Summary
rejecting defendants' attack on econometric regression model based on plaintiffs' expert's "selection of particular data to utilize, the variables he has elected to include in his regressions, and similar details relating to the implementation of his econometric models," and noting that those challenges may be "considered by a fact-finder evaluating the persuasiveness of that expert's conclusions but do not provide a "basis for wholly rejecting those conclusions at this stage in the proceedings as methodologically unsound."
Summary of this case from In re Korean Ramen Antitrust Litig.Opinion
CASE NO. 3:10-md-2143 RS
02-08-2016
IN RE OPTICAL DISK DRIVE ANTITRUST LITIGATION This Document Relates to: ALL ACTIONS
ORDER GRANTING INDIRECT PURCHASERS' RENEWED MOTION FOR CLASS CERTIFICATION
I. INTRODUCTION
In 2009, the Department of Justice disclosed the existence of an ongoing criminal investigation of possible antitrust violations within the optical disc drive ("ODD") industry. That investigation ultimately resulted in guilty pleas by Hitachi-LG Data Storage, Inc. and some of its employees. Substantial fines were imposed and the individuals were sentenced to prison terms. This Multi-District Litigation consolidates the many civil actions that were filed against ODD manufacturers in the wake of the DOJ investigation.
From the outset of this matter, defendants have conceded, that as the guilty pleas reflect, there were at least some instances of illegal anticompetitive conduct in connection with the sale of ODDs. Specifically, defendants do not deny there were multiple incidents involving "bid rigging" during procurements of ODDs by Dell, HP, and Microsoft. The controversy therefore has centered on whether the wrongdoing was limited to those events and defendants, or whether, as plaintiffs contend, the bid rigging was merely one part of a vast industry-wide price-fixing conspiracy, also involving inter-competitor agreements, and exchanges of price, output, and other types of confidential information.
Previously, two groups of plaintiffs, so-called "direct purchasers" ("DPPs") and "indirect purchasers," ("IPPs") who are separately represented and proceeding under separate complaints, sought class certification. The crux of the disputes in the two certification motions was whether plaintiffs' experts had presented a viable methodology for establishing class-wide antitrust injury and damages. Because neither group of plaintiffs had made a persuasive showing that the expert analyses they proffered sufficiently address the relevant question, the motions were both denied. Subsequent to that ruling and to the Ninth Circuit denying leave to appeal, the DPPs entered into settlement agreements disposing of their remaining claims.
The IPPs also reached settlement agreements with certain defendants—court approval of those settlements is pending. As to their remaining claims, however, the IPPs have renewed their request for class certification. The IPPs support their renewed motion with further analysis from Dr. Kenneth Flamm, their economic expert, consisting of both modified and additional approaches to the calculations he performed, and consideration of more data. Defendants insist the additional material proffered is not substantively different from that previously found to be inadequate as a means for establishing class-wide antitrust injury and damages. Defendants also continue to argue that certain alleged errors in Dr. Flamm's methodology render his conclusions unreliable.
As plaintiffs state, "there is a meaningful line drawn at class certification between methods and merits." That line, dividing what is an appropriate inquiry into the soundness of plaintiffs' proffered statistical models on the one hand, and an improper adjudication of merits issues on the other hand, remains challenging to draw. Here, some of the issues defendants have raised may ultimately be persuasive to a trier of fact that plaintiffs cannot show antitrust injury or establish damages across the class. The methodologies and theories plaintiffs proffer, however, will stand or fall on a class-wide basis and are not so lacking in substance as to permit rejection at the certification stage. Accordingly, with consideration of the choice of law issue discussed below, the motion for certification will be granted, as to 23 states and the District of Columbia, under California law.
The various motions to file material under seal in connection with the renewed motion for class certification reflect appropriate efforts to limit the material withheld from the public record, and are therefore granted, without prejudice to any subsequent motion to unseal, should good cause exist. The parties are requested to submit jointly a single proposed order specifying the materials that will remain sealed and identifying by docket number all of the sealing motions to which it pertains.
II. BACKGROUND
The factual background and long procedural history of this action have been set out in prior orders and will not be recounted in full detail here. Briefly, an optical disc is a medium for storing data. Familiar forms include CDs (compact discs) typically containing music or computer software, and DVDs (digital video disc or, officially, Digital Versatile Disc), often containing movies or other video content, and also used for computer software. Depending on the stage of technological development, optical discs were "read-only" or "recordable" or "rewritable." In a broad sense, the technology evolved generationally from CDs to DVDs to Blu-Ray Discs, each with a progression from read-only, to recordable, and then to rewritable.
Although the caption of this action is "In re Optical Disk Drive Litigation," the preferred spelling appears to be "disc," for optical media, as opposed to "disk," for magnetic media.
Optical disc drives—ODDs—are devices that allow data to be read from and, where applicable, written to, optical discs. ODDs are typically "backwards-compatible," - that is, an ODD that is designed to read (and perhaps write to) a more recently-developed format of optical disc usually will also be able to process older formats as well. ODDs have applications in a variety of consumer electronic devices, including desktop and laptop computers, game consoles, and camcorders. In these applications, the ODD is typically a built-in component of the device. ODDs are also available as stand-alone units, in a number of forms. Where ODDs are incorporated into other products such as computers, they typically represent a relatively small percentage of the cost of the product as a whole. Plaintiffs have argued, however, that an ODD can be one of the more significant individual cost components in a computer, even if only a fraction of the total.
During the putative class period, the prices of ODDs were generally marked by steep declines. While each generational advance in the technology was often introduced at a higher price, the overall trend was downward, and dramatically so, as has generally been the case in the high technology arena. Plaintiffs' basic theory in this action is that defendants were highly motivated to attempt to slow, or at least stabilize, the inevitable decline in prices. In short, plaintiffs are not arguing that the alleged conspiracy drove prices upward, merely that it kept prices from falling as rapidly and/or as far as they otherwise would have.
At the center of this action are multiple instances of alleged "bid rigging" involving procurements of ODDs by Dell, HP, and Microsoft. In connection with the now-closed DOJ investigation into that conduct, defendant Hitachi-LG Data Storage, Inc. pleaded guilty to criminal antitrust violations and paid a $21.1 million criminal fine. As alluded to above, plaintiffs contend that lengthy and voluminous discovery has revealed evidence the bid rigging was merely one part of a vast industry-wide price-fixing conspiracy, involving a broad range of improper anti-competitive behavior.
Although much of the evidence to which plaintiffs point involves the bidding events, they contend continuous illegal information exchanges occurred among all defendants relating to customer accounts other than HP and Dell. Plaintiffs point to certain "alliances" among defendants that allegedly allocated customers and markets. They contend "supply arrangements" existed among certain defendants that nominally were competitors in the market. Plaintiffs aver that defendants reached oral agreements at various meetings in Asia and in the United States at numerous points in time. While plaintiffs argue that cumulatively the evidence is indicative of a "pervasive" industry-wide conspiracy, at least at this juncture they have not proffered evidence or allegations of any instances in which the defendants' executive decision-makers entered into express agreements to fix prices across the board on an ongoing basis.
III. LEGAL STANDARD
Class actions are governed by Rule 23 of the Federal Rules of Civil Procedure, which represents much more than a mere pleading standard. To obtain class certification, plaintiffs bear the burden of showing that they have met each of the four requirements of Rule 23(a) and at least one subsection of Rule 23(b). Zinser v. Accufix Research Inst., Inc., 253 F.3d 1180, 1186, amended by 273 F.3d 1266 (9th Cir. 2001). "A party seeking class certification must affirmatively demonstrate . . . compliance with the Rule." Wal-Mart Stores, Inc. v. Dukes, 131 S.Ct. 2541, 2551 (2011). Rule 23(a) provides that a district court may certify a class only if: "(1) the class is so numerous that joinder of all members is impracticable; (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." Fed. R. Civ. P. 23(a). That is, the class must satisfy the requirements of numerosity, commonality, typicality, and adequacy of representation to maintain a class action. Mazza v. Am. Honda Motor Co., Inc., 666 F.3d 581, 588 (9th Cir. 2012).
If all four prerequisites of Rule 23(a) are satisfied, a court must also find that plaintiffs "satisfy through evidentiary proof at least one of the three subsections of Rule 23(b). Comcast Corp. v. Behrend, 133 S.Ct. 1426 (2013). Relevant here is Rule 23(b)(3), which permits certification if a court finds that "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).
"[A] court's class-certification analysis must be 'rigorous' and may 'entail some overlap with the merits of the plaintiff's underlying claim.' " Amgen Inc. v. Conn. Ret. Plans and Trust Funds, 133 S.Ct. 1184, 1194 (2013) (quoting Dukes, 131 S.Ct. at 2551); see also Mazza, 666 F.3d at 588 ("'Before certifying a class, the trial court must conduct a 'rigorous analysis' to determine whether the party seeking certification has met the prerequisites of Rule 23.'" (quoting Zinser, 253 F.3d at 1186)). This "rigorous" analysis applies to both Rule 23(a) and Rule 23(b). See Comcast, 133 S.Ct. at 1432 (discussing how Congress included "addition[al] . . . procedural safeguards for (b)(3) class members beyond those provided for (b)(1) or (b)(2) class members (e.g., an opportunity to opt out)" and how a court has a "duty to take a 'close look' at whether common questions predominate over individual ones.").
Nevertheless, "Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage." Amgen, 133 S.Ct. at 1194-95. "Merits questions may be considered to the extent—but only to the extent—that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied." Id. at 1195. If a court concludes that the moving party has met its burden of proof, then the court has broad discretion to certify the class. Zinser, 253 F.3d at 1186.
IV. DISCUSSION
A. Class definition
The IPPs' renewed motion proposes a slightly narrowed class definition:
All persons and entities who, as residents of [the United States or State] and during the period April 2003 to December 2008, purchased new for their own use and not for resale: (i) a computer with an internal ODD; (ii) a stand-alone ODD designed for internal use in computers; or (iii) an ODD designed to be attached externally to a computer. ODD refers to a DVD-RW, DVD-ROM, or COMBO drive manufactured by one or more Defendants or their coconspirators. Excluded from the class are any purchases of Panasonic-branded computers.
This narrower language removes CD and Blu-ray ODDs from the scope of products involved in the litigation. Additionally, the products are limited to computers containing ODDs and stand-alone ODDs only, removing videogame consoles such as the Xbox. Panasonic-branded computers are also now excluded. Finally, although IPPs continue to claim the conspiracy did not end until sometime in 2009, they limit the damages period to extend only through December 2008. The IPPs contend, and defendants do not particularly dispute, that the effect of these changes is to remove significant volumes of products to which defendants pointed in the prior motion as reflecting heterogeneity sufficient to defeat class certification.
Plaintiffs have also proposed a "subclass," consisting of purchasers of products from Dell and HP, the two entities most directly affected by the alleged bid-rigging. IPPs assert that because Flamm's economic models now provide separate estimates for the alleged overcharge on Dell and HP products, certifying a subclass will give the trier of fact the option to conclude the object and impact of the conspiracy was limited to ODDs sold to and incorporated in Dell and HP computers. Whether the IPPs advance the possibility of certifying a subclass as a hedge against outright denial of their motion, or rather out of a genuine concern that such a subclass needs separate definition, remains unclear. A fundamental premise of the IPPs' motion is that prices in the ODD industry during the class period were linked, and, moreover, that prices paid by Dell and HP effectively set a "floor," given their combined share of the market and purchasing power. At this juncture, at least, the IPPs have not demonstrated that defining a subclass is warranted, and that aspect of the motion will be denied, without prejudice. If, in the course of trial preparation or trial itself, it becomes apparent that defining a subclass of Dell and HP would have salutary purposes, the issue may be revisited.
B. Predominance
The IPPs seek class certification under subsection (3) of Rule 23(b), which requires them to show (1) "questions of law or fact common to the members of the class predominate" and (2) "a class action is superior to other available methods for the fair and efficient adjudication of the controversy." As set out in the prior order, to obtain class certification, the IPPs must satisfy the predominance requirement with respect to all the key elements of their claims: (1) whether there was a conspiracy to fix prices in violation of the antitrust laws; (2) the fact of plaintiffs' antitrust injury, or "impact" of defendants' unlawful activity; and (3) the amount of damages sustained as a result of the antitrust violations. In re Dynamic Random Access Memory (DRAM) Antitrust Litig., 2006 WL 1530166, *7 (N.D. Cal. 2006); In re Vitamins Antitrust Litig., 209 F.R.D. at 257.
The prior order explained why the existence or non-existence of the alleged conspiracy presented an appropriate question for resolution on a class-wide basis, and the parties have not revisited that issue in the present motion. While the parties briefly discuss the question of damages, there is no basis to conclude that damages issues would be a basis for denying certification if otherwise warranted. A plaintiff's damages "[c]alculations need not be exact, but . . . 'must be consistent with its liability case.'" Comcast Corp. v. Behrend, ___U.S.___, 133 S. Ct. 1426, 1433 (2013) (internal citation omitted). Although defendants have offered reasons they contend that Dr. Flamm's damages models are flawed, those arguments may appropriately be considered by the trier of fact, and do not rise to a level that would require denial of certification. Accordingly, the critical issue is antitrust injury or impact.
1. Class-wide injury
The prior order denying the motions for certification brought by both the DPPs and the IPPs centered on the question of whether plaintiffs had presented a plausible methodology for establishing class-wide injury. See In re Dynamic Random Access Memory (DRAM) Antitrust Litig., 2006 WL 1530166, *7 (N.D. Cal. 2006) ("plaintiffs must establish the predominance of common issues . . . [as to] the fact of plaintiff's antitrust injury, or "impact" of defendants' unlawful activity." That order explained the question arises under subsection (3) of Rule 23(b), which, as noted above, places a burden on plaintiffs to show, among other things, that "questions of law or fact common to the members of the class predominate." Predominance requires "that the common issues be both numerically and qualitatively substantial in relation to the issues peculiar to individual class members." In re Bulk [Extruded] Graphite Prod. Antitrust Litig., 2006 WL 891362 at *9. Generally speaking, the test for predominance is met "when there exists generalized evidence which proves or disproves an [issue or element] on a simultaneous, class-wide basis, since such proof obviates the need to examine each class member's individual position." In re Vitamins Antitrust Litig, 209 F.R.D. 251, 262 (D.C. Cir. 2002).
Defendants' booklet of slides submitted at the hearing as an aid to oral argument repeatedly invokes "law of the case" when describing various findings set out in the prior order. "Under the 'law of the case' doctrine, a court is generally precluded from reconsidering an issue that has already been decided by the same court, or a higher court in the identical case . . . . The doctrine is not a limitation on a tribunal's power, but rather a guide to discretion . . . . A court may have discretion to depart from the law of the case where: 1) the first decision was clearly erroneous; 2) an intervening change in the law has occurred; 3) the evidence on remand is substantially different; 4) other changed circumstances exist; or 5) a manifest injustice would otherwise result." United States v. Alexander, 106 F.3d 874, 876 (9th Cir. 1997) (citations and internal quotation marks omitted). While certain statements quoted in the booklet represent basic legal principles that remain valid, resort to "law of the case" is not necessary to make that so. Other statements offered as "law of the case" merely describe the record presented in the prior certification motion and the legal conclusions reached on that record. The issue now is what legal conclusions are supported by the record presented in this motion. Invocation of "law of the case" is therefore inapt in this context.
As in their prior motion for certification, the IPPs proffer evidence and expert opinion that various conditions in the ODD industry were conducive to the formation of an antitrust conspiracy, including barriers to entry, high concentration among sellers, substantial cross-ownership, including joint ventures, and an allegedly commoditized product. The IPPs also argue that the DOJ investigation and resulting penalties, an enforcement action in Taiwan, and rumors of penalties to be issued by the European Union, all further support an inference of a conspiracy, with class-wide impact. The real meat of plaintiffs' motion, however, lies in Dr. Flamm's presentation of econometric analysis offered to show, in essence, that prices in the ODD market during the class period were linked, such that all direct purchasers would likely have been impacted by a conspiratorial overcharge.
Showing impact to direct purchasers is, of course, only the first step in the IPPs' burden. As discussed below, they must also show "pass-through" of the overcharges to the IPP class.
In addressing the expert testimony offered to show class-wide injury to direct purchasers, the prior order first focused on the analysis provided by Dr. Gary French, who had been retained by the DPPs. In addition to pointing to certain characteristics of the industry, Dr. French offered (1) an empirical correlation analysis and (2) a regression analysis which he contended could serve as reliable, economically sound, methodologies to demonstrate that all, or nearly all, members of the class suffered damage as a result of defendants' alleged anti-competitive conduct. In finding the correlation analysis lacking, the prior order pointed to Dr. French's admission that "[w]ith or without a conspiracy I would expect to see high correlation of prices across customers in this industry." The order further concluded that Dr. French's regression analysis, which calculated a single overcharge percentage for all purchasers across all models of ODDs and throughout the entire class period, effectively assumed the proposition it was intended to prove—i.e. class-wide impact.
In the IPPs' renewed motion, Dr. Flamm stops short of explicitly endorsing the prior order's rejection of Dr. French's model, but he makes clear his view that his own analysis previously included elements that were lacking in Dr. French's approach, and which he contends address the issues raised by the prior order, at least in part. Indeed, the prior order expressly noted that Dr. Flamm's analysis was "more complex" than that of Dr. French. Significantly, the prior order noted, Dr. Flamm included a "cointegration" analysis designed to filter out "spurious" correlations that would arise even in the absence of a conspiracy, and his model allowed for more than one estimated overcharge coefficient. Nevertheless, the prior order concluded the IPPs had failed to make a persuasive showing that Dr. Flamm's somewhat more complex analysis could serve to establish class-wide impact, rather than assuming it.
While Dr. Flamm does not endorse that view of his prior analysis, for the present motion he has done additional work designed to eliminate any doubt that his methodologies are reliably designed to measure class-wide impact (if any), rather than to assume it. First, Dr. Flamm now offers a "more extensive" cointegration analysis, which he contends controls for declining ODD prices in the market. The "refined" analysis now includes data from a greater number of defendants—Dr. Flamm contends it incorporates all the "sufficiently useable data" produced in discovery. Dr. Flamm asserts the analysis empirically tests for (and supports) the existence of market-wide impact by studying the prices for ODDs sold to different customer segments.
As Dr. Flamm explains it, a "cointegration test is a test for structural economic cohesion among variables with time trends." Dr. Flamm has tested for economic cohesion between ODD products, prices, producers, and customers. In his view, if there is cointegration, there is a strong likelihood of market-wide impact from the conspiracy. Acknowledging that a correlation analysis does not necessarily imply causation, he opines that cointegration does. From his cointegration analysis Dr. Flamm concludes that the relationships among ODD prices in the class period "are NOT a spurious artifact of these prices simply declining over time." (Emphasis in original.) Rather, in Dr. Flamm's opinion, the co-integration revealed by his analysis supports a conclusion that "the economic forces of substitution in supply and demand link prices for different drives to different customers together in the market."
Second, Dr. Flamm has performed a new "Granger causality" analysis. That analysis asks the question, "does the history of a series x assist in forecasting a series y when one also knows the history of y?" See In re Amaranth, 269 F.R.D. at 384. Here, Dr. Flamm's analysis concludes that after controlling for the entire past history of prices and costs for any single type of drive, the past history of drive prices for Dell or HP has statistically significant value in predicting the prices for other customers. Thus, Dr. Flamm opines, changes in ODD prices in sales to HP and Dell caused price movements among drives sold to other ODD purchasers.
"Granger causality" analysis refers to the Engle-Granger regression form created by Robert F. Engle and Clive W.J. Granger—who shared the Nobel Prize for Economics in 2003 on the basis of their development of this methodology. See In re Amaranth Nat. Gas Commodities Litig., 269 F.R.D. 366, 384 n. 122 (S.D.N.Y. 2010).
Third, Dr. Flamm has modified his overcharge regression model in four respects:
(a) The overcharge model now is expressly delineated to estimate separate overcharge coefficients first for Dell and HP on the one hand, and then for "other" customers. As a result, Dr. Flamm contends, the overcharge model now expressly is capable of testing whether the cartel impacted only its two largest customers (Dell and HP) or also the other direct purchasers.
(b) The overcharge regression model is also capable of measuring the overcharge on a monthly basis, by each different drive in the class (DVD-RW, DVD-ROM, and COMBO) for each of these customer groups.
(c) The model now integrates all "useable" sales and cost data produced - from 86 percent of the market (all but two defendants).
(d) IPPs provide further detail on the multivariable regression analysis, which they contend shows that all factors other than conspiracy are being adequately controlled for in the overcharge model.
Finally, the IPPs present additional graphic evidence and argument to support their claim that prices paid by Dell and HP effectively served as a "floor," and that any artificial maintenance of the level of that floor would necessarily impact prices paid by other customers. The IPPs take issue with a characterization in the prior order that the prices charged to other customers did not "cluster within an especially narrow range above the supposed 'floor' of the prices paid by Dell and HP." The relevant question, the IPPs insist, is not the existence or non-existence of any such clustering, but whether the economic evidence supports linkage between the prices. The IPPs point out that, with one exception, the prices paid by other customers were maintained at a generally consistent level above those paid by Dell and HP.
The exception involves the prices of "combo drives" in 2008. The IPPs attempt to explain that exception away by asserting that Dell and HP were simply overcharged more for these particular drives than were other customers. It is unclear whether this explanation can be supported by other objective evidence. There is a significant concern that instead the IPPs are being unduly dismissive of data that does not fit their theory. Ultimately, however, that will be for the trier of fact to evaluate further.
Defendants argue that none of these changes or additions to Dr. Flamm's analysis warrants a different conclusion from that reached in the prior order. Defendants fully endorse the order's finding that Dr. Flamm was effectively assuming class-wide impact, and argue he is still doing so. As previously explained, the standard under which expert opinion like that offered by Dr. Flamm is to be evaluated at class certification has been evolving. The DRAM decision suggested that courts "must avoid engaging in a battle of expert testimony." 2006 WL 1530166, at *9. Particularly in light of the "rigorous analysis" required under Dukes and Comcast, however, the caution offered in GPU is apt. "[C]ertification [should not be] automatic every time counsel dazzle the courtroom with graphs and tables." 253 F.R.D. at 491. If the presumption were otherwise, "nearly all antitrust plaintiffs could survive certification without fully complying with Rule 23." Id. at 492; see also, In re High-Tech Employee Antitrust Litig, 289 F.R.D. 555, 567 (N.D. Cal. 2013)("conducting a thorough review of Plaintiffs' theory and methodology is consistent with the requirement that the Court conduct a 'rigorous analysis'" to ensure that the predominance requirement is met"); In re Rail Freight Fuel Surcharge Antitrust Litig.,725 F.3d 244, 255 (D.C. Cir. 2013) ("It is now clear . . . that Rule 23 not only authorizes a hard look at the soundness of statistical models that purport to show predominance—the rule commands it."). Put another way, the inquiry must be to determine if the proffered expert testimony has the requisite integrity to demonstrate class-wide impact.
GPU notes that "antitrust plaintiffs have in recent years trended toward presenting an econometric formula or other statistical analysis to show class-wide impact" and that such analysis has often been accepted at the certification stage. 253 F.R.D. at 491. GPU concluded that "such methods, where plausibly reliable, should be allowed as a means of common proof. To rule otherwise would allow antitrust violators a free pass in many industries." Id. Accordingly, it is clear that statistical and economic methodologies of the sort advanced here may be employed to establish class-wide impact. While defendants have strenuously quarreled with Dr. Flamm's implementation of econometric models, they do not challenge the propriety of employing such models in the first instance.
Defendants and their expert advance a litany of supposed defects in Dr. Flamm's analysis. At heart, though, the biggest dispute between the parties as to the appropriate mode of analysis lies in the degree to which the available data should be "aggregated" or "disaggregated." As defendants point out, where averages and conclusions are drawn from overly-broad data sets, it may conceal a reality that many members of the data set do not share the characteristics reflected in the averages. In crude form, for example, if it were the empirical case that 1 million combo drives were sold with a $10 overcharge each, and 1 million DVD-ROM drives were sold with no overcharge, aggregating the data would show that 2 million drives were sold with an average overcharge of $5, when in fact fully half of the drives were not impacted by conspiracy at all.
The IPPs and Dr. Flamm do not suggest that misleading results could never emerge from over-aggregation. They point out, however, that reducing sample size too far undermines the statistical reliability of any results, possibly masking effects that truly exist by impeding their detection at conventional levels of statistical significance. Thus, they contend, the calculations offered by defendants to suggest the impact of any conspiracy was not-class wide are unreliable, or, at the most, represent a challenge to the merits of Dr. Flamm's calculations, not appropriately resolved at the class certification stage.
Other quarrels defendants have with Dr. Flamm's analysis involve his selection of particular data to utilize, the variables he has elected to include in his regressions, and similar details relating to the implementation of his econometric models. While some or all of these points certainly could be considered by a fact-finder evaluating the persuasiveness of Dr. Flamm's conclusions, defendants have not shown a basis for wholly rejecting those conclusions at this stage in the proceedings as methodologically unsound. Nor does it appear that it would be legitimate to characterize Dr. Flamm's expanded analysis as one which assumes class-wide impact. Rather, he has presented theories that explain why, in his view, class-wide impact would have existed, and he has offered means for testing the data to demonstrate that it did, and to calculate what he believes the overcharges were. Defendants will be free to show why they think Dr. Flamm is wrong, but for purposes of satisfying the requisites of Rule 23, the IPPs have now made an adequate showing of a methodology for proving antitrust injury to all or nearly all direct purchasers, on a class-wide basis.
2. Pass-through
As noted above, IPPs, unlike DPPs, have a two-fold burden in securing class certification in matters such as this. Not only must they show that all or nearly all of the original direct purchasers of ODDs bought at inflated prices, they must also show those overcharges were passed through all stages of the distribution chain. See In re Graphics Processing Units Antitrust Litigation, 253 F.R.D. 478, 499 (N.D. Cal. 2008) ("GPU") ("[I]ndirect-purchaser plaintiffs must demonstrate that defendants overcharged their direct purchasers . . . and that those direct purchasers passed on the overcharges to plaintiffs. In so doing, they must find a way to account for the decision-making of a variety of resellers and manufacturers.").
Because the prior order concluded the IPPs had not shown a viable method for establishing impact to all or nearly all of the underlying direct purchasers, it made no conclusive determinations as to the adequacy of their showing on pass-through. The order, however, expressed skepticism as to whether the issue could ever be resolved on a class-wide basis. Of concern was whether, as in GPU, "the only way to fully assess pass-through in this action would be to conduct a wholesaler-by-wholesaler and re-seller-by-re-seller investigation, which would essentially result in 'thousands of mini-trials, rendering this case unmanageable and unsuitable for class action treatment.'" 253 F.R.D. at 505 (quoting McCarter v. Abbott Labs., Inc., 1993 WL 13011463, at *5 (Ala.Cir.Ct. 1993)). That concern was driven, in large part, by an assumption that the existence and amount of any pass-through might vary greatly, depending on the particular sales channels and other circumstances surrounding how any given ODDs were sold.
The prior order also specifically posited that it might be challenging to establish pass-through given the existence of price points—i.e., a common practice in the industry of selling products costing in the hundreds of dollars at prices just under the next $100 mark. As an example, the prior order suggested it would be implausible for a retailer to raise the price of a computer that otherwise would sell for $999 to $1003, if the overcharge paid by the direct purchaser on the ODD installed in a computer was only four dollars.
The IPPs' renewed motion for class certification makes the important point that the price-fixing in this case is alleged to have consisted of efforts to slow the decline in prices that otherwise naturally was occurring, and there is no allegation prices of ODDs were ever driven up, in absolute terms. Moreover, the naturally occurring cost declines were not limited to ODDs, but took place in most or nearly all of the components typically making up computer systems. As such, the hypothetical that presumed a retailer would be unlikely to increase a price from $999 to $1003 may not be the most effective way of evaluating what might happen in the "but-for" world without the alleged conspiracy. Nevertheless, to the extent "price points" are a significant factor in how computers containing ODDs are sold, they still potentially present issues for which the IPPs will ultimately have to account.
The IPPs present evidence that price points may not be as ubiquitous as implied by the prior order. Unless they are extremely rare, however, which seems unlikely, IPPs presumably will still have to account for their potential effects on pass-through.
For example, if a retailer wants to sell a computer at the $999 price point, and ordinarily would take a profit of $50 on the sale, it must buy from the manufacturer at $949. Assume that in the but-for world, the manufacturer's cost to build that computer would have been $899, given how inexpensive ODDs had become and how far all of the other relevant costs had declined at that particular point in time. The manufacturer could have sold at $949, and also have taken a $50 profit. With a $5 overcharge on the ODD, however, the manufacturer's cost is $904. It is not intuitively obvious that the overcharge would, in all or nearly all cases, get passed on to the end consumer—the IPP. At least in theory, the manufacturer could still sell for $949, and just take a smaller profit. Or, the manufacturer could sell at $954, and the retailer could settle for a $45 profit on the $999 sale to the IPP. In either circumstance, or in any scenario where the manufacturer and retailer each sacrificed some of their profit for a total of $5, none of the overcharge would be passed through to the IPP.
At this juncture, however, the IPPs have presented a sufficient theory and methodology to permit them to attempt to prove pass-through on a class-wide basis. In broad terms, plaintiffs have proffered evidence that in competitive markets, economic theory (supported by empirical studies) consistently predicts that pass-through rates will be at or near 100%. The IPPs assert this is because in a cost-declining market, competition will generally drive prices down to costs. The IPPs offer the explanation that if a firm tries to keep a cost reduction for itself - to increase margins over costs - other firms which also receive similar cost reductions will lower their prices (passing on the cost reductions) to siphon market share from the firm(s) which do not pass-on costs reductions. As the IPPs put it, "[t]his is the nature of competition," and in competitive markets, profit-maximizing behavior virtually always passes on cost-savings in full into lower prices.
There is no significant dispute that the relevant markets here are generally highly competitive. The IPPs offer that they have now measured pass-through rates for over 273 million ODD products. While results show the pass-through rates are not uniform, they are uniformly high and positive - which IPPs contend is sufficient to show that overcharges were consistently passed through to consumers. IPPs contend their analysis covers nearly all of the market in the United States, including companies responsible for approximately 80 percent of personal computer retail sales, and 45 percent of top distributor sales.
Dr. Flamm also offers the results of his efforts to test the impact of price points through a quantile regression analysis. That regression is designed to test the relationship between cost and price changes for computer price points at 99 dollar increments (e.g., $299, $399, $499, etc.). Dr. Flamm opines that these studies show the pass-through rate was 100 percent or greater at all price points.
One aspect of the IPPs' arguments and Dr. Flamm's analysis in the renewed motion has provoked an especially vigorous challenge from defendants. To account for issues raised by initial price setting for new products and/or for efforts by manufacturers and retailers to offer products at particular price points, or within certain price ranges, the IPPs contend that in some instances manufacturers will adjust the "quality" of particular computer systems, rather than the price. In other words, the manufacturer can select the particular components and features to include or omit so as to preserve the expected profit margins for a particular target retail price. Thus, as a simplified example using the hypothetical above, if a manufacturer is paying a $5 overcharge for an ODD, it may choose to omit a camera feature that would have cost it $%. Then, its cost is still $899, and it and the retailer can each still realize a $50 profit on the $999 sale to the IPP.
Defendants insist this theory of pass-through in the form of "lower quality" is fatally flawed for at least two reasons. First, defendants contend it would require a consumer-by-consumer inquiry as to the value each consumer places (or does not place) on any such omitted features. Defendants make the inarguable point that while one consumer may care very much whether or not a computer he or she buys has some particular feature, another may care not at all. Someone intending to use a computer to place a Skype call needs a camera—someone intending only to word-process and surf the internet may not.
Defendant's focus on the subjective desires of individual consumers is misplaced, and is not supported by legal precedent requiring any such approach. Objectively, if a person pays, for example, $999 for a computer system that has no $5 camera when, in a but-for world, an otherwise identical computer sold for $999 would include a $5 camera, then he or she has paid $5 more than the computer is actually worth. A person who has no desire or intention of driving a car at 120 miles an hour is still damaged if he or she pays Ferrari prices for an economy commuter car. Although losing a few dollars' worth of features in a computer is less extreme, the principle is the same. The harm lies in paying more than the product is objectively worth, and does not turn on the individual user's desire or lack of desire for a specific feature.
Defendants contend the court in GPU denied class certification in part because individual inquiries would be necessary regarding what particular consumers would have purchased if an antitrust conspiracy had not allegedly limited choice and suppressed innovation in the market. In contrast, the alleged reduction in "quality" here involves products being sold for more than they would be objectively worth in the but-for world. The concern expressed by the GPU court is not implicated by these circumstances.
Defendants' second attack on the "lower quality" theory relies on language in GPU that "to satisfy Section 4 of the Clayton Act, plaintiffs must demonstrate that they paid a higher price for their graphics card or computer than they otherwise would have paid in the absence of a conspiracy." GPU, 253 F.R.D. at 507. The GPU court, however, made that response to a suggestion by the plaintiffs that even if they could not show they had suffered any economic injury, they should still be entitled to recover nominal damages for having "lower quality, less choice, and reduced innovation." Id. Here, IPPs are not offering the "reduced quality" theory as an alternative to economic injury, but as the means by which they intend to argue they overpaid in some instances for computers containing ODDs.
Defendants' further challenges to the adequacy of the IPPs' showing on pass-through fall into three basic categories. First, defendants argue that Dr. Flamm's studies are flawed and misleading because, in their view, he has not looked at sufficient relevant categories of data, and/or because he has again improperly aggregated data in ways that conceal the existence, or potential existence, of ODD sales in which there was no pass-through. As before, the IPPs have adequately shown that any questions regarding the appropriate degree to which data should be aggregated to derive reliable results fall on the "merits" side of the line and do not present grounds for denying class certification at this point in time.
While it may be that Dr. Flamm and the IPPs have not "studied" pass-through in all of the various marketing chains and manners through which ODDs are sold, this too falls into the category of arguments defendants may present to a fact-finder as reasons to reject Dr. Flamm's conclusions, rather than a basis for denying certification. Furthermore, as noted, the IPPs' basic theory is that cost-savings are virtually always passed through in general circumstances like those present here. The mere fact that the IPPs have not "tested" the validity of that theory in all the relevant factual permutations does not give rise to a contrary inference that pass-through likely does not occur in the particular instances the IPPs have not closely examined.
Next, defendants contend the IPPs have failed to account for "real world" evidence that cost-savings on component parts are not always passed through. To the extent defendants rely on testimony from industry participants denying that cost-savings are always passed through, they have, in the language of summary judgment motion practice, merely created a triable issue of fact. In the context of class certification, they have not shown that the existence of such testimony compels a conclusion that the IPPs' basic economic theory is so lacking in possible validity that it can be legally foreclosed at this juncture.
Defendants also rely on evidence of other counter examples, including the fact that Dell charged the same price for adding an ODD to customer-configured computer systems over an extended time period during which Dell's costs for those ODDs presumably was falling. While the IPPs would have that evidence disregarded on various grounds at this juncture, it seems likely that defendants will be able to cure any technical deficiencies in their presentation by the time of trial. Nevertheless, the mere possibility that defendants ultimately will be able to prove on the merits that cost savings were not always passed through is not sufficient reason to deny certification now.
Finally, defendants argue that the IPPs have not accounted for, and will never be able to account for, instances in which retailers sold computer systems below cost, provided discounts or rebates, or bundled products together. Defendants contend in such instances any overcharges were necessarily not passed-through, or the amount of pass-through is inherently not ascertainable. The IPPs' response is simple. If, for instance, a retailer obtained a computer at $600, including a $5 overcharge, and for whatever reason sold it "below cost" at $500, then it is reasonable to assume in the but-for world the retailer would have paid $595 for the computer and sold it at $495. The overcharge is passed through to the IPP without regard to whether the retailer sells above or below cost. A similar analysis can be applied where there is "bundling," discounts, or rebates. Defendants have not shown why the IPPs should be legally precluded from proceeding on this theory.
None of this is to say, of course, that the IPPs ultimately will be able to prove pass-through in fact took place across the board as they contend. The crucial point is that whether the IPPs theory is right or wrong, it is something that can be decided on a class-wide basis. While defendants have offered some compelling evidence apparently contradicting the IPPs' contention that cost savings are virtually always passed through, that evidence, as well as defendant's attacks on the completeness and accuracy of Dr. Flamm's pass-through studies, present issues to be decided on the merits.
Because plaintiffs acknowledge that even under their theory and studies pass-through rates were not entirely uniform, there may be individualized issues of damages that will require further attention. At least at this juncture, however, there is no indication the issue is one that would defeat certification in the first instance. See Jimenez v. Allstate Ins. Co., 765 F.3d 1161, 1167 (9th Cir. 2014) ("[i]n this circuit . . . damage calculations alone cannot defeat class certification.") (citation omitted).
C. Rule 23(a) certification requirements
The discussion above focused on class certification requirements arising under Rule 23(b), as those have been the central focus of the controversy, both in the original motions by the DPPs and the IPPs, and in this renewed motion by the IPPs alone. The prior order suggested that the IPPs faced few challenges in showing that the requirements of Rule 23(a) for certification of a class could be satisfied. Rule 23(a)(1) requires that a class be so numerous that joinder of all members is impracticable. It remains the case that there is no dispute the "numerosity" requirement is met here. To satisfy this requirement, plaintiffs need not state the "exact" number of potential class members, nor is there a specific minimum number that is required. See In re Rubber Chem. Antitrust Litig., 232 F.R.D. 346, 350-51 (N.D.Cal.2005); In re Bulk [Extruded] Graphite Prod. Antitrust Litig, 2006 WL 891362, *5 (D.N.J.2006). Here, the number of indirect purchasers of computers containing ODDs and standalone ODDs plainly is sufficient to support class treatment.
Defendants, however, now challenge the IPPs' ability to establish "ascertainability," a requirement not expressly stated in the rule. See In re High-Tech Employee Antitrust Litig., 289 F.R.D. 555, 563 (N.D. Cal. 2013) (members of the class must be sufficiently definable and identifiable). Although the concern raised by defendants that class members will be subjected to a "memory" test as to the source of their ODDs may present some issues, at this juncture it appears that retailers' records likely will suffice to identify a substantial and sufficient portion of the class.
Rule 23(a)(2) requires that there exist "questions of law or fact common to the class." Where an antitrust conspiracy has been alleged, courts have frequently held that "the very nature of a conspiracy antitrust action compels a finding that common questions of law and fact exist." See In re Rubber Chem. Antitrust Litig., 232 F.R.D. at 351; In re Bulk [Extruded] Graphite Prod. Antitrust Litig., 2006 WL 891362 at *5. As in the prior order, at a minimum, the existence, scope, and efficacy of the conspiracy to fix or stabilize prices of ODDs sold in the United States are all common questions that IPPs must address. Also as before, the extent to which such common questions predominate, and/or can be answered through common proof, has been addressed above, in connection with Rule 23(b)(3).
Rule 23(a)(3) also requires that the claims of the named plaintiffs be typical of those of the class. "Typical" does not mean "identical to." Rather, typicality results if the representative plaintiffs' claims "arise[ ] from the same event, practice or course of conduct that gives rise to the claims of the absent class members and if their claims are based on the same legal or remedial theory." In re Auction Houses Antitrust Litig., 193 F.R.D. 162, 164 (S.D.N.Y. 2000).
In evaluating typicality, the court should consider whether the named plaintiffs' "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." See In re Bulk [Extruded] Graphite Prod. Antitrust Litig., 2006 WL 891362 at *5. In cases involving an alleged price-fixing conspiracy, the representative plaintiff's claim is often considered typical even where the plaintiff followed different purchasing procedures, purchased in different quantities or at different prices, or purchased a different mix of products than did the members of the class. See, e.g., In re Flat Glass Antitrust Litig, 191 F.R.D. 472, 479 (W.D.Pa.1999); In re Indus. Diamonds Antitrust Litig, 167 F.R.D. 374 (S.D.N.Y.1996).
Addressing a closely-related issue, Rule 23(a)(4) requires that the representative parties fairly and adequately protect the interests of the class. This "adequacy" requirement mandates that no conflicts of interest exist between the named plaintiffs and the absent class members. See, e.g., In re Rubber Chem. Antitrust Litig., 232 F.R.D. at 351, citing Local Joint Exec. Bd. of Culinary/Bartender Trust Fund v. Las Vegas Sands, Inc., 244 F.3d 1152, 1162 (9th Cir. 2001).
The adequacy requirement also contains an element relating to the competence and diligence of class counsel in pursuing the claims. See id.; see also In re Linerboard Antitrust Litig., 203 F.R.D. 197, 207 (E.D. Pa. 2001). There is no concern in this instance with the adequacy of class counsel.
While the prior order raised questions as to the typicality and adequacy of the representative plaintiffs in the DPP action, it expressly stated that there did not appear to be serious concerns as to the named IPPs. Defendants now argue, however, that because some major IPPs were business entities, the individuals named as representative plaintiffs may not adequately advance their interests. While there are surface similarities between this issue and the concerns raised in the prior order as to the DPP representatives, there is nothing that suggests it presents a serious barrier to certification.
Finally, Rule 23(b)(3) also requires that a class action be superior to other methods of adjudication. Some authority suggests that the superiority requirement is satisfied almost a fortiori if common questions are found to predominate in an antitrust action. See In re TFT-LCD (Flat Panel) Antitrust Litig, 267 F.R.D. 583, 608 (N.D. Cal. 2010). Again, while there were concerns on this point with respect to the DPPs' motion given the identity and size of some of the largest direct purchasers of ODDs, in the case of this renewed motion by the IPPs, there is nothing to suggest the requirement of "superiority" presents any particular issues not subsumed elsewhere in the analysis.
Accordingly, because the IPPs have now made an adequate showing that they can satisfy the requisites of Rule 23, their motion for class certification may and will be granted. The remaining issue is the scope of the class to be certified.
As tentatively suggested in the prior order, and confirmed now, there is no reason to conclude defendants' intent to raise a defense pursuant to the Foreign Trade Antitrust Improvements Act, 15 U.S.C. § 6a ("FTAIA") precludes certification or cannot be decided on a class-wide basis. See In re TFT-LCD (Flat Panel) Antitrust Litig., 267 F.R.D. 583, 599 (N.D. Cal. 2010) ("The Court concludes that the FTAIA defense can be resolved on a class basis, and thus the possibility of this defense does not provide a basis for denying certification. Courts have emphasized that the FTAIA analysis focuses on the defendants' conduct."). While defendants contend there will be variations among ODD sales as to the applicability of the FTAIA and its exceptions depending on the circumstances of sales in the supply chain, such differences turn on defendants' actions and do not appear to require individualized inquiries of the sort that present a barrier to class certification.
D. Scope of the Class—due process and choice of law
In order of preference, the IPPs seek certification of:
1. A nationwide class under California law,
2. A class of plaintiffs from 23 states plus the District of Columbia, under California law, or
3. 24 separate classes, under the laws of each of the relevant jurisdictions.
Either of the first two options would require applying California law to purchases made in other states. That is permissible only if the IPPs show that doing so comports with both (1) due process, and (2) California's choice of law rules. See Mazza v. Am. Honda Motor Co., 666 F.3d 581, 589-90 (9th Cir. 2012).
In AT&T Mobility LLC v. AU Optronics Corp, 707 F.3d 1106 (9th Cir. 2013) the Ninth Circuit held that California's "Cartwright Act can be lawfully applied without violating a defendant's due process rights when more than a de minimis amount of that defendant's alleged conspiratorial activity leading to the sale of price-fixed goods to plaintiffs took place in California." Id. at 1113. As the court further observed, "[s]uch a defendant cannot reasonably complain that the application of California law is arbitrary or unfair when its alleged conspiracy took place, at least in part, in California." Id.
Although AT&T Mobility was decided at the pleading stage, here the IPPs have presented sufficient evidence of each defendant's alleged conspiratorial conduct in California to satisfy their obligation to establish it was "more than a de minimis amount." Although the showing is admittedly thinner as to some defendants than others, it is relevant that none of the remaining defendants are based anywhere other than in California or overseas. As such, defendants have no reasonable basis to contend application of California law offends due process.
The choice of law analysis proceeds in three steps. First, the court determines whether the laws of the affected jurisdictions are "the same or different." Mazza, 666 F.3d at 589 (quoting McCann v. Foster Wheeler LLC, 225 P.3d 516, 527 (Cal. 2010)). Second, if the laws are different, the court "examines each jurisdiction's interest in the application of its own law" to determine whether a "true conflict" exists. Id. Finally, if a true conflict exists, the court determines "which state's interest would be more impaired if its policy were subordinated to the law of the other state." Id.
Here, all parties agree the most significant difference in law among the various jurisdictions lies in whether or not a particular jurisdiction follows the rule of Illinois Brick or is instead an "Illinois Brick repealer." The IPPs concede this is a "true conflict." IPPs would lack standing to pursue this action in the non-repealer jurisdictions, absent application of California law.
Illinois Brick Co. v. Illinois, 431 U.S. 720 (1977).
The IPPs argue, however, that the conflict can be disregarded because non-repealer states have no particular interest in precluding California from applying its laws to provide remedies to anyone injured by antitrust and unfair competition emanating from the state. Given that the action simply could not go forward in non-repealer states, however, it is too much of a stretch to employ California law as an end run around the limitations those states have elected to impose on standing. See In re TFT-LCD (Flat Panel) Antitrust Litig., 2013 WL 4175253, at *2 (N.D. Cal. July 11, 2013) (rejecting application of California law to a non-repealer state, in light of that state's interest in limiting standing). Accordingly, a nationwide class will not be certified.
Defendants argue that California law should not even be applied to repealer states, asserting that there still exist numerous differences between the California antitrust and unfair competition laws and those in many of the other 24 jurisdictions. The parties were specifically given the opportunity to submit additional briefing to flesh out their arguments on the conflicts that would exist solely in the context of a 24-jurisdiction class. Apart from the Illinois Brick issue, however, the potential differences identified between California and some of the other jurisdictions do not appear to stand as true conflicts, or as ones that should not yield to California's interests. Accordingly, the IPPs motion for class certification will be granted as to the 24 jurisdictions they have identified, under California law.
Defendants suggest that in at least some of the non-California jurisdictions, the principles of Associated General Contractors of California v. California State Council of Carpenters, 459 U.S. 519 (1983) might be applied under local law to preclude IPP recovery. If so, that would present a conflict similar to that existing in non-Illinois Brick repealer jurisdictions. At least at this juncture, though, the record does not support a conclusion that the outcome in those jurisdictions would be materially different under local law than under California law.
--------
V. CONCLUSION
The motion for class certification is granted, to the extent set out above. Within 30 days of the date of this order, the parties shall submit jointly a report as to their respective positions on how the litigation should proceed from this juncture. IT IS SO ORDERED. Dated: February 8, 2016
/s/_________
RICHARD SEEBORG
UNITED STATES DISTRICT JUDGE