. Hughes' opinion that these consumers are all uninjured for the same reason I rejected Dr. Hughes' opinion that seventeen percent of the class is uninjured. Dr. Hughes appears to have improperly compared data from the year prior to generic entry to data derived from the year after generic entry, as opposed to data from the same quarter.
While Dr. Hughes could not quantify the prevalence of many of these groups of uninjured class members among purchasers of Provigil, he testified that in his experience, these categories of TPPs and consumers exist within the pharmaceutical market place, a point not disputed by Plaintiffs. Dr. Hughes reliably and credibly stated that at least five percent of consumers had no out-of-pocket payment, and thus were uninjured, from 2005 to 2010. I find that this five percent, combined with the substantial likelihood that some of the other categories mentioned above of uninjured class members identified by Dr. Hughes would be within the proposed class, indicates that the prevalence of uninjured class members is more than de minimis.
This case is similar to Sheet Metal Workers Local 441 Health & Welfare Plan v. GlaxoSmithKline, PLC, 2010 WL 3855552 (E.D. Pa. Sept. 30, 2010), where Judge Lawrence Stengel of this district denied class certification for a group of end payors in a case involving delayed generic entry. There, Judge Stengel found that the class contained several categories of uninjured class members, such as brand loyalists and those whose insurance plan terms insulated them from overcharges. Id. at *26. Judge Stengel commented that "I cannot fathom, and the plaintiffs have not put forth, a method for identifying which individual purchasers would [be uninjured] through analysis of common information." Id. at *25. The plaintiffs' evidence did "not show that all class members paid supra-competitive prices for generic or branded [Wellbutrin SR], or that this determination c[ould] be made with common proof." Id. at *27. Accordingly, the plaintiffs were unable to meet the predominance standard in light of Hydrogen Peroxide's rigorous analysis requirement. I reach the same conclusion here. Without a means of identifying these uninjured persons using common evidence, every class member would need to be reviewed on an individualized basis to see if they were impacted by Defendants' alleged anticompetitive actions.
In summary, Plaintiffs have not sufficiently proven that they are able to demonstrate antitrust impact on a class-wide basis. This is due to various groups of uninjured persons that remain within the class, and because identifying and removing these uninjured class members would require extensive individualized inquiry.
Plaintiffs argue that they have demonstrated predominance with respect to antitrust damages through Dr. Hartman's yardstick methodology. (Hartman Damages Exp. Rep., Apr. 26, 2011, ¶ 42.) For overcharge damages, Dr. Hartman subtracts the but-for price of generic Provigil from the prices of branded and generic Provigil in the actual world during the relevant time period. He then multiplies that price difference by the number of prescriptions written. (Id. at ¶¶ 43-44.) The aggregate amount of overcharges to the Class, as calculated by Dr. Hartman, is $2.449 billion. (Hartman Supp. Exp. Rep., Dec. 20, 2013, ¶ 4.) In calculating unjust enrichment damages, Dr. Hartman subtracts the profits that Defendants would have realized in the but-for world from the amount of profits Defendants realized in the actual world during the relevant time period. (Hartman Damages Exp. Rep., Apr. 26, 2011, ¶¶ 47-48; Hartman Supp. Exp. Rep., Dec. 20, 2013, ¶ 12.) Dr. Hartman opined that the amount of unjust enrichment owed to the class, in the aggregate, is $2.507 billion. (Hartman Supp. Exp. Rep., Dec. 20, 2013, ¶ 11.)
Defendants argue that Plaintiffs are unable to demonstrate predominance as to damages for several reasons. First, citing Comcast Corp. v. Behrend, 133 S. Ct. 1426 (2013), Defendants argue that "[q]uestions of individual damages calculations will inevitably overwhelm questions common to the class." (Defs.' Resp., p. 24.) Second, in supplemental briefing, Defendants assert that Plaintiffs' damages calculation did not match their theory of antitrust impact. For the reasons that follow on this point, I disagree with Defendants' arguments.
"At the class certification stage, the plaintiffs are not required to prove damages by calculating specific damages figures for each member of the class, but rather they must show that a reliable method is available to prove damages on a class-wide basis." In re Wellbutrin XL, 282 F.R.D. at 144 (citing In re Neurontin Antitrust Litig., 2011 WL 286118, at *9 (D.N.J. Jan. 25, 2011)). Courts have held that proof of aggregate damages is appropriate in class actions. In re Pharm. Indus. Average Wholesale Price Litig., 582 F.3d 156, 197 (1st Cir. 2009) ("The use of aggregate damages calculations is well established in federal court and implied by the very existence of the class action mechanism itself").
"Some variation of damages among class members does not defeat certification," In re Flonase, 284 F.R.D. at 232 (quoting Behrend v. Comcast Corp., 655 F.3d 182, 204 (3d Cir. 2011)), and damages "calculations need not be exact." Comcast, 133 S. Ct. at 1433 (citing Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 563 (1931)). Once injury has been established, "the jury is permitted to calculate the actual damages suffered using a reasonable estimation, as long as the jury verdict is not the product of speculation or guess work." Rossi v. Standard Roofing, Inc., 156 F.3d 452, 484 (3d Cir. 1998) (quoting In re Lower Lake Erie Iron Ore Antitrust Litig., 998 F.2d 1144, 1176 (3d Cir. 1993)) (quotation marks omitted).
In Comcast, the Supreme Court addressed the issue of predominance as it relates to antitrust damages. The district court had certified the class, but had only accepted one out of four of the plaintiffs' theories of antitrust impact as capable of class-wide proof—"the theory that Comcast engaged in anticompetitive clustering conduct, the effect of which was to deter the entry of overbuilders" in the Philadelphia area. Comcast, 133 S. Ct. at 1431. However, the damages model the plaintiffs' expert had used to calculate damages for the class included damages from all of the various theories of antitrust impact, including the ones not certified for class treatment. Id. The Supreme Court reversed class certification, finding that plaintiffs had not shown that damages were capable of measurement on a class-wide basis, as required to establish predominance. Id. at 1433. The primary takeaway from Comcast has been that a "plaintiff's damages case must be consistent with its liability case, particularly with respect to the alleged anticompetitive effect of the violation." Id.
In the wake of Comcast, some defendants, including those here, argue that plaintiffs cannot have variations in damages calculations, and that diverse individual damages calculations prohibit class treatment. This is because in Comcast, the district court held (without a challenge on appeal) that in order to meet the predominance requirement, plaintiffs had to show "that the damages resulting from that injury were measurable 'on a class-wide basis' through use of a 'common methodology.'" Id. at 1430. Furthermore, the Supreme Court stated that "[q]uestions of individual damage calculations will inevitably overwhelm questions common to the class." Id.
Circuit courts have largely rejected the interpretation urged by Defendants—that variations in damages calculations between and among class members defeat predominance. See Butler v. Sears, Roebuck & Co., 727 F.3d 796, 801 (7th Cir. 2013) ("It would drive a stake through the heart of the class action device, in cases in which damages were sought . . . to require that every member of the class have identical damages"); see also In re Nexium Antitrust Litig., 777 F.3d 9, 18-19 (1st Cir. 2015) (limiting its interpretation of Comcast to the principle that the plaintiff's theory of impact must match his damages model); In re Deepwater Horizon, 739 F.3d 790, 817 (5th Cir. 2014) (same); In re Whirlpool Corp. Front Loading Washer Prods. Liab. Litig., 722 F.3d 838, 860 (6th Cir. 2013) (same); Leyva v. Medline Indus. Inc., 716 F.3d 510, 514 (9th Cir. 2013) (same). Indeed, "[i]f the issues of liability are genuinely common issues, and the damages of individual class members can be readily determined in individual hearings, in settlement negotiations, or by creation of subclasses, the fact that damages are not identical across all class members should not preclude class certification." Butler, 727 F.3d at 801. Accordingly, Comcast has largely been limited to its unique set of facts, and interpreted as precluding class treatment where the class-wide measure of damages does not match the theory of antitrust impact.
In response to Defendants' argument that variations in damages calculations overwhelm questions common to the class, Dr. Hartman demonstrated that his aggregate damages model is able to account for the variations between and among class members. To establish this point, Dr. Hartman compiled a list of individuals with copays for branded and generic Provigil ranging from $0 to $180. Some members of the sample had a flat copay, and some experienced significant savings from purchasing generic Provigil. Dr. Hartman added up the total injury to this small sample individual-by-individual—that is, he computed the overcharge experienced by each individual, and added those numbers together to find a total overcharge for the sample. He then used his averages formula to calculate the average overcharge to the sample, and multiplied that figure by the number of persons in the sample. Whether Dr. Hartman added up the overcharges individual-by-individual or took the average overcharge and multiplied it by the total number of class members, Dr. Hartman reached the same exact amount of total damages to the sample. Through this demonstration, Plaintiffs have satisfied me that their damages were derived using a reliable method that took individual variations among class members into consideration. (Hrg. Tr., Mar. 24, 2015, pp. 95-103.) Therefore, I do not find that individual variations in Plaintiffs' damages calculations prevent a finding of predominance.
In fact, Defendants do not seem to dispute that the use of yardsticks and averages to compile aggregate damages numbers is a reliable method. During the hearing, defense counsel sought to stipulate that "one and one equals two" and stated that he was "not going to be attacking the individual calculation to get to the total number." (Hrg. Tr., Mar. 24, 2015, p. 99.)
Defendants separately urge that when this Court granted Defendants' motions for summary judgment on Plaintiffs' overall conspiracy claims, Dr. Hartman's damages model suffered the same defect that the Court addressed in Comcast, in that his damages model no longer fit Plaintiffs' theory of liability and antitrust impact.
In a prior Opinion, I considered whether Plaintiffs had provided sufficient evidence to survive summary judgment on a claim for one overall antitrust conspiracy among all Defendants. King Drug Co. of Florence, Inc. v. Cephalon, Inc., 2014 WL 2813312 (E.D. Pa. June 23, 2014). In their complaint, Plaintiffs had alleged that Cephalon and the Generic Defendants had all conspired together to keep generic Provigil off of the market and to share in the generated monopoly profits. Id. at *4. In support of their claim for overall conspiracy, Plaintiffs pointed to the 180-day period of exclusivity shared by the Generic Defendants, as well as the substantially identical contingent launch provisions found within each of the settlement agreements. Id. Defendants had argued that each Generic had an individual incentive to demand a contingent launch provision, and that Plaintiffs' circumstantial evidence was insufficient to establish an overarching conspiracy as a matter of law. Id. I agreed with Defendants, holding that the Private Plaintiffs had not provided direct evidence of an overall agreement encompassing Cephalon and all of the Generic Defendants, nor had they presented circumstantial evidence that supported "an inference of concerted, as opposed to independent, action." Id. at *14. Accordingly, summary judgment was granted in favor of the Defendants as to the Private Plaintiffs' claims of overall conspiracy. Id.
As a result of this decision, which was issued after briefing had been completed on the instant motion for class certification, Defendants submitted a supplemental argument under Comcast urging that Plaintiffs' damages model did not match their remaining theories of liability. Defendants argue that Dr. Hartman's damages model presents aggregate damages to the entire market, which he attributes collectively to all Defendants. To match Plaintiffs' remaining theories of liability, Defendants posit that Dr. Hartman needed to analyze the potential damages attributable to each of the separate agreements between Cephalon and a Generic Defendant standing on its own. Such an analysis would allow a jury to determine, if only one or two Defendants were ultimately found liable, the amount of damages attributable to those specific parties' conduct. (Doc. No. 425, pp. 4-5.)
Plaintiffs respond that their damages calculation does not run afoul of Comcast because their "damage analysis is exactly the same whether there is one conspiracy or four." (Doc. No. 429, pp. 1-2.) This is due to the contingent launch provisions in each settlement agreement, urge Plaintiffs, which allowed the Generic Defendant bound by that settlement agreement to enter the market earlier than 2012 if any other generic Provigil product entered the market. Therefore, "[i]f any one of the Generic Defendants had not accepted a payment from Cephalon to stay off the market then that company would have launched a generic, all other Generic Defendants would have entered and prices would fall for the entire class." (Id. at p. 2.)
I agree with Plaintiffs and find that Dr. Hartman's damages model comports with the remaining theories of liability. Dr. Hartman opines that but for the settlement agreements, the following would have occurred: (1) the four Generic Defendants would have launched on June 24, 2006; (2) Apotex would have launched a generic Provigil product on December 24, 2006; and (3) Cephalon would have launched its own authorized Generic on June 24, 2006, which would lead to six generic Provigil products on the market. (Hartman Damages Exp. Rep., Apr. 26, 2011, ¶ 26.)
According to Dr. Hartman:
But-for the settlement agreements between Cephalon and each Defendant, generic launch would have occurred as posed above. . . . If Cephalon had settled with only a subset of the Generic Defendants, those settling [G]eneric Defendants would likely have launched by triggering the clause that allowed each settling Generic Defendant to launch if another generic launched.(Id. at ¶ 27 (emphasis in original).) Dr. Hartman then measures the difference between what the price would have been with six generic entrants in 2006 and the price actually paid by consumers and TPPs in the actual world. (Id. at ¶¶ 42-46.) A hypothetical provided by Plaintiffs helps to illustrate this issue:
If End-Payors prove at trial that Ranbaxy accepted a payment from Cephalon for the purpose of staying off the market in violation of the various state antitrust[,] consumer protection and unjust enrichment laws, End-Payors' damage model assumes that, but for the illegal agreement, Ranbaxy would have entered the market in 2006. The damage model also assumes that other generic competitors would have followed Ranbaxy's entry shortly thereafter.(Doc. No. 429, p. 3.) Therefore, I do not agree with Defendants' Comcast argument, and I find that Plaintiffs' damages model matches their remaining theories of liability and impact. Accordingly, although I find that Plaintiffs have not been able to establish predominance as to antitrust impact, I disagree with Defendants' arguments regarding Plaintiffs' damages calculations.
Defendants do not squarely attack Dr. Hartman's calculations for unjust enrichment damages; however, similar logic would apply. According to Dr. Hartman's analysis, Defendants' profits would have dropped by the same amount in the but-for world, whether one or all of the settlement agreements were anticompetitive, due to the contingent launch provisions. (See id. at ¶¶ 47-48.)
Because I find that numerous individualized inquiries prevent Plaintiffs from establishing predominance for the element of antitrust impact, I also find that superiority has not been established as to the state antitrust claims. A class action would not be superior to other available methods for trying these claims because the prevalence of individualized inquiries would make the case unmanageable on a collective basis. See Fed. R. Civ. P. 23(b)(3). Further, conflicts between the various states' consumer protection laws, which will be discussed infra, defeat superiority for the antitrust/consumer protection law class.
VI. UNJUST ENRICHMENT CLAIMS - PREDOMINANCE & SUPERIORITY
A. Choice of Law
I must first determine which laws apply to the unjust enrichment claims of the proposed class. "A necessary precondition to deciding Rule 23 issues is a determination of the state whose law will apply." Powers v. Lycoming Engines, 328 Fed. Appx. 121, 124 (3d Cir. 2009). A court "must apply an individualized choice of law analysis to each plaintiff's claims" raised by a proposed class action. Georgine, 83 F.3d at 627 (citing Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 823 (1985)).
Despite being a necessary prerequisite to the Rule 23 inquiry, Plaintiffs failed to brief the relevant choice-of-law analysis with respect to which laws should govern the state antitrust, consumer protection and unjust enrichment claims of the proposed classes. As such, Plaintiffs failed to meet their burden of showing that common questions of law predominate. See Spence v. Glock, Ges.m.b.H., 227 F.3d 308, 313 (5th Cir. 2000) ("The burden of proof lies with the plaintiffs; in not presenting a sufficient choice of law analysis they have failed to meet their burden of showing that common questions of law predominate").
I will nonetheless undertake a choice of law analysis because it is a prerequisite to an evaluation of the Rule 23(b)(3) factors. See In re LifeUSA Holding Inc., 242 F.3d 136, 147 (3d Cir. 2001) (finding error where the "District Court failed to consider how individualized choice of law analysis of the forty-eight different jurisdictions would impact Rule 23's predominance requirement").
When a federal court is sitting in diversity, the court must apply the choice of law rules of the forum state to determine what substantive state law will govern. Klaxon Co. v. Stentor Elec. Mfg. Co., 313 U.S. 487, 496 (1941). This action was commenced in the United States District Court for the Eastern District of Pennsylvania. As such, I will apply Pennsylvania's choice of law rules.
Under Pennsylvania's choice of law rules, the first step is to determine whether there is an actual or true conflict between the potentially applicable laws. Hammersmith v. TIG Ins. Co., 480 F.3d 220, 229-30 (3d Cir. 2007). If there are no relevant differences or the laws would produce the same result, the court need not engage in a choice of law analysis and may refer to the laws "interchangeably." Id. at 229.
However, if there are relevant differences, then the court must examine the governmental policies which underlie the laws. Id. at 230. Based on the result of that analysis, the court then characterizes the case as a "true conflict, false conflict, or unprovided-for case." Id. (citations omitted). If the relevant policy interests of both jurisdictions would be impaired by application of the other jurisdiction's law, there is a true conflict. Id. Where there is a true conflict, the court must then determine which state has the "greater interest in the application of its law." Id. at 230-31.
Pennsylvania requires that courts making such a determination use a "combination of the approaches of both [the] Restatement II (contacts establishing significant relationships) and interests analysis (qualitative appraisal of the relevant States' policies with respect to the controversy)." Id. at 231 (citing Melville v. Am. Home Assur. Co., 584 F.2d 1306, 1311 (3d Cir. 1978) (quotation marks omitted)). The interest analysis requires a weighing of "the contacts on a qualitative scale according to their relation to the policies and interests underlying the [particular] issue." Id. at 231 (citing Shields v. Consol. Rail Corp., 810 F.2d 397, 400 (3d Cir. 1987) (alterations in original)).
A false conflict exists when "only one jurisdiction's governmental interests would be impaired by the application of the other jurisdiction's law." Lacey v. Cessna Aircraft Co., 932 F.2d 170, 187 (3d Cir. 1991). If there is a false conflict, the court applies the law of the only interested jurisdiction. Id. Finally, a case is unprovided-for where neither jurisdiction's interests would be impaired if its laws are not applied. Garcia v. Plaza Oldsmobile Ltd., 421 F.3d 216, 220 (3d Cir. 2005). In unprovided-for cases, "the principle of lex loci delicti, the law of the place of the wrong, supplies the substantive law to be applied." Id.
Applying the above choice of law framework, I must first determine whether there is a true conflict between the twenty-six unjust enrichment laws under which Plaintiffs seek certification as well as any other potentially applicable unjust enrichment laws. Several courts in this circuit that have been confronted with the issue have determined that no material differences distinguish the various states' unjust enrichment laws and, therefore, no conflict exists. See Pennsylvania Employee, Benefit Trust Fund v. Zeneca, Inc., 710 F. Supp. 2d 458, 477 (D. Del. 2010) (concluding that the "basic elements" required under various unjust enrichment laws do not create an actual conflict); Agostino v. Quest Diagnostics Inc., 256 F.R.D. 437, 464 (D.N.J. 2009) ("the Court concludes that there are no actual conflicts among the laws of unjust enrichment"); Powers v. Lycoming Engines, 245 F.R.D. 226, 231 (E.D. Pa. 2007), rev'd on other grounds, 328 Fed. Appx. 121 (3d Cir. 2009) ("there are few real differences amongst the unjust enrichment cause of action in the various states and no conflict exists").
However, other courts in this circuit have reached the opposite conclusion and determined that the elements necessary to establish an unjust enrichment cause of action in various jurisdictions differ in material ways and, therefore, give rise to a conflict. See In re Actiq Sales & Mktg. Practices Litig., 2015 WL 1312015, at *11-12 (E.D. Pa. Mar. 23, 2015) (concluding that a true conflict exists as the variances in the states' unjust enrichment law could lead to differential treatment of the claims of the proposed nationwide class).
The conclusion that a conflict exists finds support in courts outside of this circuit. See, e.g., Mazza v. Am. Honda Motor Co., 666 F.3d 581, 591 (9th Cir. 2012) ("[t]he elements necessary to establish a claim for unjust enrichment also vary materially from state to state"); Thompson v. Bayer Corp., 2009 WL 362982, at *4 (E.D. Ark. Feb. 12, 2009) ("[a]fter an extensive review of the law, the Court finds that the states' different approaches to, or elements of, unjust enrichment are significant"); In re Aqua Dots Products Liab. Litig., 270 F.R.D. 377, 386 (N.D. Ill. 2010) (the law of unjust enrichment "varies too much" from state to state and poses "insurmountable choice-of-law problems"); Clay v. Am. Tobacco Co., 188 F.R.D. 483, 501 (S.D. Ill. 1999) ("variances exist in state common laws of unjust enrichment"); Thompson v. Jiffy Lube Intern., Inc., 250 F.R.D. 607, 627 (D. Kan. 2008) (finding a conflict between state unjust enrichment laws).
In considering a proposed nationwide unjust enrichment class, the In re Actiq opinion catalogues some of the many material ways in which unjust enrichment laws vary. First, states apply statutes of limitations of varying lengths to unjust enrichment claims. 2015 WL 1312015, at *11. Second, states have different rules as to when and how the statute of limitations accrues. Id. Third, some states do not recognize unjust enrichment as an independent cause of action. Id. at *12. Fourth, some but not all states require a plaintiff to demonstrate that they lack an adequate remedy at law. Id. Fifth, some states require that a plaintiff establish that the benefit was directly conferred on the defendant. Id. Sixth, the states also vary as to the level of misconduct, if any, a plaintiff must prove. Id. Lastly, the states follow different rules as to the availability of defenses, including laches and unclean hands. Id.
According to Plaintiffs, the "basic elements" common to the relevant unjust enrichment laws are as follows: "(1) Plaintiff confers benefit on defendant; (2) defendant accepts/retains benefit; (3) circumstances make it unjust for defendant to do so." (Pls.' Mot., Meltzer Decl., Ex. 25.) However, although Plaintiffs do not seek certification of a nationwide class, several of the differences catalogued in In re Actiq distinguish the unjust enrichment laws relevant to Plaintiffs' class proposal. In fact, the state law chart submitted by Plaintiffs in support of their motion for class certification demonstrates the presence of three such material jurisdictional differences. (See id.)
First, according to Plaintiffs' chart, Arizona, Louisiana, Massachusetts, New York, North Carolina, North Dakota, Tennessee and Utah require that a plaintiff demonstrate that there is not an adequate remedy at law in addition to the "basic elements" which compromise an unjust enrichment claim. (Id.)
My review of the case law confirms that these states do indeed recognize this requirement. Trustmark Ins. Co. v. Bank One, Ariz., NA, 48 P.3d 485, 491 (Ariz. Ct. App. 2002) (to establish unjust enrichment a party must show "the absence of a legal remedy"); La. Civ. Code Ann. art. 2298 (unjust enrichment "shall not be available if the law provides another remedy for the impoverishment or declares a contrary rule"); Santagate v. Tower, 833 N.E.2d 171, 176 (Mass. App. Ct. 2005) ("equitable remedy for unjust enrichment is not available to a party with an adequate remedy at law"); Samiento v. World Yacht Inc., 883 N.E.2d 990, 996 (N.Y. 2008) (cause of action for unjust enrichment "does not lie as plaintiffs have an adequate remedy"); Jones Cooling & Heating, Inc. v. Booth, 394 S.E.2d 292, 294 (N.C. App. 1990) (plaintiff may not recover under a theory of unjust enrichment where an adequate remedy at law exists); Lochthowe v. C.F. Peterson Estate, 692 N.W.2d 120, 124 (N.D. 2005) (to establish unjust enrichment a party must demonstrate "an absence of a remedy provided by law"); Thorpe v. Washington City, 243 P.3d 500, 507 (Utah App. 2010) (plaintiff must show absence of an adequate remedy at law); Freeman Industries, LLC v. Eastman Chem. Co., 172 S.W.3d 512, 525 (Tenn. 2005) (a plaintiff must demonstrate the absence of a legal remedy).
Additionally, although not mentioned in Plaintiffs' chart, a review of the case law also discloses that Hawaii and Minnesota recognize the same no adequate remedy requirement. Porter v. Hu, 169 P.3d 994, 1007-08 (Haw. Ct. App. 2007) (unjust enrichment is only appropriate in the absence of an adequate remedy at law); Caldas v. Affordable Granite & Stone, Inc., 820 N.W.2d 826, 842 (Minn. 2012) (citing Service Master of St. Cloud v. GAB Bus. Servs., Inc., 544 N.W.2d 302, 305 (Minn. 1996) ("A party may not have equitable relief where there is an adequate remedy at law available")). Pennsylvania does as well. Meehan v. Cheltenham Twp., 189 A.2d 593, 595 (1963) (holding that unjust enrichment is not available where an adequate remedy at law exists).
Second, according to Plaintiffs' chart, in addition to the "basic elements," California, Florida, Kansas, Maine, Massachusetts, Nevada, New Mexico, North Carolina, South Dakota, Tennessee, Utah and Wisconsin require that a plaintiff establish that the defendant appreciates or has knowledge of the benefit conferred. (Pls.' Mot., Meltzer Decl., Ex. 25.)
My review of the case law confirms Plaintiffs' assertion regarding the foregoing jurisdictions. See Ghirardo v. Antonioli, 924 P.2d 996, 1003 (Cal. 1996) (plaintiff must demonstrate that the defendant knew of the benefit); Hillman Const. Corp. v. Wainer, 636 So. 2d 576, 577 (Fla. 4th Dist. App. 1994) (must establish that the "plaintiff has conferred a benefit on the defendant, who has knowledge thereof"); J.W. Thompson Co. v. Welles Products Corp., 758 P.2d 738, 745 (Kan. 1988) (plaintiff must establish "an appreciation or knowledge of the benefit by the defendant"); In re Est. of Anderson, 988 A.2d 977, 980 (Me. 2010) (plaintiff must establish that receiving party "had appreciation or knowledge of the benefit"); Stevens v. Thacker, 550 F. Supp. 2d 161, 165 (D. Mass. 2008) (plaintiff must establish "appreciation or knowledge of the benefit by the defendant"); Certified Fire Prot. Inc. v. Precision Constr., 283 P.3d 250, 257 (Nev. 2012) (plaintiff must establish that the defendant "appreciates" the benefit); Ontiveros Insulation Co., Inc. v. Sanchez, 3 P.3d 695, 698 (N.M. App. 2000) (plaintiff must establish that "another has been knowingly benefitted at one's expense"); D.W.H. Painting Co., Inc. v. D.W. Ward Const. Co., Inc., 620 S.E.2d 887, 893 (N.C. App. 2005) (the defendant must have "consciously" accept the benefit); Action Mech., Inc. v. Deadwood Historic Preservation Commn., 652 N.W.2d 742, 750 (S.D. 2002) (defendant must be "aware" that he is receiving a benefit); Freeman Industries, LLC, 172 S.W.3d at 525 (plaintiff must establish "appreciation by the defendant of such benefit"); Rawlings v. Rawlings, 240 P.3d 754, 763 (Utah 2010) (plaintiff must establish "an appreciation or knowledge by the conferee of the benefit"); Puttkammer v. Minth, 266 N.W.2d 361, 363 (Wis. 1978) (plaintiff must establish "an appreciation or knowledge by the defendant of the benefit"). Pennsylvania does as well. Mitchell v. Moore, 729 A.2d 1200, 1203 (Pa. Super. 1999) (appreciation of the benefit by defendant is a necessary element of unjust enrichment).
Third, Plaintiffs' chart notes that Minnesota requires a showing of defendant's wrongful conduct in addition to the "basic elements." See Service Master, 544 N.W.2d at 306 ("it must be shown that a party was unjustly enriched in the sense that the term 'unjustly' could mean illegally or unlawfully.")
Plaintiffs' chart stops short of offering a full analysis of all the ways that the various unjust enrichment laws vary. For example, Plaintiffs fail to note that New York requires a plaintiff to demonstrate a "relationship or connection between the parties that is not too attenuated," Georgia Malone & Co., Inc. v. Rieder, 973 N.E.2d 743, 746 (N.Y. 2012), or that North Dakota, Arizona and Louisiana have a similar yet distinct requirement that a plaintiff must demonstrate a "connection between the enrichment and the impoverishment." Zuger v. N. Dakota Ins. Guar. Ass'n, 494 N.W.2d 135, 138 (N.D. 1992); City of Sierra Vista v. Cochise Enter., Inc., 697 P.2d 1125, 1131 (Ariz. Ct. App. 1984); USA Disaster Recovery, Inc. v. St. Tammany Parish Govt., 145 So. 3d 235, 236 n.1 (La. 2013). Plaintiffs' failure to account for these additional variations underscores the divergent nature of the body of relevant state law.
The foregoing variances are significant as some, if not all, could result in differential treatment of Plaintiffs' claims. See In re Actiq, 2015 WL 1312015, at *12. The resulting differential treatment is not accidental—the unique tailoring of the unjust enrichment laws reflects the policy choices of the state as to when the equitable remedy should be made available. As such, imposition of another state's more permissive law could impair the interests of a state with a more stringent law. In light of these material differences which implicate the states' interests in providing a forum for redress, I conclude that a true conflict exists amongst the relevant unjust enrichment laws.
Having found that a true conflict exists, I must determine which state or states have a greater interest in application of its unjust enrichment law. Pursuant to Pennsylvania's choice of law rules, the first step is to consider the relevant Restatement factors.
The Restatement instructs that the following contacts are to be considered when assessing which jurisdiction has the most significant relationship to the occurrence giving rise to an unjust enrichment cause of action:
(a) the place where a relationship between the parties was centered, provided that the receipt of enrichment was substantially related to the relationship,Restatement (Second) of Conflict of Laws § 221(2) (1971); Powers, 328 Fed. Appx. at 126 (applying § 221(2) to an unjust enrichment claim under Pennsylvania's choice of law rules).
(b) the place where the benefit or enrichment was received,
(c) the place where the act conferring the benefit or enrichment was done,
(d) the domicil, residence, nationality, place of incorporation and place of business of the parties, and
(e) the place where a physical thing, such as land or a chattel, which was substantially related to the enrichment, was situated at the time of the enrichment.
Regarding the first factor, the relationship between the parties was centered in the state in which Plaintiffs purchased, paid and/or reimbursed for Provigil or its generic equivalent. The class members allegedly conferred the benefit on Defendants in the state where the purchase was made. Since unjust conferral of a benefit is the gravamen of the unjust enrichment cause of action, the first factor weighs strongly in favor of applying the unjust enrichment law of the purchase state to each claim raised. Additionally, this factor is generally "given the greatest weight in determining the state of the applicable law." Restatement (Second) of Conflict of Laws § 221 cmt. d.
Regarding, the second factor, where the benefit was received, Cephalon and Teva's principal places of business are located in Pennsylvania, Barr's principal place of business is located in New York, and Ranbaxy's principal place of business is located in Florida. (See Answers, Doc. Nos. 118, 129, 131 and 134). Therefore, Defendants likely received the alleged overpayments in Pennsylvania, New York and Florida. As such, the second factor weighs in favor of applying Pennsylvania law to the unjust enrichment claims against Cephalon and Teva, New York law to the unjust enrichment claims against Barr and Florida law to the unjust enrichment claims against Ranbaxy.
However, the third factor militates in favor of application of the law of the purchase state because the "act bestowing the enrichment" is payment and/or reimbursement of Provigil or its general equivalent. As noted above, this act occurred in the state in which Plaintiffs made the relevant purchases.
The fourth factor also weighs in favor of application of the law of purchase states. Although Defendants' principal places of business are Pennsylvania, Florida and New York, purchases were allegedly made in all jurisdictions in which Plaintiffs seek certification. As such, no single state has a greater connection to the case than any other state.
Lastly, the fifth factor weighs in favor of applying the law of the purchase states as well because that is where the Provigil or its generic equivalent—the "physical thing" substantially related to the enrichment—was located at the time of the alleged unjust enrichment.
In sum, four out of five of the relevant Restatement factors, including the first factor which is often the most significant, see § 221 cmt. d, demonstrate that the state in which the particular purchase was made has the most significant connection to the related claim. As such, I find that the Restatement factors weigh in favor of applying the laws of the purchase states.
Pursuant to Pennsylvania's choice of law rules, I must also consider the relevant state policies at issue to determine which state has the greatest interest in application of its law. This "qualitative appraisal" also suggests that application of the laws of the purchase states is appropriate. Regarding the relevant state policies at issue, Pennsylvania, Florida and New York have a clear interest in regulating the conduct of corporations transacting business within their borders. However, the states in which the purchases were made "also have an interest in ensuring that corporations conducting business within their borders are doing so fairly." In re Actiq, 2015 WL 1312015, at *13.
In addition to these interests in regulating business, the purchase states have a strong interest in providing a forum for redress to their citizens. Furthermore, "those states with more protective unjust-enrichment laws have an interest in ensuring that their citizens have full recourse to those laws." Rapp v. Green Tree Servicing, LLC, 302 F.R.D. 505, 518 (D. Minn. 2014).
I agree with other courts which have concluded that a state's interest in providing its citizens with the level and type of redress set forth in its unique unjust enrichment law "outweigh a state's interest in regulating a resident corporation." In re Actiq, 2015 WL 1312015, at *13 (citing Rapp, 302 F.R.D. at 518) (finding "no reason to believe that the unjust-enrichment laws of [the non-forum state] could not equally or more effectively hold corporations accountable")). As such, I find that the relevant policy concerns at issue weigh in favor of applying the unjust enrichment laws of the purchase states.
In light of the preceding choice of law analysis, I conclude that the laws of the purchase states govern the proposed class' unjust enrichment claims. Having identified the law applicable to the claims of the proposed class, I now turn to whether Plaintiffs have satisfied the predominance and superiority requirements of Rule 23(b)(3).
Despite Plaintiffs' attempt to characterize the variances between the various unjust enrichment laws of the purchase states as "minor," the laws are distinguished by the substantive and nuanced differences discussed above. Plaintiffs respond that these differences can be "easily addressed through a limited number of jury instructions."
Plaintiffs urge that this conclusion finds support in In re Flonase Antitrust Litig., 284 F.R.D. 207 (E.D. Pa. 2012), In re Wellbutrin XL Antitrust Litig., 282 F.R.D. 126 (E.D. Pa. 2011) and Sullivan v. DB Investments, Inc., 667 F.3d 273 (3d Cir. 2011). Plaintiffs contend that these cases demonstrate a willingness of courts in this circuit to certify "indirect purchaser classes under multiple state laws."
The proposed classes at issue in In re Flonase, In re Wellbutrin XL and Sullivan are distinguishable from the proposed classes before me. In re Flonase involved a proposed class of indirect purchasers pursuing claims under seven different state laws (antitrust, consumer protection, and unjust enrichment) of four states. 284 F.R.D. at 210-11. In granting certification, the court held that the predominance requirement had been met because the evidence would be the same with regard to each of the plaintiffs' state law claims. Id. at 219-20. However, in reaching this conclusion, the court noted and relied upon the fact that the defendants had conceded that the plaintiffs' claims were subject to common proof. Id. That is not the case here.
Likewise, In re Wellbutrin XL involved a proposed class of indirect purchasers pursuing claims under seven different antitrust and consumer protection laws of six states. 282 F.R.D. at 131. The court found that the predominance requirement had been met because "[t]he issues of relevant market, monopoly power, and exclusionary conduct can be proven using common, class-wide evidence because such issues focus on the defendants' conduct rather than individual class members." Id. at 140.
Sullivan, on the other hand, involved proposed certification of a nationwide settlement class. The Third Circuit stated:
Because we are presented with a settlement class certification, we are not as concerned with formulating some prediction as to how [variances in state law] would play out at trial, for the proposal is that there be no trial. As such, we simply need not inquire whether the varying state treatments of indirect purchaser damage claims at issue would present the type of "insuperable obstacles" or "intractable management problems" pertinent to certification of a litigation class. . . . The proposed settlement here obviates the difficulties inherent in proving the elements of varied claims at trial or in instructing a jury on varied state laws, and the difference is key.Sullivan, 667 F.3d at 303-04 (internal citations omitted).
The cases relied on by Plaintiffs are distinguishable from the instant case on multiple grounds. Both In re Flonase and In re Wellbutrin involved significantly fewer state laws than are at issue in Plaintiffs' proposed classes. Additionally, neither certification decision involved a substantive analysis of the variances in the elements of the relevant state laws. And, unlike In re Flonase, Defendants here dispute Plaintiffs' ability to demonstrate that common issues of fact or law predominate.
Sullivan also involved a settlement class and is of limited relevance to certification of a litigation class. See In re Warfarin Sodium Antitrust Litig., 391 F.3d 516, 529 (3d Cir. 2004) (variations in state law are "irrelevant to certification of a settlement class"); Amchem, 521 U.S. at 620 (in a settlement-only class certification, "a district court need not inquire whether the case, if tried, would present intractable management problems . . . for the proposal is that there be no trial").
Moreover, nothing in In re Flonase, In re Wellbutrin or Sullivan relieves Plaintiffs of their burden of demonstrating that common questions of law or fact predominate. This burden includes providing the Court with an extensive analysis which demonstrates that the variations in the applicable state laws do not defeat predominance. See In re Sch. Asbestos Litig., 789 F.2d 996, 1010 (3d Cir. 1986).
I recognize that the existence of variations in state law does not automatically foreclose Plaintiffs ability to satisfy the predominance requirement. The Third Circuit has "endorsed the general procedure of grouping multiple state laws into a few categories for the purposes of class litigation." Grandalski v. Quest Diagnostics Inc., 767 F.3d 175, 183 (3d Cir. 2014). Grouping is permissible where differences in state law fall "into a limited number of predictable patterns, and any deviations could be overcome at trial by grouping similar state laws together and applying them as a unit." Id. However, when taking such an approach, "plaintiffs face a significant burden to demonstrate that grouping is a workable solution." Id.
After careful consideration, I conclude that Plaintiffs' unjust enrichment claims under the laws of the purchase states are not amenable to concise explanation. See In re Actiq, 2015 WL 1312015, at *14 ("[t]he elements of Plaintiffs' unjust enrichment claim cannot be succinctly identified because . . . the law of each [plaintiffs'] home state will govern"). Some states require that a plaintiff satisfy five elements to prevail on a claim of unjust enrichment. See, e.g., Freeman v. Sorchych, 245 P.3d 927, 936 (Ariz. App. 1st Div. 2011) (a plaintiff must demonstrate "(1) an enrichment, (2) an impoverishment, (3) a connection between the enrichment and impoverishment, (4) the absence of justification for the enrichment and impoverishment, and (5) the absence of a remedy provided by law"). Other states require a plaintiff to satisfy three or four elements. See, e.g., Stevens, 550 F. Supp. 2d at 165 ("a plaintiff must prove (1) a benefit conferred upon the defendant by the plaintiff; (2) an appreciation or knowledge of the benefit by the defendant; and (3) the acceptance or retention of the benefit by the defendant under circumstances which make such acceptance or retention inequitable"); Com. Partn. 8098 Ltd. Partn. v. Eq. Contracting Co., Inc., 695 So. 2d 383, 386 (Fla. 4th Dist. App. 1997) (a plaintiff must prove "(1) the plaintiff has conferred a benefit on the defendant; (2) the defendant has knowledge of the benefit; (3) the defendant has accepted or retained the benefit conferred and (4) the circumstances are such that it would be inequitable for the defendant to retain the benefit without paying fair value for it").
In an attempt to reconcile these variations, Plaintiffs' proposed jury instructions organize the state laws into a limited number of permutations. (See Pls.' Reply, Meltzer Decl., Ex. 35.) However, as noted above, Plaintiffs' accounting of the state variations is not comprehensive and glosses over important differences. As such, Plaintiffs have not met their burden of demonstrating that grouping is a feasible method for addressing the variations amongst the purchase states' unjust enrichment laws. Consequently, I conclude that Plaintiffs have failed to meet their burden of demonstrating that common questions of law predominate.
Notwithstanding the variances in the applicable state law, the equitable nature of the unjust enrichment remedy also compounds the predominance issues with Plaintiffs' proposed unjust enrichment class. When considering an unjust enrichment claim, "a court must examine the particular circumstances of an individual case and assure itself that, without a remedy, inequity would result or persist." Vega v. T-Mobile USA, Inc., 564 F.3d 1256, 1274 (11th Cir. 2009); Grandalski, 767 F.3d at 185 ("individual inquiries would be required to determine whether an alleged overbilling constituted unjust enrichment for each class member"); Hernandez v. Ashley Furniture Indus., Inc., 2013 WL 2245894, at *9 (E.D. Pa. May 22, 2013) (unjust enrichment claim demands inquiry into the unique factual circumstances of each case to determine whether inequity would result). "Due to the necessity of this inquiry into the individualized equities attendant to each class member, courts . . . have found unjust enrichment claims inappropriate for class action treatment." Vega, 564 F.3d at 1274.
In light of this necessary inquiry, the Eleventh Circuit concluded "common questions will rarely, if ever, predominate an unjust enrichment claim, the resolution of which turns on individualized facts." Id. Other courts have reached similar conclusions. See Grandalski, 767 F.3d at 185 ("District Court properly found that individual inquiries would be required to determine whether an alleged overbilling constituted unjust enrichment for each class member"); In re Actiq Sales, 2015 WL 1312015, at *17; Hernandez, 2013 WL 2245894, at *9.
Nonetheless, Plaintiffs urge that common questions of fact predominate because Defendants' alleged common course of misconduct lies at the heart of the proposed class' unjust enrichment claims. Although Plaintiffs' claims do rely on some common proof, the existence of some common evidence as to Defendants' conduct does not dispose of the need for individualized inquiry into the equities surrounding the claims of individual Plaintiffs. See Commander Properties Corp. v. Beech Aircraft Corp., 164 F.R.D. 529, 540 (D. Kan. 1995) ("Even if it could be said that [the] general theory of liability for unjust enrichment . . . is uniform among class members, individual questions remain about whether" any plaintiff actually conferred a benefit).
As such, even if Plaintiffs had offered a sufficient analysis accounting for the variations in state law, I find that common factual issues do not predominate as to Plaintiffs' proposed unjust enrichment class.
The second inquiry under Rule 23(b)(3) is whether "a class action is superior to other available methods for fairly and efficiently adjudicating the controversy." Rule 23(b)(3) enumerates the following factors for assessing superiority:
(A) the class members' interests in individually controlling the prosecution or defense of separate actions;Fed. R. Civ. P. 23(b)(3).
(B) the extent and nature of any litigation concerning the controversy already begun by or against class members;
(C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and
(D) the likely difficulties in managing a class action.
In establishing superiority, a plaintiff must demonstrate that resolution by class action will "achieve economies of time, effort, and expense, and promote . . . uniformity of decision as to persons similarly situated without sacrificing procedural fairness or bringing about other undesirable results." Amchem, 521 U.S. at 615. The court must "balance in terms of fairness and efficiency, the merits of a class action against those of 'alternative available methods' of adjudication." In re Flonase, 284 F.R.D. at 234 (quoting Georgine., 83 F.3d at 632).
Where there are numerous state law causes of action at play, I must consider "whether variations in state laws present the types of insuperable obstacles which render class litigation unmanageable." In re Warfarin Sodium, 391 F.3d at 529. However, the superiority requirement may be satisfied where "varying state laws can be grouped by shared elements and applied as a unit in such a way that the litigation class is manageable." Id.
Plaintiffs argue that class litigation is superior to individual litigation because absent a class action, the same facts pertaining to Defendants' alleged course of conduct would have to be proven repeatedly in numerous cases. Plaintiffs urge that individual litigation would, therefore, waste resources and potentially cause inconsistent results.
Plaintiffs' concerns are not immaterial. However, for the reasons discussed above in the context of the predominance analysis, I find that the variations in state law also render class litigation unmanageable. Other courts have reached similar conclusions regarding the manageability of proposed multi-state unjust enrichment classes. See, e.g., Lilly v. Ford Motor Co., 2002 WL 507126, at *2 (N.D. Ill. Apr. 3, 2002) ("variations in state common laws of unjust enrichment demonstrate that class certification of such a claim would be unmanageable").
As noted above, Plaintiffs' proposed jury instructions do not address all of the variations which distinguish the unjust enrichment laws of the purchase states, nor do Plaintiffs' proposed jury instructions adequately account for the nuanced ways in which different states explain even seemingly similar elements. Therefore, Plaintiffs have failed to satisfy Rule 23(b)(3)'s superiority requirement as to their unjust enrichment claims.
VII. CONSUMER PROTECTION CLAIMS - PREDOMINANCE & SUPERIORITY
A. Choice of Law
Applying Pennsylvania's choice of law framework, I must first consider whether there is a true conflict between the relevant consumer protection laws. I begin by noting that other courts in this circuit confronted with proposed multi-state consumer protection classes have concluded that the laws vary in significant ways. See, e.g., Karnuth v. Rodale, Inc., 2005 WL 1683605, at *4 (E.D. Pa. July 18, 2005) ("[t]he consumer fraud statutes of the various states are not uniform"); Lyon v. Caterpillar, Inc., 194 F.R.D. 206, 219 (E.D. Pa. 2000) ("consumer protection acts vary on a range of fundamental issues"). Similarly, the Supreme Court has remarked that "no one doubts that a state may protect its citizens by prohibiting deceptive trade practices. . . . But the states need not, and in fact do not, provide such protection in a uniform manner." BMW of N. Am., Inc., v. Gore, 517 U.S. 559, 568-69 (1996).
Plaintiffs assert that the consumer protection laws share a few common "basic elements," which are as follows: "it is unlawful for a person to commit a deceptive or unfair act or practice, misrepresentation, false statement, or make an omission or material fact in connection with the sale or advertisement of merchandise." (Pls.' Mot, Meltzer Dec., Ex. 25.) Despite the foregoing characterization, the state consumer protection laws under which the proposed class pursues claims vary in material ways.
Some consumer protection laws require that a plaintiff prove that the defendant undertook the prohibited act with some level of intention. Where such a requirement is recognized, the states articulate the intent element in a variety of ways.
For example, Arizona, Minnesota and North Dakota require a plaintiff to establish that the defendant undertook the unfair conduct or deceptive practice with the intent that persons would rely on the prohibited act. Ariz. Rev. Stat. Ann. § 44-1522 (prohibited act must be done "with intent that others rely"); Minn. Stat. Ann. § 325F.69 (prohibited act must be done "with the intent that others rely thereon"); N.D. Cent. Code Ann. § 51-15-02 (prohibited act must be done "with the intent that others rely thereon"). Utah requires that the prohibited act be done knowingly or intentionally. Utah Code Ann. § 13-11-4 (prohibited act must be done knowingly or intentionally). New Mexico and South Dakota require that the prohibited act be done knowingly. Stevenson v. Louis Dreyfus Corp., 811 P.2d 1308, 1311 (N.M. 1991) (prohibited act must be done knowingly); S.D. Codified Laws § 37-24-6 (prohibited act must be done knowingly). Kansas requires that the prohibited act be done "knowingly or with reason to know." Kan. Stat. Ann. § 50-626. Lastly, Vermont requires that the act be done intentionally. Winton v. Johnson & Dix Fuel Corp., 515 A.2d 371, 376 (Vt. 1986).
Massachusetts and Hawaii require plaintiffs to prove that the alleged unfair act or deceptive practice was "immoral, unethical, oppressive, unscrupulous or substantially injurious to consumers." Balthazar v. Verizon Haw., Inc., 123 P.3d 194, 202 (Haw. 2005); Renovator's Supply, Inc. v. Sovereign Bank, 892 N.E.2d 777, 786-87 (Mass. App. 2008) (conduct is unfair if it is "within at least the penumbra of some common-law, statutory, or other established concept of unfairness; . . . it is immoral, unethical, oppressive, or unscrupulous; [and] . . . whether it causes substantial injury to consumers, competitors, or other business") (alterations in original). Defendants also state that Florida recognizes this requirement but that contention appears to be in dispute. See Porsche Cars N.A., Inc. v. Diamond, 140 So. 3d 1090, 1098 (Fla. 3d Dist. App. 2014) review denied, 157 So. 3d 1042 (Fla. 2014) (holding that this requirement no longer defines the term "unfair" as used in Florida's consumer protection law).
Additionally, Nebraska and Florida require a plaintiff to prove that the unfair act or deceptive practice affected public policy or the public interest. Nelson v. Lusterstone Surfacing Co., 605 N.W.2d 136, 142 (Neb. 2000) (to be actionable "the unfair or deceptive act or practice must have an impact upon the public interest"); PNR, Inc. v. Beacon Prop. Mgt., Inc., 842 So. 2d 773, 777 (Fla. 2003) (prohibited act must offend "established public policy").
Defendants assert that Hawaii also recognizes this requirement. However, the Hawaii law expressly provides "[n]o showing that the proceeding or suit would be in the public interest . . . is necessary in any action brought under this section." Haw. Rev. Stat. § 480-2.
Further complicating matters, Florida, Maine and Vermont require an additional showing that the unfair act or deceptive practice was likely to mislead consumers. State v. Beach Blvd Automotive Inc., 139 So. 3d 380, 387 (Fla. 1st Dist. App. 2014), reh'g denied (June 12, 2014) ("Deception occurs if there is a representation, omission, or practice that is likely to mislead consumers acting reasonably in the circumstances"); State v. Weinschenk, 868 A.2d 200, 206 (Me. 2005) (an act is deceptive if it is "likely to mislead consumers acting reasonably under the circumstances"); Greene v. Stevens Gas Serv., 858 A.2d 238, 244 (Vt. 2004) (act must be "likely to mislead the consumer").
Wisconsin also mandates that a plaintiff prove that the unfair act or practice was made to the public. K & S Tool & Die Corp. v. Perfection Mach. Sales, Inc., 732 N.W.2d 792, 798-99 (Wis. 2007).
States further vary as to whether a plaintiff must prove that they relied on the defendant's prohibited act. Pennsylvania requires a plaintiff to demonstrate that he relied on the defendant's act and that the reliance was justified. Kern v. Lehigh Valley Hosp., Inc., 108 A.3d 1281, 1289 (Pa. Super. 2015). Arizona also requires reliance but, unlike Pennsylvania, the reliance need not be reasonable, let alone justified. Correa v. Pecos Valley Dev. Corp., 617 P.2d 767, 771 (Ariz. App. 2d Div. 1980). In Michigan, class litigants "need not individually prove reliance on the alleged misrepresentations," as it is sufficient if the class can establish that "a reasonable person would have relied on the representations." Dix v. Am. Bankers Life Assur. Co. of Florida, 415 N.W.2d 206, 209 (Mich. 1987). Other states, such as Florida, do not require any showing of reliance. Davis v. Powertel, Inc., 776 So. 2d 971, 974 (Fla. 1st Dist. App. 2000).
Although not exhaustive, the foregoing list illustrates the many varying elements and nuanced articulations that distinguish the state consumer protection laws. Contrary to Plaintiffs' characterization, these laws cannot be neatly distilled into a core set of elements with only minor distinctions.
Furthermore, these differences are of such a nature that a state's interests would be impaired by application of another state's law. For example, some consumer protection laws require proof of the defendant's intent and others do not. The states without an intent requirement or even those states with a less searching intent requirement create a lower evidentiary hurdle for the plaintiffs than states with a strict intent requirement. The particular formulation of these variables reflects a state's policy decisions regarding how and when redress should be available to consumers. As such, I find that a conflict exists between the relevant state consumer protection laws.
Other courts have reached similar conclusions. See In re Actiq Sales and Mktg. Practices Litig., 790 F. Supp. 2d 313, 321 (E.D. Pa. 2011) (finding a true conflict between the consumer protection laws of Indiana and Pennsylvania in light of differences in the range of prohibited acts and intent requirements); Zeneca, 710 F. Supp. 2d at 471 (finding "an actual conflict exists between the laws of Delaware and Pennsylvania on the issue of whether reliance is a necessary element under the respective consumer fraud statutes").
Having found that a true conflict exists, I must determine which state or states have a greater interest in application of its consumer protection laws. Pursuant to Pennsylvania's choice of law rules, the first step is to consider the relevant Restatement factors.
The Restatement (Second) of Conflict of Laws § 148 sets forth contacts that are to be considered when assessing which jurisdiction has the most significant relationship to the occurrence giving rise to a fraud or misrepresentation claim. See Lyon, 194 F.R.D. at 214 (applying § 148 to a consumer protection claim under Pennsylvania's choice of law rules).
Section 148 provides that "[w]hen the plaintiff's action in reliance took place in whole or in part in a state other than that where the false representations were made," courts should weigh the following factors:
(a) the place, or places, where the plaintiff acted in reliance upon the defendant's representations,Restatement (Second) of Conflict of Laws § 148(2).
(b) the place where the plaintiff received the representations,
(c) the place where the defendant made the representations,
(d) the domicil, residence, nationality, place of incorporation and place of business of the parties,
(e) the place where a tangible thing which is the subject of the transaction between the parties was situated at the time, and
(f) the place where the plaintiff is to render performance under a contract which he has been induced to enter by the false representations of the defendant.
Factors (a) and (b) weigh in favor of applying the laws of the purchase states. Plaintiffs received and acted in reliance on Defendants' representations in the state in which Plaintiffs purchased, paid and/or reimbursed for Provigil or its generic equivalent.
Regarding factor (c), a representation is "made" in the "location from which the representation emanated." Maniscalco v. Bro. Intern. (USA) Corp., 709 F.3d 202, 208 (3d Cir. 2013). Defendants likely made the representations from their principal places of business, Pennsylvania, Florida and New York. As such, this factor weighs against application of the consumer protection laws of the purchase states.
Regarding factor (d), Defendants' principal places of business are located in Pennsylvania, Florida and New York. On the other hand, Plaintiffs likely reside or conduct business in all states in which Plaintiffs pursue claims. At first glance, factor (d) appears neutral However, the Restatement notes that "[t]he domicil, residence and place of business of the plaintiff are more important than are similar contacts on the part of the defendant." Restatement (Second) of Conflict of Laws § 148 cmt. 2(i). As such, factor (d) also weighs in favor of applying the consumer protection of laws of the purchase states.
The relative importance of the foregoing factors should be further assessed in light of the following principles set forth in Section 6 of the Restatement: "(1) the interests of interstate comity; (2) the interests of the parties; (3) the interests underlying the field of tort law; (4) the interests of judicial administration; and (5) the competing interests of the states." Restatement (Second) of Conflict of Laws § 6 cmt 2(e).
Regarding the second and third principles, the basic policies underlying the particular field of law are those of "consumer protection, suggesting that any balancing of the parties' contacts or expectations should be weighted toward those of consumers." In re Relafen Antitrust Litig., 221 F.R.D. 260, 277-78 (D. Mass. 2004). As such, the location of consumers' purchases assumes "special significance." Id. Regarding the first and fifth factors, "[s]tates have a strong interest in protecting consumers with respect to sales within their borders . . . but they have a relatively weak interest, if any, in applying their policies to consumers or sales in neighboring states." Id.
As such, the Restatement factors viewed through the prism of the Section 6 principles weigh in favor of applying the law of the purchase state. However, under Pennsylvania's choice of law rules, I must also examine the relevant state policies at issue to determine which state has the greatest interest in application of its law.
A "qualitative appraisal" of the relevant states' policies at issue also suggests that application of the law of the purchase states is appropriate. Consumer protection laws "are intended to protect consumers from being overcharged." In re Flonase, 815 F. Supp. 2d at 883. Therefore, "the purchase states have a serious interest in applying their law to allow consumers . . . to recover the money that they were overcharged in a transaction occurring in their states." Id.
However, states also clearly have "an independent interest in preventing deceptive or fraudulent practices by companies operating within their borders." Pilgrim v. Universal Health Card, LLC, 660 F.3d 943, 946-47 (6th Cir. 2011). Nonetheless, the "primary aim of antitrust and consumer protection laws generally—and those of indirect purchaser states particularly—is compensating consumers, not policing corporate conduct." In re Relafen, 221 F.R.D. at 277.
I agree, that given that the primary policy interest is consumer protection, the state with the strongest interest in regulating such conduct is the state "where the consumers—the residents protected by its consumer-protection laws—are harmed by it." Pilgrim, 660 F.3d at 946-47 (emphasis in original). As such, I find that the purchase states' interest in application of their own consumer protection law to be the greatest.
For reasons similar to those discussed in the context of Plaintiffs' unjust enrichment claims, Plaintiffs have not demonstrated that common questions of law predominate as to their consumer protection claims. Again, Plaintiffs failed to provide the extensive analysis of the variations in the consumer protection laws necessary for determining whether there are insuperable obstacles to class certification.
Rather, Plaintiffs attempt to meet their burden by distilling a common core of "basic elements" from the various consumer protection laws. Plaintiffs then annotate these purported "basic elements" with additional elements where recognized by a particular state.
However, Plaintiffs' state law charts and proposed jury instructions gloss over material differences. For example, one of Plaintiffs' proposed jury instructions states:
The consumer protection laws of Arizona, Maine, Minnesota, New Mexico, North Dakota, South Dakota, Utah and Vermont additionally require that the defendant know or intend that the unfair deceptive act, practice or omission was likely to mislead others, or that others would rely on, or be deceived by, said act, practice or omission.(Pls.' Reply, Meltzer Dec., Ex. 35.)
As noted above, the states vary considerably in their formulation of the intent a plaintiff must prove. The jury instruction proposed by Plaintiffs fails to distinguish which state requires which particular intent formulation and disregards important differences amongst those formulations. The proposed jury instruction also fails to explain how each state defines operative terms. Plaintiffs' effort to explain away these differences through a generalized explanation is overly simplistic. Therefore, I find that Plaintiffs failed to provide a basis for concluding that common legal questions will predominate as to their consumer protection claims.
Other courts have reached similar conclusions regarding proposed multi-state or nationwide consumer protection classes. See Pilgrim, 660 F.3d at 946-48 ("the consumer-protection laws of the affected States vary in material ways, no common legal issues favor a class-action approach to resolving this dispute"); Karnuth, 2005 WL 1683605, at *5 (declining to certify a nationwide consumer protection class in light of the variations in state law); Lyon, 194 F.R.D. at 220 (denying certification under forty-one state consumer protection laws in light of variations in the applicable law).
As discussed above, Plaintiffs failed to offer the requisite extensive analysis of how the differences in the state consumer protection laws would be overcome. Relevant to the manageability inquiry, Plaintiffs have failed to demonstrate how the jury could be instructed in a manageable and accurate fashion. See Powers, 328 Fed. Appx. at 127 ("[a]ttempting to apply the law of a multiplicity of jurisdictions can present problems of manageability for class certification under Rule 23(b)(3)").
Plaintiffs state that a jury could be instructed as to each state's relevant laws individually or as to the laws of grouped states. Plaintiffs contend that "[r]egardless of which of these two approaches the Court adopts, the instruction process is manageable, and the jury is not likely to be confused." (Pls.' Mot., Meltzer Dec., Ex. 24, p. 7.) However, Plaintiffs have not offered a plan of how that grouping may be accomplished in a manner that does not gloss over important substantive differences between the laws. Plaintiffs' assurance that such an instruction process is manageable and not likely to confuse the jury is insufficient. As such, I find that Plaintiffs have also failed to demonstrate that a class action is a fair and efficient method for adjudicating the consumer protection law claims of the proposed class.
For the reasons stated above, I find that certification of the End Payor class is not appropriate because Plaintiffs have failed to establish the Rule 23 requirements of ascertainability, predominance and superiority by a preponderance of the evidence. Accordingly, Plaintiffs' motion for class certification is denied. (See Doc. No. 433.)