Summary
holding that at the class certification stage, plaintiffs need not show that their expert's model "actually works," so long as it is "workable"
Summary of this case from In re Air Cargo Shipping Servs. Antitrust Litig.Opinion
RULING ON MOTION FOR CLASS CERTIFICATION
STEFAN R. UNDERHILL, District Judge.
This multidistrict litigation arises out of allegations that the defendants conspired to fix, raise, maintain, and stabilize the price of ethylene propylene diene monomer (" EPDM" ) synthetic rubber at artificially high, non-competitive levels in violation of federal antitrust law. The named plaintiffs, a group of EPDM purchasers, seek to certify a class in this antitrust suit on behalf of all individuals and entities that purchased EPDM from January 1, 1997 to December 31, 2001. The plaintiffs are seeking injunctive relief and treble damages, plus costs, including reasonable attorneys' fees. All the defendants named in the Second Consolidated Amended Complaint have been dismissed or have entered into settlement agreements with the plaintiffs, except for defendants DSM Copolymer, Inc. and DSM Elastomers Europe B.V. (collectively, the " defendants" ).
The plaintiffs contend they have met their burden of meeting all the requirements of Federal Rule of Civil Procedure 23, which governs the procedure for class certification. The defendants dispute that the plaintiffs have met these requirements, specifically contending that the plaintiffs have failed to demonstrate how common questions of law or fact will predominate over individual ones on the issue of antitrust injury. Because I find that plaintiffs have met their burden under Rule 23, for the reasons explained more fully below, the motion for class certification is granted.
I. Factual Background
EPDM is a synthetic rubber used primarily in the automotive and roofing industries, but it is also found in glass-run channels, radiators, garden and appliance hoses, tubing, belts, electrical insulation, rubber mechanical goods, plastic impact modifications, thermoplastic vulcanizates, and motor oil additive applications. EPDM is the third-largest type of synthetic rubber consumed worldwide and is popular for its resistance to heat, oxidation, ozone, and weather aging. It is one of the fastest growing synthetic rubber products, reaching sales of 870,000 metric tons in 2002 from 289,000 metric tons in 2000 (an increase of approximately 300%).
The factual background about the EPDM industry is drawn primarily from the Second Amended Consolidated Complaint. These facts are not material for purposes of this motion and are included for informational purposes only.
The automotive industry accounts for two-thirds of the overall demand for EPDM. Approximately 20% of the rubber used in a typical automobile is EPDM. It is also extremely prevalent in the roofing industry, where it is the most widely used roofing material, accounting for over one billion square feet of new roof coverings in the United States.
The plaintiffs have alleged that the defendants conspired to fix, maintain and/or stabilize EPDM prices in the United States in violation section 1 of the Sherman Act, 15 U.S.C § 1, by engaging in collusive discussions, meetings, agreements, and understandings to raise prices to supracompetitive levels in tandem and to stabilize the supply of EPDM from competitors outside the cartel. The plaintiffs allege that the defendants met regularly to discuss EPDM prices-at trade association meetings or at their offices-and engaged in regular written communications with one another to maintain their collusive agreement. The plaintiffs contend that, pursuant to their collusive agreements, the defendants took turns leading EPDM price increases, conferring months in advance of an announced price increase about whose turn it was to raise prices.
Plaintiffs allege that defendants sought to cut off new sources of capacity by buying up new production in Asia and eventually setting up market allocation agreements between North American suppliers and those Asian suppliers. The plaintiffs allege the global conspiracy was most effective in the U.S. and Canadian markets where prices were approximately 10-15% higher than in the Western Europe market.
The plaintiffs allege that the defendants were so protective of their cartel that when a defendant's production capacity dropped due to plant or production problems, its competitors would provide that defendant with EPDM, so that it could continue to supply its customers, rather than taking the opportunity to compete for those customers directly. In return for receiving the additional supply of EPDM, the plaintiffs allege the struggling supplier would agree not to compete for the supplying party's customers. Thus, set market share and customer allocations were maintained despite individual supplier disruptions. The alleged conspiracy was not disclosed to the public until December 2002, when defendant Crompton announced that it was being investigated by competition authorities in the United States, Canada, and Europe. The plaintiffs allege that, despite their exercise of due diligence, they would not have been able to discover the conspiracy prior to December 2002 due to the defendants' active and purposeful concealment of the cartel and its activities.
The plaintiffs wish to certify the following class:
All individuals or entities (excluding governmental entities, Defendants and their parents, predecessors, subsidiaries, affiliates, and co-conspirators) who purchased ethylene propylene diene monomer (" EPDM" ) in the United States directly from one or more of the Defendants or any predecessor, subsidiary or affiliate of any Defendant at any time during the period from January 1, 1997 to December 31, 2001.
The plaintiffs allege the class is readily definable, and can be definitively proven through the sales and invoice records of the defendants.
In support of their motion for class certification, the plaintiffs have submitted evidence of six national lock-step price list increases within two broad categories of EPDM products: clear elastomers and oil extended elastomers. Table 1 to the Report of Dr. Martin Asher (" Asher Report" ). These price list increases indicate that the five primary EPDM suppliers increased the cost of their products by the same amount, at about the same time, at least six times during the class period. The plaintiffs allege those price hikes were not in line with input prices and were the result of communications between the defendants to sustain supracompetitive prices.
Oil extended elastomer EPDM is created by adding petroleum-based extender oil during the manufacturing process, which alters the properties of the polymer and lowers the manufacturing cost. Asher Report at ¶ 22.
In addition, the plaintiffs have submitted an econometric multiple regression analysis that purports to calculate class-wide damages. The plaintiffs' experts explain that regression modeling reveals that the price of EPDM was affected by more than just the cost of its input and the level of demand for the products, meaning there was an active conspiracy at work to maintain EPDM prices at supracompetitive levels.
II. Standards for Class Certification
The Second Circuit recently clarified the standard of review for assessing motions for class certification in In re Initial Public Offerings Securities Litigation (" In re IPO" ), 471 F.3d 24, 41-42 (2d Cir.2006). The Court disavowed the previous standard of review, holding that a district court should no longer grant class certification when the plaintiffs had merely made " some showing" of the Rule 23 prerequisites, nor should an expert's report be accepted to establish a Rule 23 requirement on the basis that it was " not fatally flawed." Id. at 32, 40, 42 (quoting, respectively, Caridad v. Metro-North Commuter R.R., 191 F.3d 283, 292 (2d Cir.1999) and In re Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124, 135 (2d Cir.2001)).
Before certifying a class, the district court must undertake a " rigorous analysis" to determine whether each of the Rule 23 requirements has been met. Id. at 33, 41 (quoting General Telephone Co. of the Southwest v. Falcon, 457 U.S. 147, 161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982)). Significantly, the judge must resolve any factual dispute material to a Rule 23 requirement. Id. That obligation is not lessened if the necessary factual finding overlaps with a merits issue-rather, the court should resolve the dispute as it would any other threshold issue. Id. at 41-42. " [T]he determination as to a Rule 23 requirement is made only for purposes of class certification and is not binding on the trier of facts, even if that trier is the class certification judge." Id. at 41. The Court cautioned, however, that considering facts for purposes of determining whether the plaintiff had established all the Rule 23 requirements is not a license to delve into all merits issues, and that the district court should not assess " any aspect of the merits unrelated to a Rule 23 requirement." Id. Finally, the " district judge has ample discretion to circumscribe both the extent of discovery concerning Rule 23 requirements and the extent of a hearing ... to assure that a class certification hearing does not become a pretext for a partial trial of the merits." Id. Keeping these precautions in mind, however, " the district judge must receive enough evidence, by affidavits, documents, or testimony, to be satisfied that each Rule 23 requirement has been met." Id.
Analogizing the Rule 23 inquiry to the preliminary injunction process, the Court held that any findings on the merits at the class certification stage would not be binding for the ultimate fact-finder- " [a] trial judge's finding on a merits issue for purposes of a Rule 23 requirement no more binds the court to rule for the plaintiff on the ultimate merits of that issue than does a finding that the plaintiff has shown a probability of success for purposes of a preliminary injunction." Id. at 39 (emphasis in original).
The plaintiffs here seek to certify a class under Rule 23(b)(3), which requires them first to prove the four threshold requirements under Rule 23(a)-numerosity, common questions of law or fact, typicality of the lead plaintiff's claims or defenses, and adequacy of representation by the lead plaintiff-and two additional requirements under Rule 23(b)(3): (1) " that questions of law or fact common to class members predominate over any questions affecting only individual members," and (2) that a class action would be " superior to other methods for fairly and efficiently adjudicating the controversy." Fed.R.Civ.P. 23(a) and (b)(3); McLaughlin v. Am. Tobacco Co., 522 F.3d 215, 222 (2d Cir.2008). These requirements " ensure [ ] that the class will be certified only when it would ‘ achieve economies of time, effort, and expense, and promote ... uniformity of decision as to persons similarly situated, without sacrificing procedural fairness or bringing about other undesirable results.’ " Cordes & Co. Fin. Servs. v. A.G. Edwards & Sons, Inc., 502 F.3d 91, 104 (2d Cir.2007) (quoting Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 615, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997)).
In addition, Rule 23(b)(3) considers the following factors relevant when determining whether common issues predominate and whether class certification is a superior method of adjudication:
At the February 8, 2008 hearing on the issue of class certification, the defendants conceded that the plaintiffs have met the four threshold requirements under Rule 23(a) and I find that they have. The only real issue on this motion for class certification is whether the plaintiffs have met their burden of proving that common questions predominate over individual ones as required by Rule 23(b)(3). The defendants contest the superiority prong of Rule 23(b)(3) on the same ground that they contest the predominance prong-that a class action is not a superior method of adjudication because individual questions predominate over common questions, thus making it more efficient to litigate the claims on an individual basis. Therefore, this ruling will focus on Rule 23(b)(3)'s predominance requirement-whether or not individual questions of law and fact predominate over common questions.
III. Discussion
A. Elements of an Antitrust Claim
The plaintiffs filed suit under section 1 of the Sherman Act, alleging, inter alia, that the defendants conspired " to fix, raise, maintain and/or stabilize the price of EPDM sold in the United States, including limiting supply and/or allocating markets and customers for the sale of EPDM in the United States." Second Consolidated Amended Complaint ¶ 44a. Section 1 of the Sherman Act provides that " [e]very ... conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is hereby declared to be illegal." 15 U.S.C. § 1. The Clayton Act permits private litigants to seek treble damages for violations of the antitrust laws. 15 U.S.C. § 15. A successful antitrust claim requires the plaintiff to prove: " (1) a violation of antitrust law; (2) injury and causation; and (3) damages." Cordes, 502 F.3d at 105 (internal alterations omitted) (quoting In re Visa Check, 280 F.3d at 136). The injury and causation element has also been referred to as " antitrust injury" and " causation or impact." Id. at 105.
B. Antitrust Elements and the Predominance Requirement
As explained above, to prevail in their motion for class certification, the plaintiffs must demonstrate that common questions of law or fact predominate over individual ones on the issues relevant to the three elements of an antitrust claim. " The predominance requirement is met if the plaintiff can ‘ establish that the issues in the class action that are subject to generalized proof, and thus applicable to the class as a whole, ... predominate over those issues that are subject only to individualized proof.’ " Id. at 107-08, 502 F.3d 91 (quoting In re Visa Check, 280 F.3d at 136). In support of their class action motion, the plaintiffs have proffered the six national price lists and their econometric modeling, which they assert show that the relevant antitrust issues-and more specifically, antitrust injury-are susceptible to generalized proof. The defendants argue that the antitrust injury element is not susceptible to common proof and that the econometric model presented by the plaintiffs is incapable of demonstrating class-wide injury or damages.
1. Violation of Antitrust Law
" Horizontal price-fixing agreements are per se violations of the Sherman Act." Id. at 105. " In general, the person who has purchased directly from those who have fixed prices at an artificially high level in violation of the antitrust laws is deemed to have suffered the antitrust injury...." State of N.Y. v. Hendrickson Bros., Inc., 840 F.2d 1065, 1079 (2d Cir.1988). In Cordes, the plaintiffs alleged that the defendants, a group of IPO underwriters, entered into a horizontal price-fixing agreement to fix their fees at seven percent of the IPO proceeds for all corporations conducting " mid-sized IPOs" (IPOs that generated between $20 million and $80 million). Cordes, 502 F.3d at 96. Because " allegations of the existence of a price-fixing conspiracy are susceptible to common proof," the Court determined that common questions predominated over individual ones on this first element of an antitrust claim, violation of antitrust law. Id. at 105. See also In re Foundry Resins Antitrust Litig., 242 F.R.D. 393, 409 (S.D.Ohio 2007) (holding that the plaintiffs' allegations of horizontal price-fixing were sufficient to establish a per se violation of antitrust law).
Similarly, the plaintiffs here have alleged that EPDM producers engaged in an illegal price-fixing agreement to raise, maintain, and/or stabilize the price of EPDM in the United States. That issue is common to the class, and if proven true, would satisfy the first element of an antitrust cause of action, namely, a violation of antitrust law. Therefore, the plaintiffs have met their burden of showing predominance on this element.
2. Antitrust Injury
In Cordes, the Second Circuit held that plaintiffs must make a two-part showing to establish the antitrust injury element. 502 F.3d at 106. " One [prong] is the familiar factual question whether the plaintiff has indeed suffered harm, or ‘ injury-in-fact.’ The other is the legal question whether any such injury is ‘ injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants' acts unlawful.’ " Id. (quoting Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489, 97 S.Ct. 690, 50 L.Ed.2d 701 (1977)).
a. Legal Question Prong
To demonstrate the legal question prong of the antitrust injury element, the plaintiff must prove that the defendants engaged in an anti-competitive manipulation of the markets. Id. at 105-07. In Cordes, the Court held the plaintiffs successfully established that the legal question was common to the class and thus predominated over individual questions because the plaintiffs alleged only one type of injury in the complaint-" overcharges paid to a horizontal price-fixing conspiracy." Id. at 107. " Because each class member allegedly suffered the same type of injury, the legal question of whether such an injury is ‘ of the type the antitrust laws were intended to prevent and that flows from that which makes defendants' acts unlawful,’ is a common one." Id. (internal citations omitted) (quoting Brunswick, 429 U.S. at 489, 97 S.Ct. 690).
Like the plaintiffs in Cordes, the EPDM plaintiffs have met the first part of the antitrust injury showing: that the legal question of the antitrust injury is common to the class. The plaintiffs have alleged only one type of injury in the complaint, namely that they paid supracompetitive prices for EPDM due to an illegal conspiracy to raise, maintain, and/or stabilize the prices of EPDM on the U.S. market. The legal question of injury is common to the class.
b. Factual Question Prong
The factual question prong of the antitrust injury element requires the plaintiffs to show that harm actually resulted from the antitrust violation-the " injury-in-fact," id. at 106, or " impact." To meet the predominance requirement of Rule 23 for class certification, the plaintiffs must show that injury-in-fact, or impact, can be proven by evidence common to the class. Id. at 106. Plaintiffs have offered two forms of common proof to assert that impact was class-wide: six national price list increases adhered to by all the major EPDM suppliers, showing that the defendants conspired to maintain an elevated " base" price for EPDM, and an economic regression model.
It is important to note at the outset that " impact" (or " injury-in-fact" ) and " damages" are two distinct elements of an antitrust claim-injury-in-fact is whether the plaintiffs were harmed and damages quantify by how much. Id. at 107 n. 11; see also In re Urethane Antitrust Litig. (" Urethane II, Polyether Polyol" ), 251 F.R.D. 629, 638 (D.Kan.2008) (" ‘ [T]he issue in the common impact analysis is the fact, not the amount, of injury.’ " ) (quoting In re Potash Antitrust Litig., 159 F.R.D. 682, 695 (D.Minn.1995)). At times these terms have been conflated, which can lead to confusion about what a plaintiff's burden precisely is at the motion for class certification stage. It is true that, generally speaking, antitrust injury-in-fact and damages are often determined by comparing the " but-for" price-the price a customer would have paid in the absence of the conspiracy-and the actual price paid. Cordes, 502 F.3d at 107. " [T]he extent of the difference between the but-for [price] and the actual [price] paid is relevant to the question of damages, but it is from a comparison between the two that the court would be asked to decide question of injury-in-fact." Id. (emphasis added). In such a scenario, the plaintiff has shown impact and damages in a single calculation: if he paid a higher actual price than the but-for price, he has shown impact; by calculating the extent of that difference he has shown damages. Hence, one way of demonstrating predominance is to show that there is a common method for proving that the class plaintiffs paid higher actual prices than in the but-for world, such as using an econometric regression model incorporating a variety of factors to demonstrate that a conspiracy variable was at work during the class period, raising prices above the " but-for" level for all plaintiffs.
However, where other methods of common proof exist to show class-wide impact such as lock-step increases of national price lists in an oligopolistic market, comparing " but-for" prices with actual transaction prices is not the only way for plaintiffs to succeed in a motion for class certification. Whenever actual prices are higher than " but-for" prices, a plaintiff suffers damages and must necessarily have been impacted, or injured-in-fact, too. Using a damages calculation is often the only way to show impact in antitrust scenarios where the industry is not served by a small number of distributors, where sales occur between diverse and independent merchants and customers, or where there is no direct evidence of class-wide conspiratorial effect. In such scenarios, proving damages may be the only way to show impact.
Proving damages proves injury because damages necessarily indicate that the plaintiff has been impacted or injured by the antitrust violation; the converse, however, is not necessarily true. That is, it is possible for a plaintiff to suffer antitrust injury-in-fact and yet have no damages because it has taken steps to mitigate the actual price paid through rebates, discounts, and other non-price factors such as lowered shipping costs, technical services, or any other type of purchase incentive. By expending resources to negotiate down from the supra competive prices established by the cartel, plaintiffs who have suffered no damages may still have suffered an injury-in-fact from the antitrust conspiracy. The fact that a plaintiff may have successfully employed bargaining power to fend off the effect of the conspiratorial practices does not mean that it has not been put in a worse position but-for the conspiracy.
" Neither a variety of prices nor negotiated prices is an impediment to class certification if it appears that plaintiffs may be able to prove at trial that ... the price range was affected generally." In re NASDAQ Market-Makers Antitrust Litig., 169 F.R.D. 493, 523 (S.D.N.Y.1996).
The proof necessary to demonstrate that the defendants conspired to maintain an inflated " base" from which all pricing negotiations began and that this " base" price was higher than the " base" price which would have been established by competitive conditions would be common to all members of the class. Proof of a conspiracy to establish a " base" price would establish at least the fact of damage, even if the extent of the actual damages suffered by the plaintiffs would vary.... [T]he proof with respect to the " base" price from which these negotiations began, or the structure of the conspiracy to affect individual negotiations, would be common to the class. Accordingly ... the fact of damage is predominantly, if not entirely, a common question.
Id. (emphasis added) (quoting Hedges Enterprises, Inc. v. Continental Group, Inc., 81 F.R.D. 461, 475 (E.D.Pa.1979)). See also Urethane II, Polyether Polyol, 251 F.R.D. at 638 (" [E]vidence of a standardized pricing structure, which (in light of the alleged conspiracy) presumably establishes an artificially inflated baseline from which any individualized negotiations would proceed, provides generalized proof of class-wide impact." ); In re Urethane Antitrust Litig. (" Urethane I, Polyester Polyol" ), 237 F.R.D. 440, 450-51 (D.Kan.2006) (accepting the plaintiffs' expert's structural analysis of the urethane market and opinion that price lists can be used as proof of common impact; " conspiratorial behavior elevates list prices and list prices serve as a reference or benchmark for pricing or pricing negotiations" ); Fears v. Wilhelmina Model Agency, Inc., 2003 WL 21659373, at *6 (S.D.N.Y.2003) (noting that, even if the plaintiffs had been able to negotiate for a lower commission rate with the defendant modeling agencies, they were nonetheless impacted by the conspiracy because those negotiations still began at the artificially elevated rate established by the cartel); In re Industrial Diamonds Antitrust Litig., 167 F.R.D. 374, 383 (S.D.N.Y.1996) (noting that " [t]he theory that underlies these decisions is, of course, that the negotiated transaction prices would have been lower if the starting point for negotiations had been list prices set in a competitive market. Hence, if a plaintiff proves that the alleged conspiracy resulted in artificially inflated list prices, a jury could reasonably conclude that each purchaser who negotiated an individual price suffered some injury" ).
The plaintiffs here assert that the six across-the-board list price increases for EPDM, which the defendants do not dispute, suffice as common proof of impact to the class, particularly when the EPDM market's characteristics tending to support collusive activity are also considered, because those price list increases had the effect of raising prices for all EPDM purchasers. Defendants argue that, because the plaintiffs have not conclusively demonstrated how those price increases raised prices for each transaction with each plaintiff, they cannot serve as common proof of impact, pointing to individual discounts, price rebates, and contract terms that would have lowered the price of EPDM below the list price for certain plaintiffs. The defendants misconstrue the plaintiffs' burden at the class certification stage. The plaintiffs need not demonstrate to an absolute certainty that all EPDM purchasers ultimately suffered damages as a result of inflated list prices. Addressing a similar argument in Urethane II, Polyether Polyol, the Court noted that the absence of an actual price increase for an individual plaintiff following a price list increase was not a relevant issue at the class certification stage. 251 F.R.D. at 637 n. 6. " Aside from the fact that this argument goes to the merits of the case and not the issue of whether plaintiffs' case concerning impact is amenable to class-wide proof, the mere fact that prices did not increase at the seemingly appropriate times does not conclusively establish that they were not artificially inflated so as to keep them from falling to the extent that they might have done so in a competitive market." Id. (emphasis in original).
In re IPO directs courts not to grant motions for class certification if plaintiffs have merely made " some showing" of the Rule 23 requirements or because their expert's report is " not fatally flawed." As the In re IPO Court stated, " the important point is that the requirements of Rule 23 must be met, not just supported by some evidence.... ‘ [A]ctual, not presumed, conformance with Rule 23(a) remains ... indispensable.’ " 471 F.3d at 33 (quoting Falcon, 457 U.S. at 160, 102 S.Ct. 2364). When assessing the predominance requirement of Rule 23(b)(3), however, I need only determine whether the element of injury-in-fact can be proven by evidence common to the class. Urethane II, Polyether Polyol, 251 F.R.D. at 636 (noting that the pertinent legal inquiry at class certification is not whether the plaintiffs have proven on the merits that they paid an artificially high price due to the defendants' alleged price-fixing conspiracy, but rather " ‘ whether class-wide impact may be proven by evidence common to all class members.’ " ) (quoting In re Bulk (Extruded) Graphite Prods. Antitrust Litig., 2006 WL 891362, at *2 (D.N.J.2006)). The plaintiffs have offered sufficient proof and explanation to convince me that these six national price list increases could have the effect of affecting the base price for EPDM, from which all negotiations started, on a class-wide basis.
As discussed above, the Court held this requirement applies with " equal force to all Rule 23 requirements." In re IPO, 471 F.3d at 33 n. 3.
The plaintiffs have not merely alleged that these price lists existed and that they affected all EPDM purchasers-they have shown undisputed evidence of lock-step price increases and provided expert opinions that the structural characteristics of the EPDM market would support collusive increases of prices to artificially high levels. As discussed more fully below, the defendants do not dispute the price list increases, nor do their experts truly contest the nature of the EPDM market, but instead argue that the plaintiffs cannot prevail on the merits, because they cannot prove that every class member ultimately suffered damages. That argument is best reserved for their argument that common questions do not predominate on the issue of damages and for trial on the merits.
The plaintiffs' expert, Dr. Martin Asher, compiled the following chart, demonstrating that the defendants participated in at least six across-the-board announced list price increases during the class period:
Effective
Type of
Bayer
Crompton/
DuPont
Exxon
Date
Elastomer
(Miles)
Uniroyal
DSM
Dow
Mobil
October/
Clear
$0.04/lb.
$0.04/lb.
$0.04/lb.
$0.04/lb.
$0.04/lb.
November 1997
Oil Extended
Increased commensurate with polymer content
$0.03/lb.-50 and/or 75 part oil extended $0.02/lb.- 100part oil extended
$0.04/lb. on all grades of Nordel EPDM
March/
Clear
$0.05/lb.
$0.05/lb.
$0.05/lb.
$0.05/lb.
$0.05/lb.
April 1998
Oil Extended
$0.03/lb.
$0.03/lb.
$0.03/lb.
$0.05 lb. on all grades of Nordel EPDM
$0.03/lb.
January/
Clear
$0.05/lb.
$0.05/lb.
$0.05/lb.
$0.05/lb.
$0.05/lb.
February 1999
Oil Extended
$0.04/lb.-medium oil extension $0.03/lb.-high oilextension
$0.03/lb.
$0.03/lb.
$0.05 lb. on all grades of Nordel EPDM
$0.03/lb.
January/
Clear
$0.05/lb.
$0.05/lb.
$0.05/lb.
$0.05/lb.
$0.05/lb.
February
($0.07/lb. for
($0.07/lb. for
2000
palletized)
palletized)
Oil Extended
$0.04/lb.-medium oil extension $0.03/lb.-high oilextension
$0.03/lb.
$0.03/lb.
$0.05 lb. on all grades of Nordel EPDM
$0.03/lb.
July 2000
Clear
$0.06/lb.
$0.06/lb.
$0.06/lb.
$0.06/lb.
$0.05/lb.
($0.07/lb. for
($0.07/lb. for
palletized)
palletized)
Oil Extended
$0.05/lb.-medium oil extension $0.04/lb.-high oilextension
$0.04/lb.
$0.04 lb.
$0.06 lb. on all grades of Nordel EPDM
$0.03/lb.
February/
Clear
$0.06/lb.
$0.06/lb.
$0.06/lb.
$0.06/lb.
$0.05/lb.
March
($0.07/lb. for
($0.06/lb. for
($0.07/lb. for
2001
palletized)
palletized)
palletized)
Oil Extended
$0.05/lb.-medium oil extension $0.04/lb.-high oilextension
$0.04/lb.
$0.04/lb.
$0.06 lb. on all grades of Nordel EPDM
$0.03/lb.
(a) Plaintiffs' Expert Reports
Source: Table 1-" Announced Price Increases,"Asher Report at ¶ 11.
Plaintiffs' experts, Drs. Martin Asher, John Beyer, and Keith Leffler, also explain precisely why the EPDM market was susceptible to collusive activity. Each expert concludes that the EPDM market was conducive to creation and implementation of an antitrust conspiracy to raise and/or maintain prices, and therefore, if the alleged collusive activity did occur-i.e., in the form of the six lock-step price increases-those structural characteristics make it more likely that the impact of those price increases would be felt class-wide.
Dr. Leffler is the opt-out plaintiffs' (Parker Hannifin Corporation, PolyOne Corporation, and The Goodyear Tire & Rubber Company) expert. Although those plaintiffs are not party to the class certification motion, I will include Leffler's analysis of the EPDM market because the defense experts address Leffler's analysis in an expert report submitted in this action.
Specifically, they assert that because of: (1) EPDM's commodity-like nature; (2) the lack of economic substitutes; (3) the product's price inelastic demand; (4) defendants' dominance of the market; (5) the barriers to market entry; (6) and defendants' use of national (as opposed to regional) price lists and price increase announcements, the EPDM market was susceptible to collusive behavior, which would have had a class-wide impact on EPDM purchasers. The experts interpret these factors as limiting the possibility of competition and making the EPDM market conducive to a price-fixing scheme. They lay the groundwork for plaintiffs' argument that, if collusive behavior did occur, it would have been effective in raising prices across the class, thus demonstrating class-wide injury-in-fact.
Market power. Over the relevant class period, the defendants controlled over 90% of the domestic EPDM market. Asher Report, ¶ 15; Report of Dr. John Beyer (" Beyer Report" ), Ex. 1; Report of Dr. Keith Leffler (" Leffler Report" ) ¶ 28. According to Beyer, the EPDM market can be described as an oligopoly-" a market with relatively few producers ... account[ing] for a large percent of the total market" -which is particularly susceptible to coordinated behavior. Beyer Report, ¶ 10; Leffler Report, ¶ 22. Under basic economic theory, the more concentrated market power is (i.e., a relatively small number of firms possess a large percentage of the market), the more easily firms can assert collusive power over production/supply and prices in a market. Beyer Report at ¶ 11.
Barriers to Entry. When industry profits are high, it encourages new firms to enter the market or existing firms to add capacity. Increased supply will eventually drive prices and profits down. However, where there are high barriers to entry, new firms will be dissuaded from entering the market, potentially leaving prices artificially high, resulting in more profits for existing firms. Id. at ¶ 12. Barriers to entry also create incentives for existing firms to raise prices out of line with the cost of inputs, knowing that the higher profits will not be enough to attract new firms. Examples of barriers to entry include: high investment costs and long production facility construction schedules.
The average cost to build a 100,000-ton EPDM plant is between $300 and $500 million, taking approximately three to five years to complete. Id. at ¶ 13; Reply Report of Dr. Martin Asher (" Asher Reply Report" ), Ex. 31 (Declaration of James R. McKinley), at ¶ 37-40. Furthermore, possessing the technology to produce EPDM requires proprietary information, which also raises a barrier to entry. Id. Beyer concludes these facts present high barriers to entry. Beyer Report at ¶ 13; Leffler Report at ¶ 33-38.
Fungible, commodity-like product. Drs. Asher, Beyer, and Leffler assert that EPDM is a commodity-like product, meaning it is easy to substitute one producer's EPDM for another's (i.e., " supply-side substitutability" ). This has two implications for the EPDM market: first, the defendants could easily compete over customers and, second, that price was the primary factor for purchasers in choosing the supplier from which to buy EPDM. Beyer Report at ¶ 26. Beyer relies on various internal documents from the defendants that show which grades and brands of their competitors' EPDM products were the equivalent of their own grades and brands. Beyer Report, Ex. 2; Asher Reply Report, Ex. 27. This demonstrates not only that the defendants were aware of the commodity-like nature of their product, but also that they could easily produce each other's products if they so desired. Leffler Report at ¶ 40-42. Without a collusive agreement to maintain and/or raise prices, it is likely that any EPDM supplier could have increased market share vis-a-vis its competitors by lowering or refusing to raise its own prices. To Beyer, this indicates a high probability of collusive potential in the EPDM market. Beyer Report at ¶ 16.
No viable economic substitutes & price inelastic demand. According to Beyer, EPDM has several unique characteristics that make it difficult for purchasers to substitute another product for EPDM. This keeps demand relatively inelastic-without other options, purchasers are essentially held hostage to price increases. Id. at ¶ 17. Knowing that purchasers will not turn elsewhere if the price increases, EPDM suppliers would be able to raise prices without losing demand, painlessly increasing their profits. Id. at ¶ 19-21; Asher Reply Report at ¶ 35-43, Exs. 32-33; Leffler Report at ¶ 30-32. Furthermore, the demand for EPDM is dependent primarily on the downstream demand for the end-use products that contain EPDM. As long as end-use consumers continue to demand those products, manufacturers of those products (the plaintiffs) would need to continue to purchase EPDM. Beyer Report at ¶ 21.
Price Lists. There were no regional differences in price, meaning the defendants' price lists applied to the national market. Beyer Report at ¶ 22. The defendants consistently raised prices in tandem, alternating which supplier was the first to raise prices. Id., Exs. 3-4, Asher Reply Report, Ex. 10. Plaintiffs allege this was the result of illegal collusive activity. If so, the price lists are proof common to the class, which demonstrate a class-wide impact on prices-" [t]he character and purpose of the announced price increases demonstrates that all purchasers of EPDM would have paid elevated prices." Beyer Report at ¶ 26.
Communication. Leffler focused on the high level of communications between competitors, noting the exchanges of emails on prices and capacity and numerous meetings held in the guise of industry association meetings and planning meetings for potential joint ventures, raw material purchases, and buy-sell arrangements. Leffler Report at ¶ 44. According to Leffler, " [c]ooperation creates an atmosphere of trust in which understanding rather than explicit agreements can be implemented." Id. at ¶ 43.
(b) Defendants' Expert Reports
Defendants' expert, Dr. Kenneth Elzinga, was charged with examining the structural characteristics of the EPDM market relied upon by plaintiffs' experts Beyer and Leffler. Report of Dr. Kenneth Elzinga (" Elzinga Report" ) at 17-34. At the outset, Elzinga states that he " agree[s], for the most part, that the factors on the Beyer-Leffler list are present in the market for EPDM." Id. at 19. Rather than attacking their analyses of the EPDM market, Elzinga takes issue with the idea that the EPDM market fits as neatly into the traditional oligopoly model as they suggest it does, and questions the relevancy of those structural characteristics where they are present. Id.
On the issue of commodity-like fungibility, Elzinga concedes that EPDM is " often substitutable across defendants." Id. However, he notes that many EPDM brands do not have corresponding competing products, suggesting that EPDM purchasers of those grades are locked into a single supplier. Essentially, Elzinga disputes that EPDM is as easily substitutable as the plaintiffs' experts claim.
On the issue of demand inelasticity, Elzinga argues that there are available EPDM substitutes, though he acknowledges that " [m]arket demand may tilt to being inelastic." Id. at 22. Plaintiffs' experts do not assert a complete absence of viable substitutes, but note that the available substitutes are not so interchangeable with EPDM that a price increase would cause significant numbers of purchasers to switch from EPDM to a substitute. Elzinga appears to agree with that point, but emphasizes that switching products is easier than Beyer and Leffler claim it is.
Moving to market concentration, Elzinga again agrees that the EPDM market is " characterized by a small number of sellers." Id. at 23. He takes issue, however, with the fact that the plaintiffs have not concentrated enough on ExxonMobil's participation in the alleged conspiracy, noting that they have not called any ExxonMobil witnesses. Id. The plaintiffs' experts did note ExxonMobil's market share when analyzing market concentration, so this point is not particularly relevant to the issue of market concentration. Elzinga also believes that Beyer and Leffler have not fully accounted for producers not named as defendants, which he claim hold a 16% stake in the global EPDM market. Id. at 24. Given that plaintiffs are arguing that there was a conspiracy to set prices in the domestic EPDM market, I am not persuaded that the global market share held by foreign EPDM suppliers is particularly relevant to this antitrust suit.
Elzinga does not disagree that the cost to build a new EPDM plant is high and that barriers to entry generally help cartel members to control prices. Id. at 25-26. Elzinga strives to point out that, despite these high costs, new and existing EPDM suppliers added capacity during the alleged conspiracy period, pointing to one new plant by a new EPDM supplier and one new plant by an existing EPDM supplier. Id. at 25. As he later notes in his report, however, one of those plants-the DDE Plaquemine plant-" struggled to produce acceptable material and to operate at capacity through 2001." Id. at 52-53. The other new plant that opened during the class period, the Seadrift Plant, never produced at rates higher than 15% through 2001. Id. at 54. Also, Elzinga notes that during the class period, several existing EPDM producers announced plans to add additional capacity, though I am not sure that announcing plans to open new plants during the class period would affect the plaintiffs' barriers to entry analysis. Those plants opened well after the alleged conspiracy period and one was in the Netherlands, not the United States. Id. at 26. Furthermore, the effect of the new capacity on the alleged conspiracy's ability to control prices above what they would otherwise have been is an issue for the merits. The bottom line is that Elzinga does not contest that the EPDM market has high barriers to entry, just that those barriers did not entirely thwart new capacity.
Finally, Elzinga takes strong issue with the plaintiffs' experts' lack of explanation regarding the mechanics of the cartel's operations. Elzinga disputes Leffler's " atmosphere of trust" theory, arguing that cartels do not operate in such a nebulous fashion, but actually require explicit agreements and enforcement between players to operate properly. Id. at 30-31. As a contrast to the EPDM scheme alleged by plaintiffs, Elzinga points to studied cartels in the vitamin, lysine, and citric acid industries. He notes that, in those industries, executives met frequently to discuss pricing of their products and that " none of these cartels ... left the functioning of their conspiracy up to a vague understanding that their members would like prices to be higher or complaints that prices were too low." Id. at 33. Elzinga notes that the few documents produced by the defendants suggesting the need to avoid price wars or make industry-wide lock-step price increases contrast with the level of " intense competition" demonstrated by the majority of documents produced in this litigation. Id. at 34. Elzinga believes that there is simply not enough evidence to show that EPDM producers met with sufficient frequency and with the required attendance by all major players to adequately form and manage a working cartel. Id. Beyer's response to this criticism is that there is enough documentation of industry-wide price increases to support the allegation of an explicit conspiracy. Beyer Response Report ¶ 9. This is an issue for the merits, because it touches on the ultimate question of the of the antitrust lawsuit-whether the defendants explicitly agreed to engage in price-fixing, thus violating the antitrust laws prohibiting collusion to stifle competition in the market.
My understanding is that the primary purpose of Elzinga's report on EPDM structural characteristics is not to show that the plaintiffs' experts have mischaracterized the dynamics of the EPDM market, but rather to assert that there is not enough evidence to prove that a conspiracy occurred. At this stage of the proceedings, however, the issue for the court is not whether the conspiracy actually occurred-that is the ultimate issue in the case and one for the factfinder to decide at the merits stage. The narrow issue at this juncture is whether the elements of an antitrust cause of action are demonstrable across the class on the basis of common proof, or whether individual questions predominate.
The other defense expert, Dr. Edward Snyder, disputes Asher's theory that the lock-step price increases can serve as common proof of impact. In his report, Snyder discusses how it is not necessarily true that the lock-step price list increases resulted in higher " but-for" prices for all class plaintiffs due to list price rebates, discounts, and non-price factors such as research and development, shipping costs, and technical services provided to customers in lieu of discounts. Although those issues may be relevant for calculating damages -those rebates, discounts, and non-price factors might mean that a particular plaintiff suffered no damages-they do not mean that the plaintiffs were not impacted in some way. Having a higher but-for price is one way of proving impact, but not the only way. Snyder's disagreements with Asher are primarily focused on the merits of whether the plaintiffs can ultimately demonstrate the three necessary elements of an antitrust claim on the merits. That is not the issue for this motion for class certification. I need only determine whether the plaintiffs have demonstrated that the issue of antitrust impact is susceptible to proof applicable to the whole class.
The fact that Elzinga supports the plaintiffs' characterization of the EPDM market is noteworthy: they make their point on structural characteristics primarily to show that the market was conducive to a conspiracy, i.e., that if those six price list increases indicate the workings of a conspiracy, then it had a good chance of being successful, not whether a conspiracy actually existed. Therefore, in light of the general agreement among the experts that the EPDM market was susceptible to a conspiracy, I find that the plaintiffs have met their burden of demonstrating that the element of injury-in-fact can be proved by evidence common to the class-the six national price lists that ostensibly applied to every customer, in conjunction with the general analysis of the characteristics of the EPDM market, support the notion that the price list increases had class-wide impact.
Pursuant to In re IPO, of course, these findings about the EPDM market are not binding on the ultimate fact-finder who would remain free to determine whether the price lists and structural characteristics demonstrate adequate classwide antitrust impact on the merits. I am merely finding that the plaintiffs have met their burden of demonstrating how proof common to the class is available to demonstrate the injury-in-fact element of an antitrust claim.
3. Damages
Proving damages is the final element of a successful antitrust claim, although courts have routinely held that class certification is not defeated by the need to undertake individualized damages determinations. See, e.g., In re Visa Check, 280 F.3d at 139-40. Nonetheless, the plaintiffs have offered the econometric multiple regression analysis of their experts, Asher and Beyer, to demonstrate that damages can be proven on a class-wide basis (and simultaneously demonstrate class-wide impact).
a. Methods and Standards
Multiple regression analysis is " a statistical tool for understanding the relationship between two or more variables." Daniel L. Rubinfeld, Reference Guide on Multiple Regression 181 (Federal Judicial Center, 2d ed.2000). The " dependent variable" is the variable to be explained (here, the price of EPDM) and the " independent variables" are factors that are thought to influence the dependent variable. Using multiple regression analysis allows economists to estimate whether a certain market factor has an effect on the dependent variable and to what degree. By determining whether a particular variable in the equation influences the dependent variable, the economist can accept or reject that variable as having an influence on the dependent variable. This statistical tool is used frequently in employment discrimination and antitrust lawsuits. For example, in employment discrimination suits, an employee may argue that sex or race is a statistically significant variable on the value of the dependent variable-the employee's salary. The Supreme Court has already accepted the use of multiple regression analysis in employment discrimination suits. See Bazemore v. Friday, 478 U.S. 385, 106 S.Ct. 3000, 92 L.Ed.2d 315 (1986). In an antitrust suit, plaintiffs will generally use multiple regression analysis to demonstrate that a " conspiracy" variable has influence over the dependent variable (price)-that is, class members paid a higher price than the basic economic principles of supply and demand would otherwise dictate, thus demonstrating collusive behavior was at work.
In a multiple regression analysis, the model must be correctly set up for the results to be valid. The expert's choices are crucial for the success of the analysis and the ultimate fact-finder should consider the following when evaluating the efficacy of the expert's analysis: has the expert correctly identified the dependent variable; has he or she chosen the correct explanatory variable that is relevant to the question at issue; are the additional variables chosen all correct or are some missing and/or irrelevant; is the form of the analysis correct? Rubinfeld, Reference Guide on Multiple Regression, at 186-88.
Notwithstanding the unique posture of this case-full merits discovery is essentially complete and the parties have submitted their expert reports, ordinarily presented for the first time at summary judgment or trial-what the court must decide at the class certification stage remains the same. Based on the standard of proof required by IPO Securities, the task of the district court at the class certification stage is to determine whether the defendants have successfully debunked the plaintiffs' claim that a class-wide approach can be used to prove impact and damages. Cordes, 502 F.3d at 107. The defendants should be focused on disputing the use of the methodology itself, not the results of the methodology. As the Supreme Court noted in Bazemore, the failure to include certain variables in a multiple regression analysis " will affect the analysis' probativeness, not its admissibility." 478 U.S. at 400, 106 S.Ct. 3000. In other words, if the defendants' experts are merely disputing the results of the plaintiffs' experts' analysis rather than the feasibility of using a single formula methodology, that would be a merits issue, not a class certification issue. If plaintiffs can present a feasible multiple regression analysis that purports to establish impact and damages using data and information common to the class, they will have met the requirement of demonstrating that damages can be proved by evidence common to the class.
In Cordes, the Second Circuit was primarily concerned with whether a single formula could be constructed, because of the individualized nature of the underwriting industry. 502 F.3d at 107. The district court identified the ten factors that the defendants' expert believed accounted for an IPO underwriter's fee: " 1) underwriter costs; 2) analyst coverage; 3) price stabilization and market making; 4) other underwriter compensation; 5) regulatory issues; 6) prospective issuer investment banking business; 7) the risk of the offering; 8) underwriter prestige and reputation; 9) share pricing; and 10) syndicate composition." In re Public Offering Fee Antitrust Litig., 2006 WL 1026653, at *6-7, 2006 U.S. Dist. LEXIS 21076, at *22 (S.D.N.Y.2006), vacated by Cordes, 502 F.3d at 109. As the defense expert explained, " [b]ecause underwriting services are not fungible products but, rather, are customized by differentiated underwriter suppliers for each issuer ... an assessment of whether a particular issuer's net proceeds were reduced by the alleged conspiracy involves an individualized inquiry based on non-common factors that will vary from one issuer and IPO to the next." Public Offering Fee, 2006 WL 1026653, at *6-7, 2006 U.S. Dist. LEXIS 21076, at *21. That expert claimed that the multitude of factors accounting for an underwriter's fee made it impossible to create a working multiple regression model. Id. at 2006 WL 1026653, at *9, 2006 U.S. Dist. LEXIS 21076, at *30. The plaintiffs' expert claimed that he could create a working multiple regression model, despite acknowledging that there are at least six factors that accounted for an underwriter's fee (risk of the offering, quality and reputation of the underwriter firm, whether the IPO involved the privatization of a government-owned entity, whether the initial stock price was relatively low, the year the IPO was conducted, and the type of IPO process used). Id. at 2006 WL 1026653, at *6-8 n. 7, 2006 U.S. Dist. LEXIS 21076, at *22 n. 7, *26. The plaintiffs' expert never presented an actual equation and consequently the defendants did not offer specific methodological criticisms beyond the claim that no such formula could be created. The district court accepted the defense expert's report, which focused on the relevant issue-whether antitrust impact can be proven by evidence common to the class-whereas the plaintiffs' expert report was focused on whether damages could be proven on a class-wide basis. Id. at 2006 WL 1026653, at *8-9, 2006 U.S. Dist. LEXIS 21076, at *27-28. On appeal, the Second Circuit noted that the two experts' analyses were actually directed to the same question, " whether injury-in-fact is susceptible to common proof in this case," and concluded that, if such a formula could be constructed, " the injury-in-fact question is common to the class. Otherwise, it poses individual ones." Cordes, 502 F.3d at 107. Therefore, the Court remanded the case to the district court to determine " which expert is correct," i.e., whether a single formula can be created, or whether there really were too many factors at work for a multiple regression model to be methodologically sound. Id.
Before proceeding further, I must note several key points. First, the plaintiffs in Cordes were depending on their multiple regression models to prove class-wide injury-in-fact and damages. The plaintiffs here present not only an econometric analysis to prove damages (which also demonstrates injury-in-fact) but have presented six lock-step national price list increases for a fungible product in an industry with structural characteristics supporting the efficacy of collusive activity. Without that type of evidence, the Cordes plaintiffs were necessarily dependent on their econometric model to meet the requirements of Rule 23. Second, unlike an underwriter fee that necessarily depends on a combination of highly individualized factors, EPDM is a commodity-like product whose price results from only a small number of variables that are readily determined using publicly available data. Therefore, the plausibility of a multiple regression analysis in the EPDM context is necessarily higher than in the IPO underwriting context. In Cordes, the Second Circuit was not asking the district court to assess the efficacy of the plaintiffs' modeling on the merits, but rather to determine why, given the individualized nature of the IPO underwriting industry as acknowledged by both sides' experts, a single formula could nonetheless be constructed. As discussed more fully below, the biggest problem for the EPDM defendants is that, although they claim to be attacking the methodology employed by the plaintiffs' experts, they are, in practice, attacking the multiple regression analysis on the merits by challenging the results of the modeling rather than the conception of the modeling.
b. Plaintiffs' Economic Experts
Plaintiffs seek to show that class-wide damages can be proven by relying, in part, on Beyer's econometric modeling. Beyer's model is a " reduced-form multiple regression equation, i.e., a single equation that takes into consideration factors of both supply and demand, as well as the impact of the alleged conspiracy." Beyer Report at ¶ 28. Beyer next identifies a class period: October 1997 to December 2001. Id. at ¶ 29. Beyer explains that it is during the latter portion of 1997 that the coinciding price increases became most apparent. Id. at ¶ 29-31. By 2002, however, Beyer's analysis of the market indicates a price war had begun, thus indicating an end to collusive dealings among EPDM producers. Id. at ¶ 31. Having identified the class period, Beyer explains that he will compare prices during the class period to EPDM prices after the class period (aka the " benchmark period" ). Id. at ¶ 32. He notes that, because he believes that collusive dealings began even before 1997, pre-class period pricing would offer an unacceptable benchmark for comparison. Id.
Beyer's regression technique is described as " Ordinary Least Squares" (" OLS" ) and uses " fixed effects." Id. at ¶ 38. Beyer explains that using " fixed effects" enables the considerations of customer buying power, negotiating strength, and the like to be taken into consideration. Id.
Beyer's reduced form equation is: " PEPDM = † (DEPDM, SEPDM, Cons)" -the price of EPDM is a function of the demand for EPDM, the supply of EPDM, and the conspiracy. Id. at ¶ 36. Beyer conjectures that the supply of EPDM is a function of the cost of its inputs. Id. at ¶ 33. Identifying the primary inputs for the production of EPDM-ethylene, propylene, energy, and labor-Beyer states that, all other things being equal, he would expect that an increase in the cost of the inputs will increase the price of EPDM. Id. at ¶ 34. Beyer calculated the supply variable by multiplying each input's percentage of overall input cost for EPDM by price indices collected from the U.S. Bureau of Labor Statistics or Platts and then adding all the results together. Id. at ¶ 40.
" PEPDM" stands for the price of EPDM, " DEPDM" stands for the demand of EPDM, " SEPDM" stands for the supply of EPDM. " Cons" stands for the conspiracy variable.
According to Beyer, Platts is a " leading global supplier of energy and metals related information." ¶ 40 n. 56.
The DEPDM variable is determined by the demand in industries that use EPDM in their products. Id. at ¶ 35. As demand for those downstream products increases (or decreases), the demand for EPDM increases (or decreases). Id. As the demand increases (decreases), basic economic theory dictates that the price of EPDM would increase (decrease) as a result. " The demand index was created by multiplying the percentages for each of the individual components [shown in Exhibit 6] by its respective industrial output index as reported by the U.S. Federal Reserve Bank." Id. at ¶ 41.
The Cons variable becomes relevant in the equation during the class period by taking on a value of " 1" between October 1997 and December 2001 and " 0" at all other times. Id. at ¶ 36. The " Cons" variable is Beyer's " dummy" variable, that is, it captures all unexplained impact on the price of EPDM in a single variable. Beyer concludes that the " regression analysis shows that, while supply and demand factors were at work in determining the price of EPDM, they do not alone explain the price of EPDM during the damage period," meaning collusive activity must also have been at work. Id. at ¶ 44.
c. Defendants' Economic Experts
Essentially, the defendants' experts' critique of Beyer's methodology centers on the one-size-fits-all nature of his dummy variable, which they argue is too indefinite to capture any true data because key variables are missing. The defendants attack plaintiffs' experts' conclusions, relying primarily on Dr. Christopher James's expert report, which purports to demonstrate the fallacy of Beyer's multiple regression analysis calculations. Although James claims to show how Beyer's regression is methodologically flawed, an examination of his report reveals that his primary dispute is with the conclusions that Beyer draws from his calculations, not whether a single formula can be used to estimate class-wide damages. James's report does not appear to be disputing that the plaintiffs would be able to use common proof, i.e., a single formula, to show that there was a classwide impact or damages. Rather, he appears to dispute the efficacy of that calculation by arguing that Beyer has omitted key variables that would explain the positive value of his " conspiracy" variable. Whether the price of EPDM increased disproportionately to the cost of inputs as the result of a conspiracy to raise/maintain prices, or instead resulted from a non-collusive cause, is a question on the ultimate merits of the case.
Rather than dispute that it is not feasible to measure antitrust damages or impact in the way Beyer proposes, James disputes the accuracy of his calculations. He criticizes how Beyer calculated the demand and supply variables. James says that Beyer: (1) failed to properly calculate the variables used in the demand index, (2) failed to properly account for changes in the structure of demand from the downstream industries-i.e., that it cannot account for such factors as " changes such as long-term substitution between materials or changes in the manufacturing of downstream applications that might increase or decrease their reliance on EPDM," (3) has not captured changes in the buyers' bargaining power over the course of the conspiracy, (4) has not properly accounted for changes in U.S. consumption through 2000, and (5) makes findings inconsistent with the economic theory that cartels in general should not lead to both higher prices and higher demand. Id. at ¶ ¶ 23-30. James argues that Beyer's supply index variable is also inaccurate because it does not account for capacity-related effects on supply. Id. at ¶ ¶ 31-32. James contends that higher capacity utilization over the alleged class period is what caused the higher prices for EPDM, not a conspiracy. Id. at ¶ 32. James claims to test the Beyer model's predictive ability outside the alleged conspiracy period, when the conspiracy variable would hypothetically be statistically insignificant, and finds that even outside the class period Beyer's model shows unexplained influences on EPDM pricing. Id. at ¶ 34.
In their reply brief, defendants acknowledge that in order to defeat class certification they must demonstrate that " Beyer's own reduced form model is not capable of answering the question of antitrust injury on a classwide basis." Defendants' Reply at 3. Even accepting the structure of Beyer's model, defendants contend that they can demonstrate that nearly half of the class members paid lower prices than they would have in Beyer's hypothetical non-conspiratorial world. Declaration of Christopher James (attached to Defendant's Reply), at ¶ 4, 6. To show this using Beyer's model, James created a customer-specific conspiracy dummy for each customer in Beyer's dataset that purchased EPDM during and after the class period. James Declaration, at ¶ 5. He then replaced Beyer's single conspiracy variable with his new customer-specific variables, known in econometrics as " interaction effects." Id. With interaction effects, each customer's average price outside the class period acts as the benchmark against which its average price during the alleged conspiracy period is measured. The customer-specific conspiracy dummy variable accounts for the difference between these two averages. Id. Using that method to test whether there was a uniform effect across EPDM customers, James found that the effects varied widely and therefore, contrary to Beyer's contention that the alleged conspiracy had uniform effects, no common damages can be proven. James Report, ¶ 46. Therefore, the defendants argue, plaintiffs have not met their burden of establishing that common questions predominate.
James then goes on to construct his own model using class-wide proof that he claims better demonstrates what factors were affecting the price of EPDM during the alleged class period. Calling this model a " structural model" of supply and demand, James says it better accounts for the " joint determination of price and capacity utilization in the marketplace," which the plaintiffs' reduced form model does not adequately account for. Id. at ¶ 41. James explains that Beyer's reduced form model erroneously conflates the effects of supply and demand factors on the dependent variable, without isolating the extent to which those factors affected price.
Using a two-step analysis known as the " two-stage least squares" technique, James first estimates the effect of plaintiffs' experts' variables on capacity utilization (the first stage's dependent variable) in order to control for the effect of demand on capacity utilization. James adds in the result of the first stage calculation as the variable in the second stage, making price the new dependent variable. Any remaining unexplained supply-side effect could be attributed to a hypothetical conspiracy variable. Id. at ¶ 42. James says the results of his model show that price was barely higher during the conspiracy period and that the conspiracy variable (once he controls for his additional variables) is statistically insignificant. Thus, there was no conspiracy. Id. at ¶ 43.
In summary, James concludes that, contrary to Beyer's report, he can prove that nearly half of the class members paid lower prices during the conspiracy than in Beyer's " but for" hypothetical, non-conspiratorial world. Therefore, James concludes, there is no way of demonstrating class-wide impact or damages due to the conspiracy. James purports to offer another model for demonstrating effects on the class, which the plaintiffs in turn challenge.
Defendants also rely on Elzinga's report to refute plaintiffs' claim of proof of classwide impact. The pertinent portion of Elzinga's report is focused on substantive arguments about whether or not a conspiracy existed in the EPDM market, which is not the question before the court at the class certification stage. For instance, Elzinga devotes whole sections of his report to discussing reasons why the six price increases were the result of unilateral increases, not collusion (Section VI Part C.2); offering a variety of explanations for several plants' low output to explain that EPDM producers did not collude to limit supply of EPDM, (Section VII); and arguing that, contrary to the plaintiffs' allegations in the complaint, there was no coordination among EPDM producers for allocating customers. Elzinga cites evidence that EPDM purchasers switched between various suppliers and that market shares among EPDM suppliers were not consistent over the class period (Section VIII). These arguments all go to the ultimate issue whether there was a conspiracy among EPDM producers to raise and/or maintain EPDM prices above competitive levels. Although Elzinga's report sets forth valid arguments on the merits, those arguments are not relevant when determining whether the plaintiffs have met their burden to demonstrate all the Rule 23 requirements for class certification.
Elzinga's focus on the plaintiffs' multiple regression analysis comes in Section X of his report, where he faults Beyer and Leffler for failing to include a provision controlling for capacity in their cost input variable. Elzinga Report, at 70. Elzinga argues that Beyer and Leffler " both ignore the effects of capacity on prices, thereby undercutting their analysis of pricing trends and their attempt to identify the effect of an alleged cartel on prices." Id. at 71. Elzinga's analysis focuses on the role of the Plaquemine and Seadrift plants before and after the class period. Those plants did not open when planned, so EPDM producers experienced significant capacity constraints throughout the class period. As a result, existing plants ran at or near capacity, which put an upward pressure on prices. Id. at 71-75. Price pressure eased when new capacity finally came on line, a period that coincided with the end of the class period in 2001. Id. at 76. " The surge in productive capacity and resulting decrease in prices coincides with what Plaintiffs claim is the end of the alleged cartel period. But by elementary economic logic, prices fell because of increased supply, not the end of the alleged cartel." Id.
Next Elzinga debunks Beyer and Leffler's demand variable, arguing they have omitted " certain aspects of demand" and overall that the index " appears out of step with actual trends in industry demand." Id. at 78. First, Elzinga notes the steady growth of EPDM sales between 1997 and 2000. By 2001, however, the demand for EDPM had dropped due to the economic recession precipitated by the terrorist attacks on September 11, 2001. Elzinga says Beyer and Leffler mistakenly assign the drop in prices to the end of the cartel period, rather than to an economy-wide recession. Id. at 81. Elzinga asserts that Beyer and Leffler's demand index erroneously indicates that demand was higher in 2002 than it was in 2001, when in reality price and quantity data indicate that demand fell in 2002 from 2001. Beyer and Leffler's failure to accurately reflect a drop in demand, Elzinga argues, calls into question the accuracy of their demand index. Id. at 82.
d. Discussion
The defendants, I believe, have misconstrued the extent of the " merits" inquiry that a court must make under the class certification analysis required by In re IPO. The defendants and their experts have focused their opposition to the motion for class certification primarily on the merits of the plaintiffs' antitrust claims. Although characterized as a dispute over the very feasibility of plaintiffs' analysis, defendants are actually arguing that plaintiffs' multiple regression analysis, done a slightly different way (i.e., the " right" or " better" way), does not prove what they claim it proves, classwide damages. The defendants claim the model, done slightly differently, demonstrates that at least half the class suffered no damages and thus plaintiffs have not presented a feasible method for proving classwide damages. In essence, the defendants are asking the court to determine which multiple regression model is most accurate, which is ultimately a merits decision.
The real question before this court is whether the plaintiffs have established a workable multiple regression equation, not whether plaintiffs' model actually works, because the issue at class certification is not which expert is the most credible, or the most accurate modeler, but rather have the plaintiffs demonstrated that there is a way to prove a class-wide measure of damages through generalized proof. The defendants contend that James has successfully refuted the efficacy of Beyer's model to show class-wide damages (and injury). It seems to me, however, that rather than refuting the efficacy of Beyer's model, defendants are disputing the results by manipulating Beyer's model to show that no common impact can be shown and then, in the alternative, proffering their own model using common class-wide data to show there is not a class-wide impact. Beyer contests James's use of individual customer-conspiracy variables; James claims Beyer's model is flawed because Beyer does not use that type of variable. Each has arrived at a polar opposite finding concerning common impact to the class of EPDM purchasers, but significantly both employ single formulas.
Certainly, In re IPO heightened the plaintiffs' burden on a class certification motion. No longer may plaintiffs slide by the class certification stage on unsupported legal conclusions or merely plausible expert methodologies. Even In re IPO, however, does not require plaintiffs to prove the merits of their case-in-chief at the class certification stage. They need not demonstrate that their multiple regression analysis captures all the proper variables and thus reaches the " right" answer, as the defendants would require them to do. The defendants have failed to convince me that it is methodologically impossible to use a single formula to estimate class-wide damages. It is unnecessary to delve further into the merits by going point-by-point through each expert's theory to decide who has designed the " better" multiple regression equation.
A brief analysis of the facts of In re IPO reveals how the heightened standard is intended to work in practice. In that case, the Second Circuit vacated the district court's grant of class certification, holding that the plaintiffs' allegations and evidence could never meet the Rule 23 requirements as a matter of law. In re IPO, 471 F.3d at 42. Although the district court accepted that the plaintiffs had established the presence of an " efficient" market, enabling them to invoke the " fraud-on-the-market" theory to show that common questions of reliance would predominate over individual questions, the Second Circuit did not accept that conclusion on its face. Id.
The " fraud-on-the-market" doctrine established in Basic Inc. v. Levinson, 485 U.S. 224, 245-37, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988), creates a rebuttable presumption that (1) an issuer's misrepresentations will affect the price of securities that are traded on the open market, and (2) investors rely on market price to reflect an accurate value for that security. Id. (citing Hevesi v. Citigroup, Inc., 366 F.3d 70, 77 (2d Cir.2004)).
Under the " some showing" standard, the district court accepted that the plaintiffs had satisfied their burden of demonstrating that the fraud-on-the-market presumption was applicable in their case. Id. Upon closer examination, however, the Second Circuit determined that it would be impossible for the plaintiffs to ever establish the fraud-on-the-market theory because there is no developed market for newly offered securities and thus the market for initial public offerings (" IPOs" ) is not an " efficient" market by definition. Id. Furthermore, the Court concluded that the plaintiffs' own allegations that the market was slow to correct the collusive price inflation, despite widespread knowledge of the scheme, belied the essential theory of how an " efficient" market, which relies on price to reflect accurate value, is expected to work. Id. at 43.
The Second Circuit concluded that, in the absence of the fraud-on-the-market presumption, individual questions of reliance necessarily would predominate over common questions. Id. Because the plaintiffs could not show that they would ever meet the Rule 23 requirement of showing that common questions of reliance would predominate over individual questions as a matter of law, class certification was not warranted. Id. Even though determining whether the fraud-on-the-market theory applied was a necessary inquiry on the merits, it was also an essential stepping-stone for the plaintiffs' predominance showing under Rule 23(b)(3). Therefore, In re IPO established that a district court is required to parse the plaintiffs' legal theory on the merits where it overlaps with the threshold inquiry of Rule 23(b)(3) predominance.
Here, the defendants are not arguing that any legal theory that the plaintiffs are relying on to demonstrate predominance is inapplicable as a matter of law. They are instead claiming that plaintiffs' common proof does not sufficiently demonstrate the merits of their antitrust claim. In Cordes, the plaintiffs were attempting to prove both impact and damages through the use of multiple regression analysis. 502 F.3d at 96-97. Unlike the EPDM plaintiffs, however, the Cordes plaintiffs did not have, in addition to their econometric modeling, common proof in the form of six national lock-step price list increases, in combination with largely unchallenged structural economic characteristics that support the efficacy of such lock-step increases. The Cordes defendants' expert contested whether a single formula could be constructed to determine whether a particular plaintiff would have paid a lower " but-for" underwriter fee. Id. at 106. The defense expert believed that no such formula could be constructed because calculating the " but-for" fee, against which the actual fee paid could be compared, required " an individualized, plaintiff-by-plaintiff analysis of ten factors including underwriter costs, price stabilization, and the risk of the offering." Id. at 97. The Second Circuit remanded the case in order for the district court to determine " which expert is correct," that is, whether a single formula could or could not be employed to make a valid comparison between the but-for and actual fee paid. Id. at 107. I do not understand the Second Circuit to be asking the district judge to determine which expert's report is more persuasive on the merits, but rather which expert is correct about whether or not the plaintiffs' method of proof is a form of common evidence.
The same problem is not present in this case. I understand the defendants to be contesting the merits of the plaintiffs' analysis, that is, that their econometric model does not properly account for the correct variables, and that a model that incorporates the correct factors will show that at least half the class suffered no damages. Whether the plaintiffs' multiple regression analysis incorporates the necessary variables/factors to reach the correct economic conclusion is an issue to be reserved for the merits, because it has no bearing on whether the plaintiffs can meet the predominance prong or Rule 23(b)(3) by establishing that common proof of impact is available in this case. " While the omission of variables from a regression analysis may render the analysis less probative than it otherwise might be, it can hardly be said, absent some other infirmity, that an analysis which accounts for the major factors ‘ must be considered unacceptable as evidence of discrimination.’ Normally, failure to include variables will affect the analysis' probativeness, not its admissibility." Bazemore, 478 U.S. at 400, 106 S.Ct. 3000 (internal citations omitted, quoting the appellate court ruling).
In Hnot v. Willis Group Holdings, Ltd., District Judge Gerald E. Lynch considered and rejected a similar interpretation of In re IPO . 241 F.R.D. 204, 210 (S.D.N.Y.2007). Where the plaintiffs and defendants disagreed about whose experts' statistical findings were more persuasive, Judge Lynch noted that:
this disagreement is relevant only to the merits of plaintiffs' claim ... and not to whether plaintiffs have asserted common questions of law or fact. By asking the Court to decide which expert report is more credible, defendants are requesting that the Court look beyond the Rule 23 requirements and decide the issue on the merits, a practice In re IPO specifically cautions against.
Id. Cf. In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305 (3d Cir.2008) (vacating class certification order on ground that the district court did not undertake the necessary rigorous analysis of the parties' experts' opinions and remanding with the direction that the district court should resolve the dispute between the experts whether impact was susceptible to class-wide proof or not). In any event, regardless of the efficacy of the plaintiffs' econometric modeling to show impact and/or damages, the plaintiffs have presented factual evidence of the allegedly collusive activity-the six lock-step price increases and the EPDM market's structural characteristics-that would be common to all plaintiffs. Common questions predominate where " even if each Class Member ... were to bring suit individually, each plaintiff would have to allege and prove virtually identical facts." In re Merrill Lynch & Co. Research Reports Sec. Litig., 246 F.R.D. 156, 165 (S.D.N.Y.2007). That is certainly the case here; a primary component of an individual EPDM plaintiff's case would revolve around proving how the price list increases are evidence of illegal collusive activity.
The facts of Hydrogen Peroxide are distinguishable from the present case. At the class certification stage, the plaintiffs' expert-the same Dr. Beyer employed by plaintiffs in this case-merely asserted that he could create an economic model using proof common to the class, but never actually completed his proposed models/analyses. Hydrogen Peroxide, 552 F.3d at 313, 315. The defendants' expert, on the other hand, contended that Beyer's modeling would fail. Id. at 313-14. In this case, Beyer has actually presented an econometric model that purports to show that antitrust impact is susceptible to common proof. Although the defense experts claim to dispute the feasibility of constructing an econometric model using proof common to the class, their reports are better characterized as disputing the results of the plaintiffs' modeling. To resolve this dispute would be to place myself in the role of the ultimate factfinder by choosing which expert's econometric model or theory is " correct." Therefore, having presented a working econometric model that demonstrates class-wide impact and injury using proof common to the class, the plaintiffs here have gone further than the plaintiffs in Hydrogen Peroxide in establishing that the requirements of Rule 23 have been met.
Furthermore, even if the defendants are correct that the measure of damages is not susceptible to classwide proof, that is not enough to defeat class certification. McLaughlin, 522 F.3d at 231. " [T]he fact that damages may have to be ascertained on an individual basis is not, standing alone, sufficient to defeat class certification." Id. (citing In re Visa Check, 280 F.3d at 140). " Common issues may predominate when liability can be determined on a class-wide basis, even when there are some individualized damages issues." In re Visa Check, 280 F.3d at 139. See also In re Scrap Metal Antitrust Litig., 527 F.3d 517, 535 (6th Cir.2008) (affirming district court's grant of class certification where " the court found that the ‘ fact of damages' was a question common to the class even if the amount of damages sustained by each individual class member varied." ); Urethane II, Polyether Polyol, 251 F.R.D. at 639 (" The possibility that individual issues may predominate on the issue of damages, however, does not defeat class certification by making this aspect of the case unmanageable." ); In re Carbon Black Antitrust Litig., 2005 WL 102966, at *20 (D.Mass.2005) (" Even if the damage assessments end up being too individualized to resolve as a class, that does not warrant denial of the plaintiffs' motion at this stage." ). In the present case, it may very well be that damages will need to be calculated individually, particularly if some plaintiffs received sufficient rebates, discounts, and other quantifiable favored treatment to mitigate all damages from the defendants' allegedly illegal, collusive behavior.
The defendants have made a strong argument that it may be impossible to calculate damages on a class-wide basis, that there is no single percentage of overcharge that can be applied to the class as whole. It may be that a customer-by-customer examination is necessary to determine pricing provisions of long-term contracts, negotiated rebates, and other discounts off the artificially inflated list price. However, those pricing incentives may have been available even in the absence of a conspiracy, to lower prices below even the non-inflated competitive level. It could be that industry custom or course of dealing between a particular supplier and purchaser dictated those types of rebates and discounts even in the absence of supra competitive pricing. Thus, it seems to me that individual questions of damages could predominate over common questions.
That is not enough to defeat class certification, however, on the issue of liability. The Second Circuit has suggested at least five ways to address individualized damages issues:
(1) bifurcating liability and damage trials with the same or different juries; (2) appointing a magistrate judge or special master to preside over individual damages proceedings; (3) decertifying the class after the liability trial and providing notice to class members concerning how they may proceed to prove damages; (4) creating subclasses; or (5) altering or amending the class.
Id. at 141 (internal footnote omitted). Because the defendants have presented a strong argument that damages may need to be determined on an individual basis, one or more of these alternatives may be appropriate in this case.
For the foregoing reasons, I find that the plaintiffs have met their burden of demonstrating how common questions will predominate over individual ones on the three elements of an antitrust claim. C. Superiority
The final Rule 23 requirement the plaintiffs must meet is that a class action is superior to other methods of adjudicating this dispute. The plaintiffs contend that individual lawsuits would be duplicative and a waste of judicial resources given the large number of affected EPDM purchasers with overlapping legal claims that arise out of the same course of conduct. Furthermore, they argue that the relatively small amount of damages that each class member stands to recover makes individual adjudication financially impractical for most class members who could not afford to mount a full-scale antitrust suit against the defendants on their own. Finally, given the similarity in claims and ease of determining members of the class, the plaintiffs assert that a class action here would comply with Rule 23(b)(3)'s directive that it be a superior method for the fair and efficient adjudication of the claims. The defendants argue that it would be more efficient to litigate the claims on an individual basis because individual questions of impact and damages predominate over common questions.
Having found that the plaintiffs have successfully demonstrated that common liability questions predominate over individual ones, I also find that a class action is a superior method of adjudication. A risk of inconsistent results and conservation of judicial resources dictate that this case should be litigated as a class action, at least on the issue of liability. Considering that a large number of class plaintiffs would have little or no financial incentive to prosecute the alleged antitrust violation individually, class certification is clearly a superior method of adjudication because it is most likely to promote fairness and efficiency.
IV. Conclusion
For the foregoing reasons, the plaintiffs' motion for class certification (doc.# 184) is GRANTED.
It is so ordered.
(A) the class members' interests in individually controlling the prosecution or defense of separate actions; (B) the extent and nature of any litigation concerning the controversy already begun by or against class members; (C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and (D) the likely difficulties in managing a class action.