United States District Court, S.D. New York
April 24, 2014
IN RE: ELECTRONIC BOOKS ANTITRUST LITIGATION.
PENGUIN GROUP (USA) INC., et al., Defendants. THE STATE OF TEXAS, et al., Plaintiffs,
Steve W. Berman, George W. Sampson, Seattle, WA, Jeff D. Friedman, Shana Scarlett, Hagens Berman Sobol Shapiro LLP, Berkeley, CA, Kit A. Pierson, Jeffrey Dubner, Cohen Milstein Sellers & Toll PLLC, Washington, D.C., Douglas Richards, Cohen Milstein Sellers & Toll PLLC, New York, NY, for class plaintiffs.
Eric Lipman, Gabriel Gervey, David Ashton, Kayna Stavast-Piper, Office of the Attorney General of Texas, Austin, TX, for plaintiff States, for State of Texas, Liaison Counsel for Plaintiff States.
W. Joseph Nielsen, Gary M. Becker, Office of the Attorney General of Connecticut, Hartford, CT, for State of Connecticut, Liaison Counsel for Plaintiff States.
Robert L. Hubbard, Office of the Attorney General of New York, New York, NY, for State of New York, Liaison Counsel for Plaintiff States.
Theodore J. Boutrous, Jr., Daniel G. Swanson, Gibson, Dunn & Crutcher LLP, Los Angeles, CA, Cynthia Richman, Gibson, Dunn & Crutcher LLP, Washington, D.C., for defendant Apple Inc.
OPINION & ORDER
DENISE COTE, District Judge.
After a bench trial in two closely related cases brought by the United States Department of Justice and thirty-three states and U.S. territories,  defendant Apple Inc. ("Apple") was found to have colluded with five major publishers (the "Publisher Defendants") to fix e-book prices in violation of Section 1 of the Sherman Antitrust Act ("Sherman Act"), 15 U.S.C. § 1. On March 28, 2014, a class was certified in the related class action (the "Class Action"). In re: Electronic Books Antitrust Litig., 11 MD 2293 (DLC), 2014 WL 1282293 (S.D.N.Y. Mar. 28, 2014) (the "Class Certification Opinion"). A joint damages trial in the States' Action and Class Action is scheduled to begin on July 14, 2014.
On April 4, Apple moved to stay the Class and States' Actions pending Apple's submission and the Court of Appeals's review of a petition for interlocutory appeal of the Class Certification Decision (the "Rule 23(f) Petition"). Although the lion's share of its briefing is devoted to argument in favor of a stay during review of its Rule 23(f) Petition, Apple also suggests, in places, that these actions should be stayed pending Apple's merits appeal of the liability decision. The parties agreed on an expedited schedule for briefing this motion for a stay, and the motion was fully submitted on April 15. By Order of April 23, Apple's stay motion was denied for reasons to be set forth in a later Opinion. This Opinion gives the reasons for that denial.
The relevant procedural history is set out below. Although familiarity with the Class Certification Opinion is assumed, a brief description of the Class Action's expert's damages model is also set forth.
I. The Late Date of Class Certification, and Delay of the Damages Trial
Well before the June - liability trial, at a lengthy conference held on October 26, 2012 addressed to pre-trial scheduling and discovery matters in the DOJ, States' and Class Actions, the Court inquired why Apple wished to delay class certification motion practice until after the liability trial. Apple stated that "the class certification process, we believe, can really be handled in a reasonable time" and "[we] think it makes sense to have it handled after" the liability trial. Apple represented that "not having the class go forward and opt out [at that time], certainly in no way will harm the class." Accordingly, the Court agreed to postpone class certification until after the liability trial.
Class plaintiffs moved for class certification on October 11, 2013. In support of that motion, class plaintiffs submitted the expert report of Dr. Roger Noll ("Noll"), who reported the results of a sophisticated damages model built from a multivariate regression analysis of more than 149 million e-book sales. Noll's initial model explained 90% of the variance in prices among e-book titles. The parties' briefing on the class certification motion was fully submitted on January 21, 2014 following a sur-reply from Apple. Motions to exclude Apple's experts, who offered opinions in opposition to class certification, were fully submitted on February 4. On March 28, class certification was granted. To allow for a 45-day notice period and to accommodate certain pre-trial filings, the Court delayed the damages trial from May 2014 to July 14, 2014.
II. Two Prior Notices
Prior to the June - liability trial in the DOJ Action and States' Action, each of the five Publisher Defendants settled with the DOJ, the States, and class plaintiffs. In connection with those settlements, two notices were sent to affected e-book purchasers - including all class members in the Class Action - advising them of the pendency of these actions and noting the allegations against the Publisher Defendants and Apple.
On September 13, 2012, the Court preliminary approved a $69 million settlement between all states and U.S. territories (sans Minnesota) and three of the Publisher Defendants: Hachette Book Group, Inc. ("Hachette"), HarperCollins Publishers, LLC ("HarperCollins"), and Simon & Schuster, Inc. and Simon & Schuster Digital Sales, Inc. ("Simon & Schuster"). The Court also approved the first plan to notify affected e-book customers. Pursuant to that notice plan, e-retailers ("e-tailers") Amazon, Barnes & Noble, Apple, Kobo, Sony, and Google each identified customers who purchased one or more of the Publisher Defendants' e-books between April 1, 2010 and May 21, 2012 (the "class period"). Each e-tailer then sent by e-mail a one-page notice to affected customers. This first e-mail notice advised that those settlements "resolve an antitrust lawsuit about the price of electronic books, " while "[a] separate lawsuit against two other publishers and Apple, Inc. continues and is set for trial in -."
In addition, the Claims Administrator set up a dedicated website at www.EbooksAGSettlements.com with further information about the litigation, including a more detailed notice. Kinsella Media, LLC, an advertising and legal notification firm, arranged for supplemental notice to be made through internet banner advertising, with banners that appeared on websites including Facebook's; sponsored links on the most popular U.S. search engines; mobile device advertising; advertising in newspapers in U.S. territories and possessions; press releases distributed via a national newswire and through promoted stories; outreach to more than 300 blogs covering book-related topics; and posts to relevant Twitter accounts. Rust Consulting, Inc. ("Rust") also sent postcard notices to Apple and Sony customers whose e-mail notices were returned as undeliverable.
On July 12, 2013, the Court approved a second plan to notify consumers of $95 million settlements with Publisher Defendants Holtzbrinck Publishers, LLC d/b/a Macmillan ("Macmillan") and Penguin Group (USA) Inc. ("Penguin"), as well as Minnesota's settlement with all five of the Publisher Defendants. Notice of the former settlement went to customers in the States who purchased the Publisher Defendants' e-books during the class period, as well as to putative class members in the Class Action who purchased such e-books; notice of the latter went to Minnesota customers who purchased the Publisher Defendants' e-books during the class period.
The second notice plan largely mirrored the first. Much like the first, the second e-mail notice advised that the settlements resolved claims against these Publisher Defendants "in antitrust lawsuits about the price of electronic books, " while "[t]he antitrust lawsuit against Apple, Inc. continues."
On March 28, 2014, a class was certified in the Class Action. Rule 23 requires individual notice to all class members who can be identified through reasonable effort, advising them, among other things, of their right to opt-out of the litigation. Fed.R.Civ.P. 23(c)(2)(B). Similarly, the Clayton Antitrust Act ("Clayton Act") requires that the States, which have brought this action parens patriae on behalf of their residents, publish notice and permit residents to opt-out. 15 U.S.C. § 15c(b). Having been asked by the Court to prepare notice submissions in the event class certification was granted, plaintiffs moved for approval of a notice plan the same day.
The proposed notice plan is much like the prior plans. A short, half-page notice is to be sent by e-mail to class members and to customers in the States who purchased the Publisher Defendants' e-books during the class period, from their e-tailer, or from Rust on behalf of Apple or Sony (the "E-Mail Notice"). The E-mail Notice advises that the States' and Class Actions "claim that there was a conspiracy involving five U.S. publishers and Apple to fix and raise retail prices of E-books, " that plaintiffs "will have to prove their claims in Court, " and that "Apples denies the claims and the requested damages." Rust is to send the same notice by postcard to class members for whom no correct e-mail address was found in the two prior rounds of notice. An eight-page, detailed notice (the "Detailed Notice") is available on a dedicated case website, and will be mailed to anyone who requests it by calling a toll-free number, writing to an Apple E-Books Antitrust Litigation Post Office Box, or writing or e-mailing class counsel.
Plaintiffs and Apple met and conferred regarding notice, and largely agreed upon the notices to be sent. The parties submitted minor disputes regarding the notices' language to the Court, which were addressed during a telephone conference on March 31. After a second round of discussions, the parties submitted a revised notice on April 1. The Court approved the form of notice by Order of April 1, and approved the notice plan as a whole by Order of April 2. The automated process for disseminating notice may not be halted after April 28.
III. The Instant Motion Practice
As noted above, class certification was granted on March 28 in the Class Action. On April 4, Apple brought the instant motion for a stay of the Class and States' Actions pending Apple's submission and the Court of Appeals's review of Apple's Rule 23(f) Petition, as well as for a stay pending Apple's appeal from the Opinion finding Apple liable. Apple also moved for an administrative stay pending this Court's decision on its stay motion. The parties agreed upon a briefing schedule for the stay motion, which was fully submitted on April 15. On April 11, Apple filed its Rule 23(f) Petition.
By letter of April 22, Apple requested a ruling on its stay motion by close of business the following day, or, in the alternative, a grant of Apple's request for an administrative stay. By letters of April 22 and 23, class plaintiffs and the States, respectively, opposed any such stay. Class plaintiffs opined that any stay, including an administrative stay, of these proceedings will "almost assuredly delay the July 14 trial"; the States concurred that, in the event any stay were granted, "the feasibility of a July 14 trial is significantly decreased." By Order of April 23, the Court denied Apple's stay motion and its request for an administrative stay, with reasons to follow in this Opinion.
The standard for evaluating a stay application is well established:
(1) whether the stay applicant has made a strong showing that he is likely to succeed on the merits; (2) whether the applicant will be irreparably injured absent a stay;
(3) whether issuance of the stay will substantially injure the other parties interested in the proceeding; and (4) where the public interest lies.
S.E.C. v. Citigroup Global Markets Inc. , 673 F.3d 158, 162 (2d Cir. 2012) (per curiam) (citation omitted). These factors operate as a "sliding scale" where "[t]he necessary level' or degree' of possibility of success will vary according to the court's assessment of the other stay factors... [and] [t]he probability of success that must be demonstrated is inversely proportional to the amount of irreparable injury plaintiff will suffer absent the stay." Thapa v. Gonzales , 460 F.3d 323, 334 (2d Cir. 2006) (citation omitted). A stay is an "intrusion into the ordinary processes of administration and judicial review, and accordingly is not a matter of right." Nken v. Holder , 556 U.S. 418, 427 (2009) (citation omitted); see also Maldonado-Padilla v. Holder , 651 F.3d 325, 327-28 (2d Cir. 2011) (quoting Nken , 556 U.S. at 427).
I. Apple's Request for a Stay Pending Review of its Rule 23(f) Petition
Here, for the reasons set forth below, Apple has made no persuasive showing of harm, Apple's challenge to the class certification decision is unlikely to succeed, and any stay would injure plaintiffs and the public interest. Accordingly, Apple's request for a stay pending review of Apple's Rule 23(f) Petition was denied by Order of April 23.
A. Irreparable Injury
To demonstrate ongoing "irreparable harm" such that a stay is proper, a party must show that it will suffer injury which "cannot be remedied" absent a stay. Grand River Enter. Six Nations, Ltd. v. Pryor , 481 F.3d 60, 66 (2d Cir. 2007) (per curiam) (citation omitted). The party seeking the stay has the burden of showing "injury that is not remote or speculative but actual and imminent, and for which a monetary award cannot be adequate compensation." Dexter 345 Inc. v. Cuomo , 663 F.3d 59, 63 (2d Cir. 2011) (citation omitted). Apple has not met this burden.
Apple chiefly argues that class notice will harm Apple's reputation, but also briefly refers to harm due to the cost of a corrected class notice, confusion engendered by a corrected notice, and invasion of class members' privacy. These arguments are addressed in turn.
1. Harm to Apple's Reputation
Apple contends that class notice "risks damaging the goodwill and reputation that Apple has spent many years creating." Apple has not, however, pointed to any particular harm here that distinguishes this notice from the many other notices of pendency of a class action that are routinely issued without interlocutory appellate review of the certification decision. Moreover, the timing of this notice is largely due to Apple's request to postpone class certification until after the liability trial. As significantly, Apple ignores the fact that this will be the third notice to class members advising them of the pendency of this lawsuit. Apple argues that class notice "is a bell that cannot be unrung, " but this bell has already been rung - twice.
The present E-mail Notice states that "[t]he lawsuits claim that there was a conspiracy involving five U.S. publishers and Apple to fix and raise retail prices of E-books, " notes that the Publisher Defendants have settled, and states that plaintiffs "will have to prove their claims in Court against Apple" as "Apple denies the claims and the requested damages." In September 2012, the first e-mail notice was sent to the very same class members advising that certain Publisher Defendants had "resolve[d] an antitrust lawsuit about the price of electronic books" while "[a] separate lawsuit against two other publishers and Apple, Inc. continues and is set for trial in 2013." Less than nine months ago, a second e-mail notice was sent advising of further settlements with Publisher Defendants "in antitrust lawsuits about the price of electronic books, " although "[t]he antitrust lawsuit against Apple, Inc. continues." Apple has not established that it will suffer any harm from a third notice advising class members of the pendency of these actions.
Instead, Apple argues that this notice is different because it is "the first class notice to consumers stating that Apple violated the law in connection with its entry into the e-books market." In fact, the E-mail Notice to be sent to class members says nothing about the Court's finding last July that Apple violated the antitrust laws. The liability finding is mentioned only in the Detailed Notice, which is to be provided only to class members who expressly request it.
In the Detailed Notice, in the sixth of eleven bullet points, following a bullet point stating that "Apples denies the claims and the alleged damages, " the Detailed Notice advises:
A trial in the AG Lawsuit against Apple was conducted in June 2013, and the Court found Apple liable for violating the antitrust laws. Damages were not determined in this first trial. Apple is appealing the Court's finding of liability, and also denies the alleged damages.
This reports only the bare fact that Apple was found liable, and advises that Apple is appealing that finding. This was widely reported news just nine months ago, and has continued to make news as the litigation develops. And it bears emphasis that this Detailed Notice will only be sent to class members who expressly request it. Apple has offered little reason to believe that making available a notice that includes a single sentence referencing this finding of liability will harm Apple.
Apple's only response is that "there is a significant difference between a newspaper article, reporting on the Court's prior findings of conspiracy, and an official class notice sent directly to Apple's consumers... bearing the imprimatur of a federal court." Again, it is the Detailed Notice, not the Email Notice, that references the liability finding, and the Detailed Notice is sent only to class members who request it. In any case, any interested class member has already been notified of this litigation, twice, and may well already know about the liability finding; if not, an interested class member could find the same information in a manner of minutes on the internet, not to mention a link to the liability opinion.
Notably, while Apple now argues that this language will cause it irreparable harm, Apple did not object to this language earlier this month when the parties proposed notice to the Court. Although Apple objected to a bullet point just below this one, which the Court struck in its Order of April 1, it raised no concerns about this reference to the liability finding. This quiescence does not accord with Apple's charge, weeks later, that this language will "undoubtedly and irreparably harm Apple's business."
Apple also suggests that it will be harmed because the Detailed Notice "explains that plaintiffs are seeking $280 million' in damages attributable to Apple's conduct." Again, Apple did not object to the statement in the proposed Detailed Notice that "[t]he two lawsuits are seeking $280 million combined." Apple suggested that the sentence following this one be stricken, and proposed adding a sentence just before this one ("The Plaintiffs allege violations of Section 1 of the Sherman Act, 15 U.S.C. § 1."), but raised no issue as to the "$280 million." And the two prior detailed notices already noted that damages estimates in claims against the Publisher Defendants totaled more than $200 million.
Although Apple contends that "many courts have found that potential injury to a defendant's reputation warrants curtailing or delaying class notice, " Apple cites but a single case, in which "both parties agree[d] that some form of stay [wa]s appropriate." Altamura v. L'Oreal, USA, Inc., CV 11-5465 (CAS), 2013 WL 4537175, at *2 (C.D. Cal. Aug. 26, 2013). And Apple does not dispute that, regardless of the success of its Rule 23(f) Petition, these notices will nonetheless be sent to affected customers in the thirty-three plaintiff states and territories in the States' Action pursuant to the Clayton Act. For these reasons, Apple has failed to establish that it will suffer irreparable harm to its reputation absent a stay.
2. Cost of Corrected Notice and Possibility for Confusion
Apple also argues that, in the event the Court of Appeals reverses class certification, the parties would be burdened with the need to issue a corrective notice, which may confuse class members. Because, as described below, Apple has failed to establish that its Rule 23(f) Petition is likely to win review by the Court of Appeals, let alone lead to decertification of the class, these costs are but a remote possibility.
3. Class Members' Privacy
Apple also argues that "the process of identifying class members and disseminating class notices inevitably burdens customers and infringes on their privacy interests." Yet class members are not being "identif[ied]" - the e-tailers have already compiled lists of affected consumers, and already sent them at least two e-mail notices, in addition to subsequent e-mails advising that the Publisher Defendants' settlements were approved and that class members' accounts have been credited. And unless the class is decertified, notice must be given. Again, as described below, Apple has not established that there is a "substantial possibility" of decertification. Accordingly, Apple has failed to show that it will suffer irreparable harm absent a stay.
B. Likelihood of Success on the Merits
The next factor, a strong showing of a likelihood of success on the merits, requires "more than a mere possibility of relief." Nken , 556 U.S. at 434 (citation omitted). To demonstrate a "strong showing that it is likely to succeed on the merits, " Apple has the burden of demonstrating "a substantial possibility, although less than a likelihood, of success" on appeal. Mohammed v. Reno , 309 F.3d 95, 101 (2d Cir. 2002) (citation omitted). Here, success on appeal requires both that Apple's Rule 23(f) Petition for interlocutory appeal is granted and that, upon review, the Court of Appeals decertifies the class. Apple has not established a substantial possibility of either.
1. Apple's Petition for Interlocutory Appeal
Pursuant to Rule 23(f), Fed. R. Civ. P., "[a] court of appeals may permit an appeal from an order granting or denying class-action certification" upon a timely petition. The Second Circuit will only grant leave to appeal where a petitioner demonstrates either "(1) that the certification order will effectively terminate the litigation and there has been a substantial showing that the district court's decision is questionable, or (2) that the certification order implicates a legal question about which there is a compelling need for immediate resolution." Hevesi v. Citigroup Inc. , 366 F.3d 70, 76 (2d Cir. 2004) (quoting Sumitomo Copper Litig. v. Credit Lyonnais Rouse, Ltd. , 262 F.3d 134, 139 (2d Cir. 2001)).
The first category of cases comprises "the so-called death knell' cases" where class certification "forces the defendants to settle." Sumitomo , 262 F.3d at 138. The second category are cases in which certification "implicates an unresolved legal issue"; "the more fundamental the [legal] question and the greater the likelihood that it will escape effective disposition at the end of the case, " the more likely the Court of Appeals is to permit an interlocutory appeal. Id . (citation omitted). The Court of Appeals has emphasized that "the standards of Rule 23(f) will rarely be met." Id. at 140.
Apple does not even address the Rule 23(f) standard in its initial memorandum in support of its motion. In response to plaintiffs' opposition memoranda, Apple quotes the standard in its reply, but simply states, without argument, that "both grounds for review... are satisfied here." This conclusory statement does not satisfy Apple's burden of establishing a substantial possibility of success on appeal.
Even if Apple had engaged with the Rule 23(f) standard, Apple's Petition does not appear to meet either ground for interlocutory appeal. Here, the class certification order cannot "effectively terminate the litigation, " as the States' Action - which accounts for $155 million of the $280 million of alleged damages - will proceed to trial regardless of class certification. Nor does class certification threaten "ruinous liability" for Apple, as the $106 million in damages alleged by the class, even trebled, will not seriously threaten a company recently reported by Moody's Investors Service to have more than $150 billion cash-on-hand. Cf. Chamberlan v. Ford Motor Co. , 402 F.3d 952, 960 (9th Cir. 2005) (noting, in "death knell" analysis, that "the potential recovery here may be unpleasant to a behemoth company, but it is hardly terminal" to defendant Ford Motor Co.) (citation omitted); Waste Mgmt. Holdings, Inc. v. Mowbray , 208 F.3d 288, 294 (1st Cir. 2000) ("[W]hat might be ruinous' [liability] to a company of modest size might be merely unpleasant to a behemoth.").
And there is no reason to believe that the certification order presents "a legal question about which there is a compelling need for immediate resolution." Hevesi , 366 F.3d at 76 (citation omitted). The Class Certification Opinion is based on well settled law applied to a conspiracy to set national ebook prices that, according to plaintiffs' expert's rigorous multivariate regression analysis, inflated e-book prices for 99.8% of sales of the Publisher Defendants' e-book during the class period. See Class Certification Opinion, at *11-25. Thus, Apple has failed to show a substantial possibility that its Petition will be granted.
2. Likelihood of Decertification
Apple has also failed to establish a substantial possibility that the Court of Appeals will decertify the class. Apple raises four arguments in support of decertification: (a) the class certified "contain[s] members who have suffered no harm" and therefore lack standing; (b) Noll's model cannot establish class-wide harm, because it is based on "average" overcharges and ignores offsets; (c) the Court erred in "resolv[ing] key merits questions" in its certification order, and consequently "abdicat[ed] its duty to rigorously analyze' plaintiffs' proof"; and (d) Noll's model does not match plaintiffs' theory of liability. These arguments are addressed in turn.
a) Article III Standing
First, Apple argues that the Court impermissibly accepted the possibility that the class might include persons who have not suffered injury from Apple's conduct. Citing to Denney v. Deutsche Bank AG , 443 F.3d 253, 263-64 (2d Cir. 2006), Apple contends that, "[t]o the contrary, the Second Circuit has expressly held that bringing a suit as a class action does not relax the standing requirement, and that no class may be certified that contains members lacking Article III standing.'" Apple misreads Denney, which fully supports certification here, mistakenly conflating standing with persuasive proof of injury.
While Article III requires an "injury in fact, " Denney itself teaches that, ordinarily, "[f]or purposes of determining standing, we must accept as true all material allegations of the complaint." Id. at 263. Where standing is challenged at the later stages of litigation, it must be "supported... with the manner and degree of evidence required at the successive stages of the litigation." Lewis v. Casey , 518 U.S. 343, 358 (1996) (quoting Lujan v. Defenders of Wildlife , 504 U.S. 555, 561 (1992)). Here, prior to trial, as a motion for summary judgment is pending, class members need do no more than establish a genuine issue of material fact exists as to the alleged injury. Id.
The class, by definition, is composed solely of consumers who purchased an e-book from a Publisher Defendant during the period of time in which those Publisher Defendants were engaged in a conspiracy with Apple to fix e-book prices. There can be no serious argument that those consumers lack Article III standing to bring a Sherman Act claim for price fixing.
Class plaintiffs have put forward evidence that in 99.8% of purchases of the Publisher Defendants' e-books during the class period, the purchaser suffered an overcharge as a result of Apple's conduct. This adequately supports each class member's standing to litigate his or her claims. Cf. Denney , 443 F.3d at 263 ("We do not require that each member of a class submit evidence of personal standing.").
Apple also argues that class certification has "depriv[ed] Apple of its right to make... individualized challenges" to claims of overcharge with respect to particular e-books. This is not so. Apple has the list of the Publisher Defendants' e-books sold to class members during the class period. Apple is free to make as many "individualized challenges" to alleged overcharges for as many titles as it would like, should it have evidence to offer that does not run afoul of the Federal Rules of Evidence.
b) Challenges to Noll's Model
Apple next charges that class certification was inappropriate because "Noll's damages model is not capable of showing class-wide anticompetitive harm.'" For the reasons set out in the Class Certification Opinion, as well as the reasons below, Apple's arguments are not persuasive.
Apple principally argues that Noll's model "does not even attempt to demonstrate injury for each class member, " as it "calculates an average overcharge for e-books within each category [Noll] created." This is not the case. In fact, there is no reason to believe that Noll's damages model would have been any different had he been asked to calculate the overcharge for a single plaintiff who purchased a half-dozen e-books. That is, the fact that Noll's model calculates the same relative overcharge for each of the e-books in a given category is not an artifact of an attempt to measure class-wide damages; one would expect the same thing to be true if Noll were only interested in that single plaintiff's half-dozen titles.
Indeed, this would be a feature of any well-constructed multivariate regression analysis. Noll's 502 categories are as fine-grained a comparison as Noll can make between the Publisher Defendants' e-books and "competitive benchmark" e-books. Noll's multivariate regression analysis uses every major quantifiable factor that might influence e-book pricing to ensure that it compares apples to apples. After accounting for all of these factors, Noll has no other way to distinguish the effect of the conspiracy on one e-book in a given category from the effect on another that falls in that same category. Where all other variables are equal, Noll's model calculates that the relative effect of collusion will be the same. Thus, Noll is not impermissibly "averaging" overcharge calculations across many e-books - Noll's method for calculating any overcharge simply depends on the objective variables that might influence pricing; when, for two e-books, those factors are identical, the relative overcharge will be the same. See Class Certification Opinion, at *30-31. It is noteworthy that neither Apple nor either of its experts identified a quantifiable variable that Noll failed to include in his regression analysis. For this reason, as well, Apple's contention that the application of Noll's model will result in a "windfall" for purchasers of e-books who "suffered no harm" misses the mark. Damages will only be awarded where individual transaction records identify the purchaser as someone who bought an e-book for which there was an overcharge.
Apple also points to evidence that 17% of the Publisher Defendants' e-book prices fell after the adoption of agency agreements as evidence that many class members were not injured. Notably, Apple omits the fact that 60% of other publishers' e-books prices fell over the same period - more than three and a half times as many. And because an e-book's price would be expected to fall for a number of independent reasons, like the release of a paperback edition, the fact that a given e-book's price fell does not prove that the price was not inflated (i.e., that the price would not have fallen more, absent Apple's conduct). See id. at *15. Apple appears to assume that any drop in the price of the Publisher Defendants' e-books during this time period was caused by Apple's conduct, but offers no evidence or analysis to support this inference.
Apple next contends that the Court erred in rejecting Apple's proposed damages offsets because "unavoidable speculation' about the but-for world, and the resulting uncertainty about any individual injury, renders certification inappropriate." This is a non sequitur. In the absence of evidence, Apple is no more able to speculate about injury to one plaintiff as to another, and thus any defense Apple may wish to make on the basis of such speculation would be applicable class-wide. Indeed, the appropriateness of offsets to any damages calculation is itself a class-wide issue.
It bears noting that in this application for a stay Apple has, for the first time, proposed a new offset to damages: it contends that undercharges enjoyed on one Publisher Defendant's e-book should be offset against overcharges suffered on another. Whether or not this offset is appropriate, it could easily be calculated on a class-wide basis and so does not inject individual issues into the trial. Noll's model, however, found no overcharges in just 0.2% of transactions.
Apple's final argument, regarding a "trial-by-formula, " is rejected for the reasons set out in the Class Certification Opinion. See id. at *22. In brief, the Supreme Court has never suggested that widely used tools of economic analysis like regression models should be banned from trials because they rely on a "formula." The Supreme Court's reference to a "Trial by Formula" in Wal-Mart Stores, Inc. v. Dukes was to a plan to try a sample set of class members' claims of sex discrimination and then multiply the average backpay award to determine the classwide recovery without further individualized proceedings. 131 S.Ct. 2541, 2561 (2011). Under this procedure, liability for all but the sample set would never have been tried. Id . Here, class plaintiffs simply seek to apply an expert economist's measure of damages to each class member's individual transaction records. Apart from the word "formula, " this method bears no relationship to the "Trial by Formula" prohibited in Dukes.
c) Resolving Key Merits Issues
Apple next argues that a class should be decertified because this Court abdicated its duty to rigorously analyze plaintiffs' proof and resolved key merits questions. Apple's chief argument on this score is that, in the first paragraph of the Class Certification Opinion, the Court made the following statement: "This is a paradigmatic antitrust class action. Virtually all class members paid inflated prices for e-books as a result of a centralized price-fixing conspiracy, and they have proffered a sophisticated damages model to reliably determine damages." Class Certification Opinion, at *1. Apple misreads the second sentence to ultimately resolve the question of damages against Apple as to "[v]irtually all class members."
As the context, as well as the remainder of the Opinion, makes clear, the Court made no such ultimate determination. Rather, the Court found that class plaintiffs had produced sufficient evidence - for purposes of class certification - that, according to Noll's rigorous damages model, class members suffered an overcharge in 99.8% of their purchases of the Publisher Defendants' e-books during the class period as a result of the centralized price-fixing conspiracy alleged by class plaintiffs. Given this fact-pattern, which was adequately established for purposes of class certification, the Court observed that this is a paradigmatic class action. This was not, as Apple charges, an attempt to "short-circuit a jury trial by resolving merits questions unrelated to certification." Whether particular class members were, in fact, injured by Apple's conduct remains a question for the jury.
Similarly, Apple complains that the Court has "accepted" Noll's model. The Court was required to rule on the class plaintiffs' motion for class certification and Apple's motion to strike Noll's expert report. To that extent, the Court has determined that Noll's model is sufficiently reliable to be admissible and to support a finding of predominance. It will be for the jury to accept or reject it as persuasive at trial.
Apple also contends, erroneously, that the Court "adopted a presumption in favor of class certification in antitrust cases." Apple's only basis for this argument is the Court's quotation from the Supreme Court's decision in Amchem Products, Inc. v. Windsor , 521 U.S. 591 (1997), noting that predominance is readily shown "in certain cases alleging... violations of the antitrust laws." Class Certification Opinion, at *13 (quoting Amchem , 521 U.S. at 625). To recognize that certain antitrust cases are prime candidates for class certification because class-wide issues so clearly predominate is not to indulge a presumption in favor of class certification in antitrust cases. The Court's Opinion recognized that the burden to show predominance rested on the plaintiffs, and required them to shoulder that burden. Id. at *11, 20, 22-23.
Apple's remaining arguments are rejected for the reasons given in the Class Certification Opinion. In particular, Apple's charge that Noll's model should be excluded because it can explain only 5% of the variation in the prices of a given e-book is misleading. Noll's model is built to estimate the effects of collusion on e-book prices; to do so, it effectively compares prices of e-books potentially affected by collusion to prices of unaffected e-books. Thus, the relevant question is whether the model can reliably explain the differences between collusive e-book prices and competitive prices; the model's incidental ability (or inability) to explain changes in a given e-book's price is largely irrelevant. See Class Certification Opinion, at *28. Noll's model is able to explain 90% of the variance in prices among e-book titles. Id. at *10.
d) Noll's Damages Model Matched Plaintiffs' Theory of Liability.
Relying on a portion of an expert report that was stricken as untimely, Apple argues that very few e-books would have been sold at $9.99 according to Noll's model. As noted in the Class Certification Opinion, it does not appear, on its face, that Apple's expert's study made any attempt to isolate the extent to which the predicted prices of New Releases and NYT Bestsellers - the e-books Amazon sold at $9.99 prior to the shift to agency - fall close to $9.99. Id. at *30 n.37. In any case, because this analysis has been stricken, Apple cannot rely on it in opposition to class certification. For all of the reasons set out above, Apple's Rule 23(f) Petition is unlikely to succeed.
C. Injury to Plaintiffs and the Public Interest
By letter of April 22, 2014, Apple advised the Court that, in order to meet current deadlines, including the joint damages trial to begin in the Class and States' Actions on July 14, the machinery of notice must be set in motion no later than April 28. A stay pending the Second Circuit's review of Apple's Petition would delay not only adjudication of damages in the Class Action, but would also delay adjudication in the States' Action - if a joint trial were maintained - or require the States to move forward with a damages trial without the Class, despite the fact that damages trials in the States' and Class Actions should be nearly identical. This would impose substantial burdens on States, class plaintiffs, and on the Court.
Delaying the trial would also delay any recovery due plaintiffs, should they prevail. Apple argues that class members have already been partially compensated by the Publisher Defendants' settlements, but class members have a strong interest in being fully compensated for any losses they have suffered. Likewise, the public interest favors a speedy trial and resolution of this matter.
Apple also argues that a stay is in class members' interests because a corrective notice would cause confusion. As Apple's 23(f) Petition is unlikely to be granted, the chance of such confusion is minimal. Accordingly, injury to plaintiffs, as well as the public interest, militates against a stay. Because Apple has not established that it will suffer irreparable harm without a stay, because it is unlikely to succeed on the merits, and because plaintiffs' interest and the public interest counsel against a stay, Apple's request for a stay pending the Second Circuit's review of Apple's Rule 23(f) Petition is denied.
Apple leans heavily on language taken from the Manual for Complex Litigation, urging that "the district court should ordinarily stay the dissemination of class notice to avoid the confusion and the substantial expense of renotification that may result from appellate reversal or modification." Manual for Complex Litigation (4th) § 21.28. But the Manual's sole support for this proposition is a single cite to Ramirez v. DeCoster , 203 F.R.D. 30 (D. Me. 2001), an aberrational case in which - unbeknownst to the court - the parties entered a settlement agreement as to all claims just before the court entered a ruling denying class certification on a number of claims. The court then considered certification of a settlement class, affirmatively "urg[ing] the Court of Appeals to accept an appeal under Rule 23(f), " making repeated references to possible reversal by the Court of Appeals and even considering, at length, the ramifications of such a reversal. Id. at 40. In that context, the court entered a stay to be triggered by filing of a Rule 23(f) petition. The unusual circumstances in Ramirez find no parallel here.
Moreover, earlier in the same paragraph quoted by Apple, the Manual for Complex Litigation notes that "[i]nterlocutory appeals can disrupt and delay the litigation without necessarily changing the outcome of what are often familiar and almost routine issues" and states that "[g]ranting a stay depends... on a demonstration that the probability of error in the class certification decision is high enough that the costs of pressing ahead in the district court exceed the cost of waiting." Apple points to no authority suggesting that a court is empowered to enter a stay where the four-factor standard is not met. Because it is not met here, Apple's stay must be denied.
II. Apple's Request for a Stay Pending the Merits Appeal
Apple principally argues for a stay pending the Court of Appeals's review of Apple's Rule 23(f) Petition for interlocutory appeal of the grant of class certification. But, Apple also makes occasional references to its pending merits appeal of this Court's judgment on liability in the DOJ Action and the States' Action and suggests that a stay should be granted pending that appeal. This Court already denied Apple's request for a stay pending the merits appeal at a conference with the parties on August 9, 2013 for the reasons stated on the record. Apple elected not to move for reconsideration of that decision, not to request a stay of all proceedings from the Court of Appeals, and not to move for an expedited appeal from the trial decision or the injunction entered against Apple. In its briefing on the present motion, Apple largely ignores the August request for a stay pending the merits appeal - Apple does not even address whether it is likely to succeed on the merits of that appeal - but occasionally adverts to it in an attempt to bolster its arguments for a stay pending review of its Rule 23(f) Petition.
To the extent Apple now moves the Court for a stay pending the merits appeal, that portion of its motion is construed as a motion for reconsideration. The motion is untimely and without merit. "A motion for reconsideration should be granted only when the defendant identifies an intervening change of controlling law, the availability of new evidence, or the need to correct a clear error or prevent manifest injustice." Kolel Beth Yechiel Mechil of Tartikov, Inc. v. YLL Irrevocable Trust , 729 F.3d 99, 104 (2d Cir. 2013) (citation omitted). Apple has not attempted to make such a showing. Accordingly, Apple's motion for reconsideration of this Court's August 9, 2013 denial of its request for a stay pending the merits appeal is denied as well. Even if this were Apple's first motion for stay pending the merits appeal, Apple's motion would be denied on several grounds - not least of which is that Apple does not even attempt to show a likelihood of success on the merits in its briefing in support of this motion.
For the reasons stated above, Apple's April 4 motion for stay pending appeal, and its request for an administrative stay pending decision on that motion, were denied by Order of April 23.