Opinion
No. 08-md-2002
08-14-2017
IN RE: PROCESSED EGG PRODUCTS ANTITRUST LITIGATION THIS DOCUMENT APPLIES TO: ALL DIRECT PURCHASER ACTIONS
MULTIDISTRICT LITIGATION
MEMORANDUM PRATTER, J.
I. INTRODUCTION
Undeterred by the Third Circuit Court of Appeals' denial of their Rule 23(f) petition to appeal the Court's certification of the Direct Purchaser Plaintiffs' Class, Defendant egg producers seek to decertify the Direct Purchaser Plaintiffs' Class. Defendants primarily contend that "three seminal events"—two of which occurred prior to the Court's initial ruling—demand decertification. The Court disagrees. After reviewing the parties' submissions and hearing oral argument, the Court denies the motion to decertify the Direct Purchaser Plaintiffs' Class.
II. BACKGROUND
A summary of the relevant factual background and the class certification standard under Rule 23 appears, in detail, in the Court's opinion granting certification for the Direct Purchaser Plaintiff Shell Egg Class. In re Processed Egg Products Antitrust Litig., No. 08-MD-2002, 2015 WL 5610834 (E.D. Pa. Sept. 21, 2015), opinion amended and superseded by In re Processed Egg Products Antitrust Litig., No. 08-MD-2002, 2015 WL 7067790 (E.D. Pa. Nov. 12, 2015) (hereinafter "Class Cert. I at ___"). It is not necessary to rehash that information here, except to the extent now of challenges to specific issues addressed, and observations made, in the Class Cert I opinion.
In Class Cert I, the Court granted in part the Direct Purchaser Plaintiffs' (DPPs) motion for class certification and certified a class of shell egg direct purchasers, finding that the shell egg subclass satisfied Rule 23's requirements. A party seeking class certification must show that the prerequisites of Rule 23(a) are met: numerosity, commonality, typicality, adequacy and ascertainability. Fed. R. Civ. P. 23(a). Of Rule 23(a)'s requirements, typicality and adequacy were the most hotly contested. The Court ultimately rejected Defendants' arguments that differences in pricing or purchasing arrangements, potential individualized defenses, and purported insufficiencies with class representatives precluded a finding of typicality and adequacy. See Class Cert I at 179-80.
A class must also show how it satisfies at least one of Rule 23(b)'s standards. The Court concluded in Class Cert I that the shell egg subclass met Rule 23(b)(3)'s "predominance" and "superiority" requirements. The Court expended most of its breath in Class Cert I on Rule 23(b)(3)'s predominance requirement and waded into the evidence put forth by the Plaintiffs' expert, Dr. Gordon Rausser.
At the certification stage, Defendants had argued that the Supreme Court's decision in Comcast Corp. v. Behrend, 133 S. Ct. 1426 (2013), commands that Plaintiffs must first measure the decrease in supply and then measure the effect on the price. The Court declined to adopt such a broad view of Comcast. It distinguished Dr. Rausser's model from the model at issue in Comcast because this case was in a different posture. None of the alleged means of reducing the supply of eggs had been found inappropriate for class treatment and Defendants had not challenged that "any disaggregated part of the alleged conspiracy should be found lawful or otherwise incapable of common proof." Class Cert I at 192. However, the Court noted at the time that while not just any adverse ruling would threaten Dr. Rausser's model, the model could face a Comcast problem "if (a) certain conduct were found to have occurred; and (b) that conduct had an impact on the price of eggs; but (c) that conduct was not legally cognizable vis-à-vis the class." Class Cert I at 192 n.15.
While the Court expressed some reservations about whether Dr. Rausser's methodologies were bulletproof, it ultimately concluded that "Plaintiffs have shown that common evidence is capable of demonstrating that Defendants engaged in a series of complementary supply-reducing actions as part of a conspiracy to increase the price of eggs. Dr. Rausser has measured whether that conspiracy was successful in increasing the price of eggs. His model fits the theory of liability and satisfies Comcast." Class Cert. I at 193. The Court thus found that common issues predominate with respect to the antitrust injury to the shell eggs subclass. See Class Cert. I at 199-200.
Pursuant to Federal Rule of Civil Procedure 23(f), Defendants then petitioned the Third Circuit Court of Appeals for permission to appeal the class certification order. The Court of Appeals denied their petition.
At the time the Court certified the Direct Purchaser Plaintiff Shell Egg Class, it left unresolved the class period determination and requested that the parties submit additional briefing on the parties' proposed cutoff dates. After reviewing the parties' submissions, the Court held that the proper cut-off date for the class is December 31, 2008. In re: Processed Egg Prod. Antitrust Litig., No. 08-MD-2002, 2016 WL 410279, at *1 (E.D. Pa. Feb. 3, 2016) (hereinafter "Class Cert. II at ___"). The Court certified the following Direct Purchaser Plaintiff Shell Egg Class:
All individuals and entities that purchased shell eggs produced from caged birds in the United States directly from Defendants during the Class Period from September 24, 2004 through December 31, 2008.Class Cert II at *8.
Excluded from the Class are the Defendants, their co-conspirators, and their respective parents, subsidiaries, and affiliates, as well as any government entities. Also excluded from the Class are purchasers of "specialty" shell eggs (such as "organic," "certified organic," "free range," "cage free", "nutritionally enhanced," or "vegetarian fed") and purchasers of hatching eggs, which are used by poultry breeders to produce breeder stock or growing stock for laying hens or meat.
In Class Cert II, the Court again addressed the question of whether Dr. Rausser's model violated Comcast's principles. In their briefing on the class cutoff, Defendants explained that from December 2008 onward, a number of states passed legislation that had the effect of legislating minimum cage size requirements. Class Cert II at *7. The Court recognized that "multiple state legislative schemes in place after 2008, mandating producers impose cage-space limitations, would create the possibility that the effect of the Defendants' conduct post-2008 could not be measured using Dr. Rausser's proposed model." Id. The Court observed that Dr. Rausser's "model does not, however, account for changes in regulatory requirements imposed by state governments upon producers" and therefore, "[i]n much the same way the Supreme Court found impermissible in Comcast, the price variance Dr. Rausser attributes to the conspiracy intermingles lawful and unlawful behavior between 2008 and 2013." Id. at 8. Accordingly, "[g]iven that the Rausser Model cannot reliably show damages resulting from the alleged conspiracy on a class wide basis after December 31, 2008, the Court f[ound] that the proper class cutoff date for the shell egg subclass is December 31, 2008." Id. at *8.
Defendants have moved to decertify the shell eggs class motivated in large part by their interpretation of the reach of the Court's holding in Class Cert II, and use their challenge to that decision as a means of raising additional, separate arguments.
LEGAL STANDARD
Under Fed. R. Civ. P. 23(c)(1)(C), an order granting class certification "may be altered or amended before final judgment." After certification has been granted, courts "remain[] free to modify it in the light of subsequent developments in the litigation." Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 160 (1982). In the Third Circuit, district courts are required to reassess their class rulings regularly as the case develops. Barnes v. Am. Tobacco Co., 161 F.3d 127, 134 n.4, 140 (3d Cir. 1998) (noting that "[t]he district judge must define, redefine, subclass, and decertify as appropriate in response to the progression of the case from assertion to facts" (quoting Richardson v. Byrd, 709 F.2d 1016, 1019 (5th Cir.1983))).
While district courts must remain vigilant in ensuring that class certification is proper as the landscape of the litigation evolves over time, decertification must be justified by actual, compelling developments in the litigation. Generally, class decertification is prompted by a change in factual circumstances or developments in applicable substantive or procedural law. See, e.g., Markocki v. Old Republic Nat. Title Ins. Co., No. CIV.A. 06-2422, 2015 WL 3421401, at *3 (E.D. Pa. May 27, 2015) ("Reconsideration of an earlier class certification order is warranted upon developments in the litigation such as changes in substantive or procedural law."); Bayshore Ford Truck v. Ford Motor Co., No. CIV. A. 99-741 JLL, 2010 WL 415329, at *2 (D.N.J. Jan. 29, 2010) (declining to find that "a court's class certification ruling should simply be reconsidered for any reason" and instead concluding that reevaluation of a class ruling should occur where new factual developments warranted it); Elias v. Ungar's Food Prod., Inc., No. CIV.A. 06-2448KSH, 2009 WL 2581502, at *5 (D.N.J. Aug. 20, 2009) ("In order to warrant motion practice to decertify the earlier-certified class, the Court must be presented with a changed factual situation that renders the original certification 'unsound.'" (citing Zenith Labs., Inc. v. Carter-Wallace, Inc., 530 F.2d 508, 512 (3d Cir. 1976))).
In the first instance and throughout the life of a case, courts must conduct a "rigorous analysis" to ensure that Rule 23's requirements are met. In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 318 (3d Cir. 2008). The Court must find that each of Rule 23's standards has been met by a preponderance of the evidence. Id. at 321. In determining whether these standards have been met, the realities of the tasks at hand dictate that, to some extent at least, the Court necessarily needs to examine issues that may partially overlap issues to be fully decided at the final merits determination. Id. at 316. However, exercising the analytical restraint often demanded of judges, for certification purposes, the Court only examines the merits of the underlying case to the extent such an examination is relevant to the Rule 23 standards. Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S. Ct. 1184, 1195 (2013).
DISCUSSION
Defendants Rose Acre Farms, Inc., Michael Foods, Inc., and Ohio Fresh Eggs, Inc. urge the Court to undo the decision to certify a class of shell egg direct purchasers. Specifically, Defendants mount three main challenges centered on purported limitations of the models by Direct Purchaser Plaintiffs' expert, Dr. Gordon Rausser. First, they present a modified iteration of their Comcast argument from Class Cert II, arguing that, in light of the Court's decision to cut off the class period on December 31, 2008, post-2008 data should not be included in Dr. Rausser's model. When Dr. Rausser's econometric model is adjusted to reflect the December 2008 damages cutoff, Defendants urge that Dr. Rausser's analysis no longer supports the Plaintiffs' theory. Second, Defendants take issue with an alleged inconsistency between Dr. Rausser's class reports and merits reports. Holding the class and merits reports against each other, Defendants argue that the Court is again faced with a Comcast issue. Finally, Defendants urge that while Dr. Rausser has recognized on the record that accounting for individual factors is paramount for accurate modeling, he has failed to follow his own admonition, rendering his results meaningless.
While Michael Foods, Inc. participated in filing the Motion, the Court has stayed all activity involving Michael Foods in light of its settlement with Direct Purchaser Plaintiffs. See Doc. No. 1477. The Court has granted preliminary approval of the Michael Foods settlement, and a fairness hearing will be held on November 6, 2017. Doc. No. 1523.
Defendants primarily take aim at the Court's predominance ruling in Class Cert I, but they also argue in passing that the class representatives fail to meet the typicality and adequacy requirements of Rule 23(a). Again, neither of Defendants' arguments is premised on any change or development taking place after the Court issued Class Cert I. Defendants' arguments on typicality and adequacy appear rooted in the observation that between 10% and 20% of the class did not purchase shell eggs after 2006, but that 2007 is the first year Dr. Rausser's layer and production analysis shows a supply reduction. Plaintiffs point out, however, that Dr. Rausser's pricing model shows price elevation throughout, even if the layer and production analysis—which is admittedly conservative—does not show a reduction in egg production until the later years of the conspiracy. Accordingly, the Court declines to revisit its holdings on typicality and adequacy in Class Cert I on the basis of the layer and production analysis.
The Court is unpersuaded by Defendants' attempts to invalidate the Court's prior predominance ruling. Most of these arguments could and should have been raised long before the Court issued its opinion certifying the class in September 2015. Indeed, the Motion to Decertify essentially recasts many of the arguments the Court has already addressed and dispensed with at various stages of this litigation. To the extent that the Motion raises a new issue for the Court to decide, it is whether the Court's Order cutting off the class period on December 31, 2008 calls into question its prior class certification. This is the argument that the Court will now address.
Defendants challenge the Court's prior finding that Plaintiffs satisfied Rule 23(b)'s predominance requirement by pointing to alleged flaws in Dr. Rausser's methodologies and conclusions upon which the Court relied in certifying the class. Defendants admitted in their briefing and at oral argument that they did not revisit Dr. Rausser's econometric analysis supporting antitrust injury and class wide damages—which was thoroughly addressed at the March 2015 class certification hearing and in detail in Class Cert I—until the Court issued its Class Cert II decision nearly a year later. Setting aside the fact that Defendants had ample time before the Court initially certified the class to raise the purported issues with Dr. Rausser's analysis that they now identify, Class Cert II does not itself alter the Court's prior conclusions on predominance. Indeed, if it did so, the Court would have concluded as much when it set the cutoff period. Instead, the Court affirmed that "Plaintiffs have established that the Rausser model is capable of quantifying damages attributable to Defendants' anticompetitive behavior for 2004 through 2008 . . . ." Class Cert II at *8.
In their motion, Defendants take the Class Cert II's 2008 cutoff date a step further by arguing that post-2008 data must be removed from Dr. Rausser's model. In Class Cert II, the Court recognized that "[i]n much the same way the Supreme Court found impermissible in Comcast, the price variance Dr. Rausser attributes to the conspiracy intermingles lawful and unlawful behavior between 2008 and 2013." Class Cert II at *8. This determination, along with the Court's warning in Class Cert I that intermingling legally cognizable conduct could present Comcast problems propels Defendants to argue, yet again, that Dr. Rausser's models cannot satisfy Comcast.
Defendants take issue with the fact that Dr. Rausser's overcharge regression includes—as the Court has recognized—data affected by state regulatory regime changes post-2008. They assert that applying an overcharge regression that incorporated post-2008 data (and therefore the price effects, if any, of the state regulatory regimes) to the Class Period itself presents the same Comcast problem that motivated the Court to cut off the Class period in 2008 in Class Cert II. The remedy to the purported problem with Dr. Rausser's model, according to Defendants, is to remove all post-2008 data. When Defendants' expert, Dr. Jonathan Walker, did so, he found that Dr. Rausser's overcharge regression was 3.6% lower between September 2000 and December 2008 than the benchmark period. Walker Decl. ¶ 8. The "disappearance" of the overcharge from September 2000 to December 2008, according to Defendants, demonstrates that Plaintiffs can no longer show that class members suffered antitrust injury or damages using common proof.
Obviously Class Cert II followed Class Cert I. Therefore, it could prompt the need for additional clarification. The argument that all post-2008 data should be removed was available before the Court issued Class Cert I, however. Indeed, Dr. Walker raised the issue of including post-2008 data in his merits report. See Supplemental Report of Jonathan Walker Related to Direct Purchaser Plaintiffs ¶ 55 ("I have re-run Dr. Rausser's regression based on data from February 1997 through December 2008. Re-running Dr. Rausser's price regression this way yields substantially different results. . . . Based on Dr. Rausser's model, shell egg prices were 3.56% (or $159 million) lower on average in the period September 2000-December 2008 than they were in Dr. Rausser's benchmark period."). While Defendants may have been emboldened by the Court's ruling in Class Cert II to revisit their strategy, they were fully able to mount the challenge they now raise before Class Cert I was issued.
Plaintiffs argue, and the Court agrees, that removing all post-2008 data from Dr. Rausser's model is unnecessary. When it set the class cutoff date, the Court recognized that "the UEP Certification program continued through 2013 and included the same types of instrumentalities as are alleged to have characterized the conspiracy pre-2008." Class Cert II at *3. The Court again recognized at summary judgment, that "it is undisputed that the defendants continued to operate as members of the UEP Certified Program after the lawsuits were filed [in 2008]. . . . [and] the plaintiffs have put forward evidence that allows them to argue that the effects of this conspiracy continued through at least 2012." In re: Processed Egg Prod. Antitrust Litig., No. 08-MD-2002, 2016 WL 4771865, at *4 (E.D. Pa. Sept. 12, 2016). In other words, there is evidence for a jury to consider that the same allegedly conspiratorial conduct that occurred pre-2008 occurred after 2008. The basis on which the Court cut off the class period in 2008 was simply that Dr. Rausser's model could not parse out conduct that was legal in some states from conduct that was illegal in other states after 2008. The model in Comcast was similarly problematic because that model was unable to isolate damages arising from the one theory of liability suitable for class treatment. Accordingly, the Court determined that the class period would end on December 31, 2008.
Cutting off the class period in 2008 would not automatically invalidate the usefulness of post-2008 data. The Supreme Court in Comcast focused on the rigidity of the model given the viable theories in the case, not the underlying data itself. It did not conclude that particular data was tainted, but rather that the model itself failed to "bridge the differences between supra-competitive prices in general and supra competitive prices attributable to [the one theory viable for class treatment]." 133 S. Ct. at 1435. That the Court here has recognized a similar rigidity for Dr. Rausser's model after 2008 does not command taking the next step of stripping post-2008 data from the model itself. The Court's holding in Class Cert II was simply a recognition and application of the model's limit on measuring damages in light of the circumstances. Accordingly, the Court rejects this attempt to, yet again, invalidate Dr. Rausser's model under Comcast. To the extent that there is disagreement over the models, or the extent to which post-2008 data itself is relevant given that the Court has concluded the allegedly conspiratorial conduct continued after 2008, the Court determines that question is best answered by a jury. See In re Mushroom Direct Purchaser Antitrust Litig., No. 06-0620, 2015 WL 5767415, at *13 (E.D. Pa. July 29, 2015 (concluding "that 'arguments about how the selection of data inputs affect the merits of the conclusions produced by an accepted methodology should normally be left to the jury.'" (quoting Manpower, Inc. v. Ins. Co. of Pa., 732 F.3d 796, 808 (7th Cir. 2013)).
In addition to Defendants' contention that, in essence, the Court's decision in Class Cert II undermines its decision in Class Cert I, they raise additional arguments supporting removal of post-2008 data. They contend, inter alia, that a "structural break" occurred in 2008 and take the position that data cannot be aggregated or pooled when there is a divergence between the effects of the explanatory variables in different time periods. The Court concludes that removal of post-2008 data is not analytically required under Comcast, but further, removal of the data could be problematic because, for instance, removing all post-2008 would involve excluding 52% of the available data. The Court declines, at this juncture of the case, to find that the phenomena raised by Defendants amount to a "structural break" that invalidates post-2008 data.
Defendants mount another Comcast argument, urging that purported inconsistencies between Dr. Rausser's overcharge regression and his egg production model (information which was available to the parties when the Court decided Class Cert I) show a disconnect between the impact and damages model and the Plaintiffs' theory of the case, undermining predominance. Defendants compare Dr. Rausser's production model, which they state shows no supply decrease until 2007, with the regression's 19.8% overcharge spanning from 2000 to 2013. Juxtaposing the production model against the overcharge progression, Defendants state that the overcharge cannot be caused by the alleged conspiracy and instead must be attributable to some other cause. This inconsistency is evidence of an "insurmountable Comcast problem," according to Defendants.
The Court previously rejected Defendants' argument in Class Cert I that "Comcast requires that Plaintiffs first measure the extent to which Defendants' actions decreased the supply of eggs and then measure the effect on the price." Class Cert I at 191. Instead, it agreed with Plaintiffs' reading of Comcast and explained that in Comcast, the model's inability to isolate damages attributable to the single theory of impact that was suitable for class treatment (among the four proposed) precluded finding that common issues would predominate. In other words, "the model could not determine whether the theory suitable for class treatment caused the damages found by the model." Id. The Court has already accepted Dr. Rausser's overcharge regression and egg production modeling, and will not take the opportunity at this time to undo any of its prior conclusions, particularly because these arguments could have been raised, or in some manner were raised, prior to the Court issuing Class Cert I.
Defendants also attack Dr. Rausser's common factors and overcharge regressions on the basis that he inappropriately pooled customer data and did not control for the individual factors that he testified affect the price of eggs. As evidenced by Defendants' ample citations to Dr. Rausser's testimony at the class certification hearing in their briefing, they and the Court were well aware of the nature of Dr. Rausser's models at the time of the hearing and when the Court issued Class Cert I. The Court will not revisit arguments at this juncture that undoubtedly could have, and in many respects were, raised previously regarding customer data and individual factors. --------
CONCLUSION
For the foregoing reasons, the Court will deny the Defendants' Motion to Decertify the Direct Purchaser Plaintiffs' Class. An appropriate Order follows.
BY THE COURT:
/s/_________
GENE E.K. PRATTER
United States District Judge