Kleen Products LLC v. International Paper

306 F.R.D. 585, 2015 U.S. Dist. LEXIS 100466, 2015 WL 2330068

• Court: District Court, N.D. Illinois
• Decided: March 26, 2015
• Docket: Case No. 10 C 5711
• Status: Published

Opinion

MEMORANDUM OPINION AND ORDER

HARRY D. LEINENWEBER, District Judge.

Before the Court are Defendants’ Motion to Strike [ECF No. 845], and Plaintiffs’ Motion for Class Certification [ECF No. 657.] These Motions have resulted in a deluge of briefing; the Class Certification Motion alone spawned seven separate briefs that total more than 300 pages (not including the attached exhibits) and that include two sur-replies and several notices of supplemental authority. The Court has rigorously analyzed all of the parties’ submissions, and for the following reasons, Defendants’ Motion to Strike [ECF No. 845] is granted in part and denied in part, and Plaintiffs’ Class Certification Motion [ECF No. 657] is granted.

I. BACKGROUND

Plaintiffs are a proposed class of entities that directly purchased Containerboard Products from Defendants. Containerboard Products include containerboard itself and the various products made out of container-board, such as containerboard sheets, which are used to make corrugated products like displays, boxes, and other containers. Plaintiffs allege that Defendants engaged in a conspiracy to artificially manipulate the market in order to increase the price of Contain-erboard Products in violation of .antitrust laws. See, 15 U.S.C. § 1. The crux of Plaintiffs’ Complaint is that Defendants agreed “to restrict the supply of containerboard by cutting capacity, slowing back production, taking downtime, idling plants, and tightly restricting inventory.” [Pl.’s Mot. for Class Cert, at 1, ECF No. 660.] According to Plaintiffs, these actions illegally increased the price of containerboard, which caused them to pay more for Containerboard Products than they otherwise would have paid absent the conspiracy.

Plaintiffs seek to certify as a class:

All persons that pm-chased Containerboard Products directly from any of the Defendants or their subsidiaries or affiliates for use or delivery in the United States from at least as early as February 15, 2004 through November 8, 2010.

The proposed class definition also excludes certain groups from being class members:

Specifically excluded from this Class are the Defendants; officers, directors, or employees of any Defendant; any entity in which any Defendant has a controlling interest; and any affiliate, legal representative, heir or assign of any Defendants. Also excluded from this Class are any federal, state or local governmental entities, any judicial officer presiding over this action and the members of his or her immediate family and judicial staff, and any juror assigned to this action.

Defendants oppose certification, arguing that Plaintiffs have not satisfied Rule 23.

II. LEGAL STANDARD

“To be certified, a proposed class must satisfy the requirements of Federal Rule of Civil Procedure 23(a), as well as one of the three alternatives in Rule 23(b).” Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 811 (7th Cir.2012). Rule 23(a) requires Plaintiffs to prove “numerosity, typicality, commonality, and adequacy of representation.” Id. Plaintiffs in this case seek certification under Rule 23(b)(3), which also requires them to prove that: (1) the questions of law or fact common to the members of the proposed class predominate over questions affecting only individual members; and (2) a class action is superior to other available methods of resolving the controversy. Id.

Plaintiffs bear the burden of satisfying Rule 23, which is not “ ‘a mere pleading standard.’ ” Comcast Corp. v. Behrend, — U.S.-, 133 S.Ct. 1426, 1432, 185 L.Ed.2d 515 (2013) (quoting Wal-Mart Stores, Inc. v. Dukes, — U.S.-, 131 S.Ct. 2541, 2551-52, 180 L.Ed.2d 374 (2011)). To meet this burden, Plaintiffs must “satisfy through evi-dentiary proof’ each of Rule 23’s elements. Id. In deciding a class certification motion, the Court must conduct a “rigorous analysis” before it can determine whether Plaintiffs have satisfied Rule 23’s requirements. Id. (internal quotation marks omitted). This often means that a Court must resolve issues that also bear on the merits of the claim, but only if those issues overlap with class certification issues. Id.

Despite the need for rigorous analysis, “the court should not turn the class certification proceedings into a dress rehearsal for a trial on the merits.” Messner, 669 F.3d at 811. Instead, the Court need only consider the evidence submitted by the parties and determine whether Plaintiffs have proven each of Rule 23’s elements by a preponderance of the evidence. Id.

III. ANALYSIS

There are two preliminary issues the Court must address. First, the Seventh Circuit has held that “[w]hen an expert’s report or testimony is ‘critical to class certification,’ ... a district court must make a conclusive [Daubert ] ruling on any challenge to that expert’s qualifications or submissions before it may rule on a motion for class certification.” Id. at 812 (quoting Am. Honda Motor Co. v. Allen, 600 F.3d 813, 815-16 (7th Cir.2010)). Expert reports in this case are indeed critical to class certification, but no Defendant has yet challenged Plaintiffs’ experts on Rule 702 or Daubert grounds. To the contrary, Defendants have “expressly reserved their right to move to exclude [Plaintiffs’ experts] under Daubert and Rule 702.” [Def.’s Mem. in Opp. to Pl.’s Mot. for Class Cert. (“Def.’s Opp. Br.”) at 40 n.35, ECF No. 763.] Although Defendants vigorously challenge Plaintiffs’ experts’ methodology and conclusions in the context of arguing that Plaintiffs’ have not satisfied Rule 23, none of those arguments are based on Rule 702 or Daubert. Defendants have not challenged, for example, Plaintiffs’ experts’ education or qualifications. The Court therefore reserves ruling on Plaintiffs’ experts’ admissibility until Defendants raise and brief that issue.

Second, Defendants seek a full evidentiary hearing prior to the Court deciding whether to certify the class. Plaintiffs' oppose such a hearing, arguing that it is unnecessary and would waste time and money. Several courts have held evidentiary hearings prior to deciding a class certification motion, see, e.g., In re Groupon, Inc. Sec. Litig., No. 12 C 2450, 2014 WL 2035853, at *2 (N.D.Ill. May 16, 2014), but as far as the Court is aware, such hearings are not required. Rather, the Supreme Court and Seventh Circuit have admonished district courts not to simply accept Plaintiffs’ pleadings as true and to conduct a “rigorous analysis” of the Plaintiffs’ class certification claims. See, Comcast Corp., 133 S.Ct. at 1432 (“Repeatedly, we have emphasized that it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question____”) (internal quotation marks omitted). As stated above, the parties have submitted an avalanche of briefing and opposing expert reports that set forth the parties’ positions on the issues. Included in this briefing are thousands of pages of documents substantiating the parties arguments. Moreover, the parties’ central dispute is legal, not factual. The dispute centers mainly on the proper legal standard under Rule 23 and whether Plaintiffs’ experts’ reports are enough to satisfy that standard. For the most part, the parties agree on the basic facts, and both parties’ experts rely upon the same data, so there are little if any factual disputes that the Court must resolve to decide class certification. Given the extensive paper record and the completeness of the parties’ briefing, an evidentiary hearing would not add much to the Court’s analysis. Thus, the Court is confident that it can determine class certification based on a careful examination of the evidentiary record the parties have submitted.

Having resolved those threshold issues, the Court must first decide Defendants’ Motion to Sti’ike some of the materials Plaintiffs submitted with their reply brief. Once that issue is decided, the Court can then determine whether Plaintiffs have satisfied Rule 23. All Defendants have joined in a single response to Plaintiffs’ Class Certification Motion, to which Plaintiffs have replied. Defendant RockTenn joined in the combined response but also filed a separate response to the Class Certification Motion based on arguments that apply only to RockTenn. Plaintiff replied to RockTenn’s response, to which RockTenn filed a Sur-reply, which prompted Plaintiffs to file a Sur-sur-reply. After deciding the Motion to Strike, the Court will first consider the joint opposition to class certification and then consider RockTenn’s unique opposition.

A. Motion to Strike Plaintiffs’ Reply Experts’ Reports

Plaintiffs’ initial Class Certification Motion contained expert reports from Drs. Mark Dwyer and Michael Harris. Defendants’ combined response to the Motion included expert reports from Drs. Janusz Ordover and Dennis Carlton, who both criticized Plaintiffs’ experts’ reports in a number of ways, including a criticism of Plaintiffs’ experts’ choice of testing and methodology to prove class-wide injury. In response, Plaintiffs obtained reply expert reports from Drs. Dwyer and Harris, and also obtained a report from a new expert, Dr. Douglas Zona. Defendants have moved to strike certain portions of Dr. Dwyer’s reply report and all of Dr. Zona’s report.

Federal Rule of Civil Procedure 26(a)(2) governs expert discovery and “requires a party to disclose to the other parties a written report of a retained expert that includes ‘a complete statement of all opinions the witness will express and the basis and reasons for them.’ ” Sloan Valve Co. v. Zurn Indus., Inc., No. 10 C 204, 2013 WL 3147349, at *1 (N.D.Ill. June 19, 2013) (quoting Fed. R.Civ.P. 26(a)(2)(b)(i)). An expert rebuttal report is meant to “contradict or rebut evidence” disclosed in the initial report, Fed. R.Civ.P. 26(a)(2)(D)(ii), and its “proper function ... is to contradict, impeach or defuse the impact of the evidence offered by an adverse party.” Peals v. Terre Haute Police Dep't, 535 F.3d 621, 630 (7th Cir.2008). Rule 26 does not address reply expert reports, but, much like reply briefs, parties may not advance new arguments for the first time in a reply expert report. Sloan Valve Co., 2013 WL 3147349, at *1. If a reply expert report is truly rebuttal evidence, then it is admissible and the opposing party is not entitled to respond to it. Id. at *4. If, however, the reply report contains new opinions that are not proper rebuttal testimony, the report must be stricken. Id. at *2-3.

1. Dr. Dwyer’s Reply Report

Plaintiffs’ disclosed Dr. Dwyer as an expert within the time frame outlined in the Court’s scheduling order. The question is whether his reply report constitutes new and alternative opinion testimony or is instead proper rebuttal testimony in support of his original report.

Sloan Valve Co. is instructive. In that case, the plaintiffs initial expert report calculated damages based on “collateral unit sales ratios,” and in that report the expert included data on both weighted and unweighted ratios. Id. at *2. Despite including both sets of data, the expert’s damages calculation relied solely on the weighted ratios. Id. In response, the defendant’s expert attacked the plaintiffs expert for relying upon only the weighted ratios. Id. Consequently, the plaintiffs expert’s reply report conducted a new calculation using the unweighted ratios in order to demonstrate that using the unweighted ratios would not change his initial conclusions. Id. The court found that this was proper reply expert testimony because, “rather than offering a new opinion and changing the basis for the calculation of the collateral unit sales, [the plaintiffs reply expert] included the [new] calculation to refute [the defendant’s expert’s] criticisms.” Id.

The plaintiffs reply expert report also included a “revised and increased estimate of incremental costs” based on data that the defendant first disclosed in its rebuttal expert’s report. Id. at *3. The court refused to strike that reply expert testimony because it was based on data that was previously unavailable to the plaintiffs expert due to the defendant’s failure to disclose it. Id. Finally, the court struck a portion of the plaintiffs expert report that constituted “a new, alternative collateral sales calculation” based on data that was available to the plaintiffs expert when he filed his initial report. Id. The plaintiff argued that the new calculation was in response to the defendant’s expert’s criticism, but the court found that the opinion was new because it included a new opinion based on data that was not a part of the plaintiffs expert’s initial report, though it was available to him. Id.

In this case, Dr. Dwyer’s initial report concluded that “all or nearly all members of the proposed class were impacted by price increases implemented by the defendants,” which is explored more thoroughly below in the Court’s analysis regarding class certification. Dr. Dwyer’s conclusion was based on “systematic, empirical comparisons of net prices paid by class members of Containerboard Products before and after price increase implementation dates previously announced by the defendants.” Defendants’ experts criticized Dr. Dwyer for conducting what Defendants call a “one penny more” analysis, where any increase in price is attributed to Defendants’ alleged conduct. This, according to Defendants, means that Dr. Dwyer’s methods did not account adequately for other factors that would have caused an increase in price even without a conspiracy. Defendants’ experts argue that Dr. Dwyer should have done a “but-for” analysis to determine antitrust impact.

In his reply, Dr. Dwyer does the analyses that Defendants’ experts argue he should have done initially. Importantly, Dr. Dwyer does not abandon his prior methods or conclusions. Rather, he conducted the additional analyses to refute Defendants’ arguments and to show that his original conclusions and opinions are sound and a reliable method of assessing antitrust impact. This makes Dr. Dwyer’s reply report remarkably similar to the reply report allowed in Sloan Valve Co. Much like the reply report in that case that included new calculations based on the same data included in the initial report, here Dr. Dwyer’s reply report is based on the same data in his original report and does not seek to include new data. Instead of abandoning his prior methods in favor of the new ones, Dr. Dwyer’s reply concludes that the new calculations support his initial methodology and opinions. Dr. Dwyer further concludes that Defendants’ experts are wrong when they say that the additional testing and methods show that there is no antitrust impact. This is the very purpose of a reply report: to refute a defendant’s expert’s arguments and to provide further support, rather than abandoning, one’s initial opinions. Dr. Dwyer will be held to the original methodology and opinions in his initial report, but that does not mean he cannot respond to Defendants’ experts’ criticisms in defending his initial conclusions. Thus, the Court denies Defendants’ Motion to Strike to the extent that it seeks to strike portions of Dr. Dwyer’s reply report.

2. Dr. Zona’s Reply Report

Dr. Zona was not initially an expert disclosed before class certification briefing. Dr. Zona is an expert Plaintiffs hired to examine “the opening expert reports submitted by Drs. Harris and Dwyer, as well as the reports submitted by Drs. Carlton and Ord-over.” He also conducted his own “but-for” analysis and concludes that Dr. Dwyer’s “methodology and opinions on both impact and damages [is] reliable and valid.”

Defendants are probably correct that Dr. Zona’s report should be stricken. The Court need not engage in lengthy analysis, however, because Plaintiffs do not need Dr. Zona’s report to satisfy Rule 23, a point Plaintiffs concede. Thus, the Court grants Defendants’ Motion to Strike Dr. Zona’s report.

B. Class Certification—Combined Arguments

Having narrowed the range of expert evidence to only the reports and deposition testimony of Drs. Dwyer, Harris, Ordover, and Carlton, the Court now considers whether to certify Plaintiffs’ proposed class based on the parties’ combined briefing.

1. Rule 23(a) Elements

In order to warrant class certification, Plaintiffs must prove “numerosity, typicality, commonality, and adequacy of representation.” Messner, 669 F.3d at 811. As Defendants correctly note, Rule 23(b)’s predominance standard often overlaps with typicality, commonality, and adequacy. Thus, Defendants have focused their arguments on predominance issues rather than individually attacking each of Rule 23(a)’s elements. Essentially, Defendants have conceded that typicality, commonality, and adequacy have been satisfied so long as Plaintiffs have adequately proven predominance.

As to numerosity, Defendants have not specifically challenged that element in their briefing, and the element is easily satisfied in this ease. The sales data relied upon by both parties’ experts establish that the proposed class numbers in the thousands. A potential class that large is sufficiently numerous for Rule 23(a) purposes. See, Schmidt v. Smith & Wollensky LLC, 268 F.R.D. 323, 326 (N.D.Ill.2010).

2. Rule 23(b)(3) Elements

Plaintiffs seek certification under Rule 23(b)(3), which requires 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.” The Court will discuss predominance first and then, if necessary, superiority.

i Predominance

The parties’ statements of the proper legal standard for determining predominance differ greatly. To read Plaintiffs’ version, one would think that predominance naturally flows from the fact that this is an antitrust ease. [Pl.’s Mot. for Class Cert, at 49, ECF No. 660.] Defendants’ version, on the other hand, would lead one to believe that the Supreme Court’s opinion in Comcast makes satisfying predominance nearly insurmountable. [Def.’s Opp. Br. at 23-27, ECF No. 763.] The truth is somewhere in the middle.

The predominance inquiry under Rule 23(b)(3) “ ‘trains on the legal or factual questions that qualify each class member’s case as a genuine controversy,’ with the purpose being to determine whether a proposed class is ‘sufficiently cohesive to warrant adjudication by representation.’” Messner, 669 F.3d at 814 (quoting Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 623, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997)). Predominance is similar to Rule 23(a)’s typicality and commonality requirements, but “the predominance criterion is far more demanding.” Id. (internal quotation marks omitted). And although the Supreme Court has said that “predominance is a test readily met” in antitrust cases, that simply means that “in antitrust cases, Rule 23, when applied rigorously, will frequently lead to certification.” Id. at 815 (internal quotation marks and citation omitted). This does not, however, make class certification automatic in antitrust cases. See, id.

Generally, predominance is satisfied when “‘common questions represent a significant aspect of [a] case and ... can be resolved for all members of [a] class in a single adjudication.’ ” Id. (quoting 7AA Wright and Miller, Federal Practice & Procedure § 1778 (3d Ed.2011)). In other words, “common questions can predominate if a common nucleus of operative facts and issues underlies the claims brought by the proposed class.” Id. (internal quotation marks omitted). The presence of some individual questions is not fatal, but individual questions cannot predominate over the common ones. Id. To determine if a question is common, the Court must look to the evidence necessary to answer that question; if “the members of a proposed class will need to present evidence that varies from member to member” to answer the question, then the question is an individual one. Id. (internal quotation marks omitted). Conversely, “if the same evidence will suffice for each member” to answer the question at issue, then the question is common. Id.

“Analysis of predominance under Rule 23(b)(3) ‘begins, of course, with the elements of the underlying cause of action.’ ” Id. (quoting Erica P. John Fund, Inc. v. Halliburton Co., — U.S. -, 131 S.Ct. 2179, 2184, 180 L.Ed.2d 24 (2011)). In the antitrust context, plaintiffs must prove: “(1) that [Defendants] violated federal antitrust law; and (2) that the antitrust violation caused them some injury.” Id. Plaintiffs must also show damages, but “[i]t is well established that the presence of individualized questions regarding damages does not prevent certification.” Id. (citing Wal-Mart, 131 S.Ct. at 2558).

To provide a clearer analysis, the Court will discuss each antitrust element separately, keeping in mind that “Rule 23(b)(3) ... does not require a plaintiff seeking class certification to prove” that each individual element is “susceptible to classwide proof.” Amgen Inc. v. Conn. Ret. Plans and Trust Funds, — U.S.-, 133 S.Ct. 1184, 1196, 185 L.Ed.2d 308 (2013). “Rather, the inquiry is more holistic.” In re High-Tech Employee Antitrust Litig., 985 F.Supp.2d 1167, 1184 (N.D.Cal.2013). And although predominance analysis is not simply “bean counting,” Butler v. Sears, Roebuck and Co., 727 F.3d 796, 801 (7th Cir.2013), analyzing each element separately is useful in isolating what questions are common and determining whether those questions predominate.

a. Plaintiffs’ Liability Proof

Plaintiffs have established that common questions regarding liability predominate over any individual issues. Plaintiffs’ theory of liability is based on Defendants’ alleged conspiracy to coordinate supply restrictions and price increase announcements in order to cause the price of Containerboard Products to increase. To prove each element of a conspiracy, virtually all class members would be relying on the same evidence that Plaintiffs have submitted in support of class certification—namely the documents, emails, phone records, and other indirect evidence necessary to prove that Defendants conspired in violation of antitrust laws. This type of alleged conspiracy is the prototypical example of an issue where common questions predominate, because it is much more efficient to have a single trial on the alleged conspiracy rather than thousands of identical trials all alleging identical conspiracies based on identical evidence. See, 7AA Wright & Miller, Federal Practice & Procedure § 1781 (3d Ed.2014) (“[W]hether a conspiracy exists is a common question that is thought to predominate over the other issues in the case and has the effect of satisfying the prerequisite in Rule 23(b)(3).”).

Defendants’ arguments on this issue go entirely to the merits of Plaintiffs’ conspiracy claims. Defendants argue that Plaintiffs have not established a conspiracy, offering up several innocent reasons for their conduct. Defendants also attach great significance to Plaintiffs’ failure to produce any explicit, direct agreement among Defendants to fix prices. Because Plaintiffs have not proven an actual conspiracy, Defendants argue that the Court should deny class certification.

But whether Defendants actually conspired is not the issue before the Court. The issue is whether the conspiracy question will be decided by evidence common to the class, and both parties have demonstrated that the evidence either proving or disproving a conspiracy will be common to the entire class. Defendants’ arguments on this point are identical to the defendants’ liability arguments in In re Polyurethane Foam Antitrust Litigation, No. 1:10 MD 2196, 2014 WL 6461355 (N.D.Ohio Nov. 17, 2014). In that case, the defendants argued that “the price-fixing conspiracy alleged in the Complaint did not exist,” based on “the failure of discovery to yield evidence of any agreement among foam manufactures to fix the timing or content of price increase letters.” Id. at *16 (internal quotation marks and alterations omitted). The defendants also argued that their conduct was “consistent with legal and economic theory predictions of behavior” in the relevant markets. Id. The court rejected those arguments at the class certification stage, noting that those arguments “do not succeed in showing liability questions—however answered—cannot be answered through common proof.” Id.

Like the defendants in In re Polyurethane Foam, Defendants’ arguments here do not demonstrate the lack of common proof; rather, Defendants’ own evidence tending to disprove a conspiracy is common to the entire class. Defendants’ arguments, if correct, might entitle them to summary judgment or a verdict in their favor, but such merits arguments are inapplicable at this stage. Id. Thus, Plaintiffs have established that common questions predominate the liability issue.

b. Plaintiffs’ Impact Proof

The heart of the battle in this case lies in the second element, i.e., causation, which is often referred to as antitrust impact. According to Defendants, individual issues overwhelm the common questions regarding impact because Plaintiffs’ experts have not provided a just and reliable method of proving that the alleged antitrust violations harmed all or nearly all class members. Thus, according to Defendants, Plaintiffs’ only avenue of proving causation is through individual proof relating to each of the thousands of class members, which makes class certification inappropriate.

Given the parties’ overlapping arguments relating to impact and damages, the Court must first outline an important distinction: “impact” and “damages” are two separate elements in an antitrust claim. In re Ethylene Propylene Diene Monomer (EPDM) Antitrust Litigation, 256 F.R.D. 82, 88 (D.Conn.2009). Impact is “whether the plaintiffs were harmed,” whereas “damages quantify by how much." Id. Parties and courts often conflate the two, leading to “confusion about what a plaintiffs burden precisely is at the motion for class certification stage.” Id. Oftentimes, demonstrating impact and damages involves “comparing the ‘but-for’ price—the price a customer would have paid in the absence of the conspiracy— and the actual price paid.” Id. When that happens, the single comparison establishes both impact and damages. Id. Thus:

[O]ne 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.

Id. Defendants in this case base a large portion of their impact arguments on the perceived lack of a “but-for” analysis.

But, “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 an motion for class certification.” Id. For example, “ ‘if it appeal’s that plaintiffs may be able to prove at trial ... the price range was affected generally,’” then the plaintiffs can show impact without a “but-for” comparison, and this is so even if there are negotiated prices or a variety of prices. Id. (quoting In re NASDAQ Market-Makers Antitrust Litig., 169 F.R.D. 493, 523 (S.D.N.Y.1996)); see also, In re Urethane Antitrust Litig., 768 F.3d at 1254-55. The court in Hedges summed up this point succinctly:

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.

Hedges Enters., Inc. v. Cont’l Grp., Inc., 81 F.R.D. 461, 475 (E.D.Penn.1979). Courts have long held that a plaintiff can demonstrate antitrust impact by showing that the conspiracy caused an increase to the standard market price of the product at issue. See, e.g., In re Urethane Antitrust Litig., 768 F.3d at 1254-55 (finding impact satisfied for class certification purposes based in part on evidence of “parallel issuance of similar product ... price-increase announcements”); In re Urethane Antitrust Litig. (Urethane II), 251 F.R.D. 629, 638 (D.Kan.2008) (“[E]v-idence 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 Indust. Diamonds Antitrust Litig., 167 F.R.D. 374, 383 (S.D.N.Y.1996) (“[I]f 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.”).

This distinction between impact and damages is crucial in a case like this, where Plaintiffs have presented (1) record and expert evidence independently showing impact and (2) an econometric model that attempts to prove both damages and therefore impact. Because Plaintiffs do not rely solely on their econometric damages model for their impact proof, Defendants’ impact arguments based on the lack of a “but-for” comparison are ineffective. See, e.g., In re EPDM Antitrust Litigation, 256 F.R.D. at 88-89. Those arguments go to damages, not impact. With this distinction in mind, the Court now addresses impact.

Demonstrating antitrust impact for class-certification purposes does not require that Plaintiffs prove antitrust impact. Instead, Plaintiffs need “only to demonstrate that the element of antitrust impact is capable of proof at trial through evidence that is common to the class rather than individual to its members.” Messner, 669 F.3d at 818 (internal quotation marks omitted). The focus, then, is on the evidence necessary to establish antitrust impact, not on whether Plaintiffs have adequately proven it. Again, the Court’s overriding concern is whether Plaintiffs have established that the proposed class is “sufficiently cohesive to warrant adjudication by representation.” Amgen Inc., 133 S.Ct. at 1196 (internal quotation marks omitted).

Although Defendants have concentrated their impact arguments on the expert evidence, the Court looks to all the evidence to determine whether Plaintiffs have established that evidence common to the class is capable of proving antitrust impact. See, In re Urethane Antitrust Litig., 768 F.3d 1245, 1255-56 (10th Cir.2014). The evidence in this case that goes to antitrust impact consists of (1) record evidence, which Plaintiffs allege shows illegal and anticompetitive conduct that increased the base price for Container-board Products, (2) testimony from Dr. Harris regarding Defendants’ industry and its susceptibility to collusive conduct, and (3) testimony from Drs. Dwyer and Harris that purport to establish that all or nearly all class members suffered harm as a result of the conspiracy.

First, Plaintiffs’ have mustered a large amount of record evidence relevant to impact, which is mostly duplicative of their conspiracy evidence. Specifically, Plaintiffs produced evidence of what appears to be coordinated price increases, coordinated supply reductions, and other similar conduct that, according to Plaintiffs, Defendants would not have engaged in unless acting as part of a conspiracy. Plaintiffs also point to memoranda, phone calls, and trade association meetings that show the ability for Defendants to communicate with and monitor each other regarding their allegedly collusive activity.

Defendants respond by pointing to other evidence showing that they added capacity and supply during the class period, which would negate some of Plaintiffs’ factual claims. Defendants also argue there are innocent reasons for their conduct. The Court need not decide at this stage which evidence to believe, however, because regardless of these factual disputes, the evidence on both sides is common to all class members. The question is whether Defendants’ industry made it possible for them to collude in a way that would allow them to harm all or nearly all class members, and the evidence that both parties rely on to answer that question is common to the class.

Second, Plaintiffs rely on Dr. Harris to establish that the Containerboard Products market was ripe for collusion during the class period, and that Defendants’ conduct is more likely the result of collusion than independent behavior. Dr. Harris conducted a “structure, conduct, performance” (“SCP”) analysis to determine whether “the structure, conduct, and performance of [Defendants’] industry [was] consistent with, and likely to facilitate, collusive conduct, thereby providing a motive to collude and suggesting that any collusion would be broadly successful.” [Declaration of Dr. Hams at 5, EOF No. 658-2.] Dr. Hams also considered whether Defendants’ conduct during the class period was more consistent with “concerted action or their unilateral self-interest.” [Id.] Ultimately, Dr. Harris concludes that (1) “the economic evidence shows that Defendants had the motive, opportunity, and means to collude and were they to do so, ... they would have succeeded,” and (2) “the conduct of the Defendants was more consistent with collusion than with independent economic decision-making.” [Id,].

In support of these opinions, Dr. Harris looked at a variety of industry-wide figures like capacity, operating rates, inventory, demand, and pricing in the containerboard market. Importantly, Dr. Harris found that the “vast majority of sales of Containerboard Products are pegged to published price indi-ces,” the most common of which is the price for “42# Kraft liner [published] in Pulp and Paper Weekly (“PPW”).” The PPW index in this ease is critical because Drs. Harris and Dwyer rely in part on the movement of that index in demonstrating that all or nearly all class members suffered antitrust impact, as discussed more thoroughly below.

As to the “structure” portion of the SCP analysis, Dr. Harris analyzed the extent to which the Containerboard Products market is highly concentrated and the barriers to entry in that market. The evidence shows that, in 2003, only six North American firms controlled 72% of the Containerboard Products market, and by 2007, those six firms became five and accounted for 74% of the market. The evidence also shows, and Defendants do not challenge, that there are significant barriers to any new firms entering the market, such as the enormous amount of capital necessary to start a new firm. Dr. Harris also analyzed Containerboard Products to determine whether they are homogenous, which would mean that class members would see all Defendants as essentially offering the same product. If this is so, then the primary method of competition in the Con-tainerboard Products market is price, rather than some other factor. Dr. Harris concludes that product homogeneity makes collusion easier, and finds that Containerboard Products are interchangeable commodities that are highly homogenous. This conclusion is based on Defendants’ own statements, presentations, and tax filings, which tend to show that Defendants admit the commodity nature of Containerboard Products.

Finally, Dr. Harris analyzed the frequency of Defendants’ interactions amongst each other and their participation and membership in trade associations. A conspiracy is much less likely to exist, Dr. Harris concludes, if firms communicate rarely, On the other hand, constant communication and participation in trade association events provide a fertile ground for collusion. Dr. Harris relies on academic economic literature for his conclusion “that trade associations tend to facilitate collusive conduct,” and the evidence indeed shows that Defendants frequently interacted with each other as a part of their business operations and at various trade association meetings. Based on all this evidence and his own expert experience and opinion, Dr. Harris concludes that the “structural characteristics of the industry ... are consistent with and would facilitate successful collusion among the Defendants.”

As to the conduct portion of the SCP analysis, Dr. Harris looked at mill closures, operating rates/inventories/trades, downtime/slowback, coordinated pricing, monitoring, direct communication among Defendants, and Defendants’ prior history of antitrust violations. In general, Dr. Harris concludes that, in the face of constant and increasing demand during the class period, Defendants reduced capacity—and therefore supply—by closing or slowing down the rate of production at mills. Specifically, Dr. Harris points to strategic mill closures or “downtime” that reduced capacity and supply in the face of “strong growth in box demand.” Also important to Dr. Harris’s analysis are the various price increase announcements that occurred during the class period. Defendants collectively announced fifteen price increases during the class period, nine of which were fully implemented. Defendants made these various announcements at near-identical times and for near-identical amounts. These announcements often occurred very shortly after various trade association meetings. Dr. Harris concludes that this conduct runs contrary to what independent firms would do when faced with similar market conditions and that Defendants’ conduct is more consistent with collusive behavior than with normal, unilateral activity.

As to the “performance” portion of the SCP analysis, Dr. Harris relies on Dr. Dwyer’s report, which Dr. Hams concludes is sound and founded in accepted economic theory. Based on Defendants’ actual economic performance, Dr. Harris concludes that Defendants’ performance is consistent with collusion.

Defendants hurl several attacks at Dr. Harris’s opinions. First, Defendants argue that Dr. Harris failed to define the relevant market for purposes of his SCP analysis. According to Defendants, this failure is fatal and makes Dr. Harris’s analysis useless. Defendants contend that Containerboard Products cannot possibly be a single market, because Containerboard Products include containerboard and corrugated products, which are not interchangeable and thus are not part of the same market. See, Sargent-Welch Scientific Co. v. Ventron Corp., 567 F.2d 701, 710 (7th Cir.1977).

Dr. Harris correctly responds, however, that the Containerboard Products market is the relevant market, and this is so despite the fact that the market includes both containerboard and corrugated products. Defendants draw an analogy between components of a personal computer and the personal computer itself, arguing that the components and the computer could not possibly be the same market, and therefore containerboard and corrugated products likewise cannot comprise a single market. The analogy fails because the analogs are not truly analogous; container-board is not a component that goes into a corrugated product, it is the component. As Dr. Harris points out, containerboard is simply a corrugated product that hasn’t been folded yet. Moreover, containerboard has no other use except for being folded into a corrugated product. And, there are no substitutes for containerboard—that is, corrugated products cannot come from some other source other than container-board. Perhaps most importantly, Defendants’ businesses are vertically integrated such that “the firms who manufacture the containerboard are the very same firms who convert that containerboard into corrugated products.” [Reply Declaration of Dr. Hams at 24, EOF No. 826-2.] This means that there is no useful or principled way to separate the two into separate markets; they appear to be part in parcel of the same market. Thus, the Court finds that Plaintiffs have established and defined the relevant market for antitrust purposes.

Defendants next criticize Dr. Harris for failing to conduct a “but-for” analysis for each event that Dr. Harris analyzed, such as mill closures and downtime. In short, Defendants argue that Dr. Harris should have looked to each individual event during that class period, such as each mill closure, and then determined whether that specific event would have occurred even absent a conspiracy, and then further analyzed whether that specific event caused all or nearly all class members to pay higher prices than they otherwise would have paid.

Aside from “but-for” analysis not being required to show antitrust impact, this argument fails because it sets the hurdle too high. Defendants have not pointed to any case law or economic theory that says an expert conducting a SCP analysis must look at all events in isolation, and then view each individual event to see if that event specifically is the one that caused antitrust impact. To the contrary, it is a well-accepted practice to look to industry events as a whole to determine whether a defendant’s conduct is consistent with collusion, as Dr. Harris does with his SCP analysis. See, In re Processed Egg Prods. Antitrust Litig., — F.Supp.3d-, -, No. 08-md-2002, 2015 WL 337224, at *11 (E.D.Penn. Jan. 26, 2015). Dr. Hams correctly notes that “[n]othing in economic theory suggests that individuals or firms act on the basis of a single reason or animating event.” [Reply Declaration of Dr. Harris at 24, ECF No. 826-2.] Dr. Harris’s analysis simply looked at all of Defendants’ capacity-reducing decisions in total to determine whether they were more likely the result of collusion or not. Plaintiffs’ theory of harm is that Defendants’ collusive actions caused an increase in the market price of Container-board Products, which harmed all or nearly all class members, and Dr. Harris’s SCP analysis is consistent with and supports that theory. Several courts have relied on an expert’s analysis on the structures and features of a market in certifying a class. See, e.g., In re EPDM Antirust Litig., 256 F.R.D. at 90 (“The plaintiffs have not merely alleged that these prices lists existed and that they affected all EPDM purchasers—they have ... provided expert opinion that the structural characteristics of the EPDM market would support collusive increases of prices to artificially high levels.”).

Third, Plaintiffs rely on Dr. Dwyer’s analysis of Defendants’ conduct and the corresponding movement of the PPW index to demonstrate impact. Dr. Dwyer looked at “industry-wide reflections of price and actual prices paid by class members before and after” Defendants’ price increase announcements. [Report of Mark Joseph Dwyer, Ph. D., at 7, ECF No. 658-4.] Dr. Dwyer’s analysis started with looking at the nature of Defendants’ price increase announcements. Of those fifteen announcements, fourteen included all Defendants. For eleven of those fifteen, the announcements were made during the same month. And, for all fifteen, the amount of the increased price was either identical or near-identical across all Defendants. Because of the “lock-step” nature of these price increase announcements, Dr. Dwyer found that “[t]he extent to which the [Defendants’] price increase announcements are reflected—both in timing and in magnitude—in a published price index for liner-board is indicative of impact on prevailing market prices and therefore of the class-wide price impact of such announcements.” [Id. at 8.] Such a method is a common way that courts have allowed experts to demonstrate impact. See, In re Urethane Antitrust Li-tig., 768 F.3d at 1254-55.

As briefly mentioned above, the PPW index is the price index Dr. Dwyer relied on for his analysis. PPW and its price index are published by “RISI,” which is a private publisher that has been publishing the PPW index consistently for thirty years, including throughout the class period. As Dr. Dwyer explains:

Every third week of the month, PPW publishes price indexes for a variety of paper products, including linerboard and corrugated medium. These transaction price indexes are based on surveys of buyers and sellers of a particular containerboard grade. RISI included in the prices it published in its surveys only those related to transactions between nonaffiliated parties. Internal transfers and trades between producers were excluded. PPW also excludes transactions that were contractually tied to a price index. These filters make the indexes more representative of the prices class members actually paid.

[Report of Mark Joseph Dwyer, Ph.D., at 9, ECF No. 658-4.] Of the various products that PPW provides an index for, Dr. Dwyer relied on the price for “42 lb. unbleached kraft linerboard.” [Id.] This index price, according to Dr. Dwyer, “is used as the industry linerboard price in [Defendants’] own analysis of industry pricing.” [Id.] This is supported by evidence indicating that Defendants do indeed rely upon the PPW index in setting their prices for Containerboard Products, negotiating prices in individual contracts, and analyzing the market. [Pl.’s Mot. for Class Cert, at 9, ECF No. 660 (citing evidence ranging from Defendants’ own documents to deposition testimony demonstrating the pervasive use of the PPW index in determining Containerboard Product pricing).]

In comparing the PPW index and Defendants’ price increase announcements, Dr. Dwyer found that nine of the fifteen announcements “are reflected by increases in the PPW Index after the effective dates of those announced increases.” [Report of Mark Joseph Dwyer, Ph.D., at 11, ECF No. 658-4.] Most importantly, in all nine instances, the dollar amounts that the PPW index increased matched Defendants’ announced increase. According to Dr. Dwyer, “[i]f a substantial portion of survey participants had reported that they did not experience a price increase, the PPW Index would not reflect an increase identical to what the • defendants had announced.” [Id.] In other words, Defendants collectively announced near-identical price increases, and shortly following those announcements, the PPW index showed that most customers experienced an increase in price exactly equal to the price increase Defendants announced. Thus, the lock-step increase in the PPW index that followed and tracked Defendants’ collective price-increase announcements demonstrates that nearly all class members suffered antitrust impact.

Defendants first counter Dr. Dwyer’s PPW index assessment by arguing that there is nothing surprising about the fact that a published index price would increase once manufacturers in a market announce that prices are increasing. In logical terms, Defendants characterize Dr. Dwyer’s analysis as running afoul of the post hoc ergo propter hoc fallacy: that the PPW index increased after Defendants announcements does not necessarily mean it increased because of those announcements.

But, as Dr. Dwyer states, there is more than simple correlation here. First, there is more to the relationship between Defendants’ price increase announcements and the PPW index than, say, the relationship between a Chicagoan jumping on one foot and not being eaten by a wild lion. It would be absurd to say that, because the Chicagoan was. not eaten by a lion after jumping on one foot, jumping on one foot must keep lions away. That is because there are several reasonable, common-sense alternatives for why the Chicagoan was not eaten by a lion—for one, the lack of wild lions roaming Chicago. But here, as Dr. Dwyer demonstrates, there does not appear to be any other reasonable explanation for such a close relationship between the PPW index and Defendants’ price increase announcements. Moreover, the source of the PPW index is known; the index represents actual prices purchasers paid following Defendants’ price increase announcements. This constitutes strong evidence that Defendants’ price increase announcements caused the PPW index to increase. And this, in turn, constitutes strong evidence that all or nearly all class members were impacted by the increased price, given Plaintiffs’ evidence regarding the paramount importance of the PPW index in setting prices.

Defendants also argue that Plaintiffs’ reliance on the PPW index in analyzing impact is misplaced. Defendants argue that the index does not reflect class member’s actual transaction prices because it is “not a mathematical reflection of actual transaction prices; it is a level that virtually nobody actually pays for tonnage within assessed specifications.” [Def.’s Opp. Br. at 14, EOF No. 763.] Additionally, Defendants argue that a large number of class members individually negotiated a price rather than simply paying the index price. These arguments miss the mark because Plaintiffs have produced evidence showing that (1) Defendants largely rely on the PPW index in setting prices, and (2) in most individually negotiated contracts, the PPW index factored into the negotiated price. At the least, Plaintiffs have presented sufficient evidence that would allow a fact-finder to infer that, even for negotiated prices, the starting point for those negotiations would be higher if the market price for the product was artificially inflated. This comports with the “prevailing view” that “price-fixing affects all market participants, creating an inference of class-wide impact even when prices are individually negotiated.” In re Urethane Antitrust Litig., 768 F.3d at 1254. Thus, Defendants’ attempt to minimize the importance of the PPW index fails to defeat class certification. And, to the extent that Defendants’ arguments regarding the merits of relying on the PPW index are correct, those arguments are still based on class-wide evidence, which supports a finding of predominance.

Although Plaintiffs have provided additional testimony from Drs. Dwyer and Harris to prove the merits of antitrust impact—that all or nearly all class members actually suffered antitrust harm—at this point the Court is satisfied that Plaintiffs demonstrated that the impact element of their antitrust claim is capable of proof at trial through evidence common to the class. Each class member, if forced to proceed on an individual basis, would be relying on the same evidence of the structure, conduct, and performance of Defendants’ industry and their uniform price-increase announcements in order to show an elevated baseline price for the Container-board Products they purchased during the class period. Thus, the impact evidence in this case is common to the class, and because the evidence, if true, establishes that Defendants’ conspiracy caused a market-wide increase to the price of Containerboard Products, Plaintiffs have established impact for class certification purposes. Numerous cases have found that when a plaintiff produces evidence that the alleged conspiracy increased the baseline price of a product, “there is an inference of class-wide impact.” In re Urethane Antitrust Litig., 768 F.3d at 1254 (collecting cases across several jurisdictions). Evidence that multiple defendants issued “parallel ... price-increase announcements” especially supports the inference of class-wide impact, and this is true “even when prices are individually negotiated.” Id. at 1254-55. Defendants’ arguments that, in fact, most class members were not impacted by the alleged conspiracy may ultimately prove successful at trial or summary judgment. But at the class certification stage, those issues are not before the Court. Therefore, Plaintiffs have established that common questions predominate over individual issues as to impact.

c. Plaintiffs’ Damages Proof

Damages are but one element that the Court considers in determining predominance under Rule 23(b)(3). Roach v. T.L. Cannon Corp., 778 F.3d 401, 408-09 (2nd Cir.2015). To establish predominance, a plaintiff must produce a reliable method of measuring classwide damages based on common proof. See, id. But, “[i]t is well established that the presence of individualized questions regarding damages does not prevent certification.” Messner, 669 F.3d at 815 (citing Wal-Mart, 131 S.Ct. at 2558). Thus, if Defendants are correct that Plaintiffs’ multiple-regression model is not capable of proving class-wide damages, that alone will not defeat certification. See, id. Instead, the Court would need to consider whether the individual issues in calculating damages would overwhelm the common issues relating to liability and impact.

Defendants resist these principles, arguing that the Supreme Court changed the law in this area with its decision in Comcast. Defendants argue that, “since Comcast, a class should not be certified if the plaintiffs fail to present a damages model applicable to individual class members on a class-wide basis or through a simple computation.” [Def.’s Opp. Br. at 55, ECF No. 763.] As several courts since Comcast have noted, however, Comcast did not change the well-established rule that the existence of individual damage issues does not automatically defeat class certification. See, e.g., In re Nexium Antitrust Li-tig., 777 F.3d 9, 21 (1st Cir.2015) (stating, post-Comcast, that “the Supreme Court ... and the circuits in other cases have made clear that the need for some individualized determinations at the ... damages stage does not defeat class certification”).

The Seventh Circuit has also stated Com-cast’s meaning in the class certification context, explaining that Comcast was concerned with a damages methodology that measured the harm resulting from four theories of liability, three of which the district court had rejected. In re IKO Roofing Shingle Prods. Liability Litig., 757 F.3d 599, 602 (7th Cir. 2014). Because the damages model did “not even attempt” to “measure only those damages attributable to” the remaining liability theory, Comcast Corp., 133 S.Ct. at 1433, “the class could not get anywhere,” In re IKO Roofing Shingle Prods. Liability Litig., 757 F.3d at 602. Also, the plaintiffs relied solely on their damages model to also prove impact, rather than presenting separate evidence of impact. Finally, as several courts have noted, Comcast was based entirely on one key concession: the plaintiffs in that ease inexplicably did not challenge the district court’s conclusion that it could not certify the class unless damages were measurable “on a class-wide basis through use of a common methodology.” Comcast Corp., 133 S.Ct. at 1430; see also, In re Urethane Antitrust Litig., 768 F.3d at 1258 (“First, unlike the claimants in Comcast, our plaintiffs did not concede that class certification required a method to prove class-wide damages through a common methodology.”). Plaintiffs in this case have not made a similar concession, and they have presented additional classwide evidence showing that impact is capable of proof at trial, as discussed above. Thus, the Court will consider whether Plaintiffs’ damages model is capable of quantifying damages on a class-wide basis. If so, Plaintiffs have established predominance as to each element of their antitrust claim and the class will be certified. If not, the Court will determine whether individual damages issues will overwhelm the common issues.

In establishing damages, Plaintiffs rely on Dr. Dwyer, who conducted a “preliminary analysis that demonstrates the feasibility of reliably estimating damages on a class-wide basis through the use of ... multiple regression analysis.” Although no expert in this case explains in simple terms what a “multiple regression analysis” entails, the method is common enough to understand through ease law and references like the Reference Manual on Scientific Evidence by Professor Daniel Rubinfeld, which the Seventh Circuit has deemed a reliable source for understanding this type of technical evidence. See, ATA Airlines, Inc. v. Fed. Express Corp., 665 F.3d (7th Cir.2011). “Multiple regression analysis is a statistical tool for understanding the relationship between or among two or more variables.” Daniel L. Rubinfeld, Reference Guide on Multiple Regression, in Reference Manual on Statistical Evidence 305 (3d Ed.2011). The tool looks at a dependent variable, which is the variable to be explained, and independent variables, which are the variables “thought to influence the dependent variable.” In re EPDM Antitrust Litig., 256 F.R.D. at 95. Here, the price of Containerboard Products is the dependent variable, and the independent variables are the various factors that might have an effect on price, like the alleged conspiracy and the costs involved in producing the products. This type of statistical tool “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.” Id. Multiple regression analysis is common in antitrust cases, where the plaintiffs use it to show that an alleged “conspiracy” has a statistically significant impact on the dependent variable—usually price. Ru-binfeld, Reference Guide on Multiple Regression 305-07.

Dr. Dwyer purports to do precisely this type of analysis in measuring damages. But, the Court cannot simply accept Dr. Dwyer’s report simply because it appears to do a multiple regression analysis. Rather, “as painful as it may be,” ATA Airlines, Inc., 665 F.3d at 896, the Court must rigorously screen expert evidence when certifying a class to ensure that the damages model only seeks to prove damages that flow from the harm alleged, Comcast Corp., 133 S.Ct. at 1435.

The reliability of an expert’s multiple regression analysis depends on the choices the expert makes in choosing variables and setting up the model. See, Daniel L. Rubinfeld, Reference Guide on Multiple Regression 311-17. In determining whether a model is set up correctly, courts consider several questions: “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 [or] irrelevant; is the form of the analysis correct?” In re EPDM Antitrust Litig., 256 F.R.D. at 95. (citing a prior edition of Professor Rubinfeld’s Reference Guide on Multiple Regression).

Here, Dr. Dwyer estimates the economic damages attributable to the alleged conspiracy by looking at prices class members actually paid and what they would have paid “but for” the alleged conspiracy. To do this, Dr. Dwyer created two categories of products: Intermediate Containerboard Products (“ICp”), which consists of roll stock and corrugated sheets, and Final Containerboard Products (“FCP”), which consists of all other Containerboard Products. [Report of Mark Joseph Dwyer, Ph.D., at 19, SCF No. 658-4.] Dr. Dwyer then selected a benchmark period of months before and after the class period to compare the prices class members paid during the class period to those prices outside the period. Next, Dr. Dwyer conducted a multiple regression analysis whereby he analyzed the change in the price of Contain-erboard Products during the class period. He included several independent variables to control for non-conspiratorial economic factors that normally influence price, and he also included a “dummy variable,” which is the independent variable that tests whether the alleged conspiracy influenced price. “A positive estimated effect for this dummy variable indicates that prices during the class period are higher than can be explained by the economic factors serving as controls and therefore supports the inference that the elevation is attributable to the collusion alleged by plaintiffs.” [Id.].

As the Court understands it, under Dr. Dwyer’s model, if the non-conspiratorial economic factors fully explain the increase in price during the class period, then the dummy variable indicator would be close to zero, which means that the conspiracy had virtually no effect on price. If this is the case, there are no damages because the conspiracy did not increase prices above what the prices would have been absent the conspiracy. Conversely, if the increase in price is attributable solely to the conspiracy and no other economic factor, then the dummy variable would be close to one. If this is the ease, the damages would be astronomical because the full amount of the price increases during the class period would be attributed solely to the conspiracy.

Dr. Dwyer accounted for numerous variables in an attempt control for economic factors that might explain the change in price during the class period. The Court need not list them all, but Dr. Dwyer grouped them into four categories: downstream demand; production and delivery; inflation; and seasonal factors. Importantly, these categories include independent variables that account for various costs that affect price like the price of pulp that goes into making contain-erboard, labor costs, and fuel costs associated with production and delivery.

With the potential variables selected, Dr. Dwyer ran regressions for the ICP and POP categories. Dr. Dwyer used two different procedures to determine which independent variables, in addition to the dummy variable, would be included in the regression. Under the first procedure, the Alkaike Information Criterion (the “AICC method”), the dummy variable reflected an overcharge of 4.05% for the ICP product category and 3.61% for the FCP product category. Under the second procedure, the Bayesian Information Criterion (the “BIC method”), the ICP overcharge was 4.04% and the FCP overcharge was 2.38%.

Dr. Dwyer also attempted to ássess the robustness of his results by including only the 10 independent variables “with the most explanatory power.” [Id] This regression shows an overcharge attributable to the conspiracy of 3.33% for the FCP products and 2.78% for the ICP products. Dr. Dwyer then used the average overcharge amounts mentioned above to calculate damages. The average overcharge was 2.92% for the FCP products and 3.81% for the ICP products.

To arrive at a total damages figure, Dr. Dwyer multiplied the average ICP and FCP overcharges by the dollar amount of purchases class members made during the class period. Dr. Dwyer also attempted to subtract out of the total purchases those that were made during the class period, but whose prices were part of a contract entered into before the class period. This method is consistent with Plaintiffs theory of harm that Defendants’ conspiracy raised the market price that all class members paid for Con-tainerboard Products during the class period. During the class period, class members paid $21.06 billion for ICP products. Based on Dr. Dwyer’s overcharge of 3.81%, class members paid $801.27 million dollars more than they would have absent the conspiracy. Class members also paid $102.25 billion for FCP products. Based on Dr. Dwyer’s overcharge of 2.92%, class members paid $2,991 billion dollars more than they would have absent the conspiracy. Dr. Dwyer’s report also breaks down the total amount of damages caused by each Defendant. In sum, Dr. Dwyer’s preliminary estimate of damages is $3.792 billion.

Defendants first argue briefly that Dr. Dwyer’s report cannot be trusted because it only shows damages based on the class period reflected in Plaintiffs Amended Complaint, rather than the period initially alleged. Defendants cite no authority for disregarding an expert’s methodology because he relied upon the class period as alleged in an amended complaint. Dr. Dwyer did what was asked of him based on the class period supplied to him by Plaintiffs. That Plaintiffs amended their Complaint does not change the Court’s analysis on whether Dr. Dwyer’s model is capable of demonstrating class-wide damages. To the extent Defendants’ arguments are relevant, they go to the weight a jury might give Dr. Dwyer’s testimony, not its reliability or admissibility. See, In re Urethane Antitrust Litig., 768 F.3d at 1263 (refusing to reverse district court based on the defendants’ argument that the plaintiffs’ expert moved the class period start date to maximize damages).

As to the substance of Dr. Dwyer’s damages model, Defendants’ experts first attack Dr. Dwyer’s use of a technique called “principal components” to address a collinearity problem. The parties’ experts appear to agree that running regressions in this case presents a eollinearity problem, which occurs when two independent variables are highly correlated or related. Including collinear variables in a regression often skews the results.

To correct for eollinearity, Dr. Dwyer inserted 150 independent variables into a computer program that would then select “principal components,” which are those components that, according to the program, best explain the covariance across the independent variables. The program produced 133 principal components, which Dr. Dwyer then inserted into two different “stepwise regression” programs—AICC and BIC. Both programs operate the same way by starting “with a core model in which the variables that are entering into the model selection are absent.” [Dwyer Dep. 137:25-138, Aug. 12, 2014, Ex. 3, ECF No. 763 -2.] The programs then “consider[] adding a variable one step at a time,” creating a model “until none of the remaining factors meet a certain threshold of explanatory power.” [Dwyer Dep. 138:2-6, Ex. 3, ECF No. 763-2.] The only difference between AICC and BIC is the relevant threshold at which the program stops selecting new variables from the 133 available.

Defendants broadly claim that, according to economists, “ ‘stepwise regression is to be avoided.’” [Def.’s Opp. Br. at 51 (quoting Peter Kennedy, A Guide to Econometrics 49 (6th Ed.2008)), ECF No. 763.] But as Dr. Dwyer points out in his reply report, the “stepwise regression” that “is to be avoided” is not the same regression that Dr. Dwyer performed, though admittedly economists often use “stepwise regression” to describe Dr. Dwyer’s approach, a point Dr. Dwyer alluded to in his deposition. [Dwyer Dep. 120:3-5 (“Stepwise can sometimes have a particular meaning that would be inappropriate here.”), Ex. 3, ECF No. 763-2.] Dr. Dwyer provides academic support for the type of model selection procedures he used, and the Court has compared the description of the stepwise method that is disfavored and Dr. Dwyer’s method, and the two are indeed different. Furthermore, Dr. Dwyer notes that the model selection methods he used are included in most or all statistical software packages. The Court finds it unlikely that a model selection method that supposedly is rejected by the academic community would be included in all the software that same community uses to run regressions. Finally, Defendants have not cited any case law that rejects an expert’s model solely because of the expert’s use of Dr. Dwyer’s model selection method. The Court therefore refuses to ignore Dr. Dwyer’s report solely because he employed a stepwise regression.

Defendants next take issue with the way Dr. Dwyer used the AICC and BIC model selection methods. Defendants assert that, despite Dr. Dwyer’s claim that his model is neutral, he did not allow either selection model to select variables in a neutral fashion. This is so, Defendants argue, because Dr. Dwyer “limited the variables the stepwise software could exclude so that the conspiracy dummy variable liad to be included.” [Def.’s Opp. Br. at 51, ECF No. 763.] Defendants interpret this method as forcing the model to assign the conspiracy variable a value even if the model would have otherwise excluded it as having no explanatory value. Defendants argue that Dr. Dwyer should have simply let the software select from all the available variables, and Defendants contend that, by doing so, the conspiracy variable would not have been selected.

Dr. Dwyer, however’, had good reason for including the conspiracy variable in each of his regressions. As he correctly states, allowing the conspiracy variable to be omitted would tell the experts nothing about the conspiracy's effect on price. Moreover, Defendants are incorrect that forcing the model to test for the alleged conspiracy’s effect on price necessarily means the model will find such an effect. Presumably, if the conspiracy had no effect on price, the model would show that the conspiracy variable had a zero or statistically insignificant effect on price. Because the independent variable at issue is the existence of a conspiracy, Dr. Dwyer reasonably included that variable in his regressions. Otherwise, his regressions would tell us nothing about the conspiracy’s effect on price.

Additionally, Defendants’ arguments are inconsistent. On the one hand, Defendants criticize Dr. Dwyer for using a stepwise regression to select which independent variables to include in each model, rather than using his own independent judgment to select those variables. On the other hand, they criticize him for using his economic judgment to use the conspiracy and inflation variables in each model, rather than letting the program pick all of the variables. Both experts state that the stepwise method can exclude variables that the program finds less explanatory, even though economic principles compel inclusion of the variable. Dr. Dwyer used his economic expertise to select certain factors that test for conspiracy and control for inflation, along with other factors selected by the stepwise programs. To the extent that Defendants challenge the specific factors included and excluded from his model, those arguments go to the weight and probative value of his testimony, not to the underlying methodology. See, In re Urethane Antitrust Litig., 768 F.3d at 1260-61.

Of course, in some instances, inclusion or exclusion of a certain variable might be egregious enough to render an expert’s report inadmissible altogether. See, In re Scrap Metal Antitrust Litig., 527 F.3d 517, 530 (6th Cir.2008). But that is not the case here. Defendants have not pointed to some specific, major factor that Dr. Dwyer excluded that shows his model is wholly without merit. To the contrary, Dr. Dwyer went to great lengths to include in his independent variables all sorts of factors that might otherwise explain the price increases during the class period. His methodology therefore appears to be firmly rooted in sound economic and econometric principles, and the large majority of Defendants’ experts’ criticisms go to the merits of whether the price of Container-board Products “increased disproportionately to the cost of inputs as the result of a conspiracy to raise/maintain prices, or instead resulted from a non-eollusive cause.” In re EPDM Antitrust Litig., 256 F.R.D. at 98. This is a merits question that the Court does not need to resolve in order to decide whether to certify the class.

Defendants finally argue that Dr. Dwyer’s damages calculations do not support class certification because his model “cannot be applied to individual plaintiffs, and Dwyer admitted that comparing a customer’s price to the price predicted by his regression model would not indicate whether or not the customer had been damaged.” [Def.’s Opp. Br. at 56, ECF No. 763.] But in a complicated antitrust case such as this, where the theory of harm is that the entire market price of a product was inflated as a result of a conspiracy, “plaintiffs are permitted to use estimates and analysis to calculate a reasonable approximation of their damages.” Loeb Indus., Inc. v. Sumitomo Corp., 306 F.3d 469, 493 (7th Cir.2002); See also, In re Scrap Metal Antitrust Litig., 527 F.3d 517 at 529 (approving a method of comparing a defendant’s profits before and after a violation in “proving antitrust damages”).

Moreover, to the extent that there are individualized damages issues, that alone will not defeat class certification, especially where, as here, common issues predominate the liability and impact elements of Plaintiffs’ claim. Butler, 727 F.3d at 801. And, should discovery demonstrate that individual damages issues indeed threaten to overwhelm the common issues such that class certification becomes inappropriate, Defendants may seek to decertify the class or modify the class so that only liability and impact are decided on a class-wide basis. Messner, 669 F.3d at 826. In light of the key, overwhelming common questions of liability and impact in this case, Defendants have not demonstrated that individual damages issues threaten to overwhelm the litigation.

In sum, the Court has poured over the parties submissions and accompanying evi-dentiary record and finds that Plaintiffs have satisfied the predominance requirement of Rule 23(b)(3) based on competent evidence common to the class.

ii. Superiority

Lastly, Plaintiffs must demonstrate that proceeding on a class basis is superior to other forms of resolving the dispute. Plaintiffs have made a sufficient showing that class treatment in this ease is superior to individual cases. The overarching liability and impact issues are common to the class and can “be resolved in one stroke.” Butler, 727 F.3d at 801 (internal quotation marks omitted). And given the breadth and importance of the common issues, “the superiority requirement ... poses no serious obstacle to class certification here.” Messner, 669 F.3d at 814 n. 5.

In attempting to defeat superiority, Defendants argue first that releases they obtained as a result of settling a prior antitrust lawsuit bar many class members’ claims in this suit. Second, Defendants argue that many class members purchased Containerboard Products pursuant to contracts that include various “disqualifying clauses,” such as forum selection clauses, jury waivers, and clauses that set the available time to bring a claim and the available damages. Defendants conclude that these issues make a class action unmanageable and create a “quagmire” that causes individual issues to predominate the common questions.

As to the first argument, when parties settle a case, a “court may release not only those claims alleged in the complaint and before the court, but also claims which could have been alleged ... in connection with any matter or fact set forth or referred to in the complaint.” Wal-Mart Stores, Inc. v. Visa, 396 F.3d 96, 107 n. 13 (2d Cir.2005) (internal quotation marks omitted). Such a “general release is valid as to all claims of which a signing party has actual knowledge or that he could have discovered upon reasonable inquiry.” Oberweis Dairy, Inc. v. Associated Milk Producers, Inc., 568 F.Supp. 1096, 1101 (N.D.Ill.1983) (citing Goodman v. Epstein, 582 F.2d 388, 402-04 (7th Cir.1978)). In other words, “a release applies only as long as the released conduct arises out of the identical factual predicate as the settled conduct.” In re Digital Music Antitrust Litig., 812 F.Supp.2d 390, 399 (S.D.N.Y.2011) (internal quotation marks omitted).

Defendants settled with various opt-in and opt-out plaintiffs the In re Linerboard, Antitrust Litigation lawsuit in the Eastern District of Pennsylvania. The lawsuit involved an alleged conspiracy that occurred between 1993 and 1995, although some individual cases did not settle until as late as 2008. According to Defendants, a “typical” settlement agreement from the Linerboard litigation released Defendants from all claims “relating in any way to any conduct prior to the Effective Date” of the agreement “concerning the purchase, sale, distribution, or pricing of linerboard, medium, corrugated containers [or] corrugated sheets.” [Def.’s Opp. Br. at 64 n. 59, ECF No. 763.] Defendants argue that such a broad agreement bars the claims at issue here, and that determining whether those claims are barred makes class action impracticable.

The Court disagrees. Although the release language is very broad, Plaintiffs’ claims in this case are not based on an “identical factual predicate as the settled conduct.” In re Digital Music Antitrust Litig., 812 F.Supp.2d at 399. The conduct at issue in the prior litigation was Defendants’ allegedly collusive behavior in the mid-nineties. The actions at issue here are coordinated market manipulation and price increase announcements that occurred nearly a decade later. And even though the collusive behavior that was alleged in the prior litigation is similar to the behavior alleged here, the claims in this case are not based on Defendants’ alleged price-fixing behavior of the nineties. Under Defendants’ argument, they are free to keep colluding in violation of antitrust laws so long as they conspire in the same way as they were alleged to have behaved in a prior settled case. The Court is unaware of any case supporting this argument; indeed, several cases are to the contrary. See, id. Because none of the Liner-board releases apply to this new case that is based on different facts, the Court finds that those releases do not defeat class certification.

Defendants’ second argument is that several contractual provisions apply to disqualify some class members from participating in the class. According to Defendants, “the multitude of individualized issues (under the laws of multiple states) presented by the disqualifying clauses ... render the class action mechanism inefficient.” [Def.’s Opp. Br. at 72.].

The Court rejects this argument for several reasons. First, Defendants’ argument relies on Lozano, in which the Seventh Circuit approved of a district court relying on class action waivers in arbitration agreements to find class certification inappropriate. Lozano v. AT & T Wireless Servs., Inc., 504 F.3d 718, 728 (7th Cir.2007). Unlike the arbitration agreements in Lozano, however, none of the clauses here preclude class actions. Lozano therefore provides no support for Defendants’ position.

Second, Defendants rely heavily on In re Titanium Dioxide Antitrust Litigation, in which the court enforced clauses contained within individual class members’ contracts. In re Titanium Dioxide Antitrust Litigation, 962 F.Supp.2d 840, 851-52 (D.Md.2013). That case also fails to support Defendants’ argument because there, the court did not find the existence of contractual provisions to be an impediment to class certification. Id. at 846 (stating that the defendants’ motion to modify the class definition to exclude class members who were subject to various contractual provisions was premature until “after the expiration of the class opt-out period”). Plaintiffs are correct that the majority of eases hold that the existence of contractual provisions does not automatically defeat class certification. See, e.g., id.; Herkert v. MRC Receivables Corp., 254 F.R.D. 344, 350 (N.D.Ill.2008) (“[A] possibility that some of the putative class members could have card-member agreements with varying terms ... does not necessarily preclude class certification.”).

In the alternative, Defendants ask that the Court at least modify the class definition to exclude those class members that purchased pursuant to a disqualifying clause. This may indeed be necessary, but the Court agrees with the court in Titanium Dioxide that the more appropriate time to address this issue is after the class is certified and Defendants can determine which specific class members are subject to potentially disqualifying contractual provisions. The Court has the authority under Rule 23(e)(1)(C) to modify the class any time before final judgment. To the extent that class members are not eligible to participate due to their specific contracts, Defendants may file a motion to modify the class to exclude those specific members once that information is known.

Because Plaintiffs have proven beyond mere pleading each of Rule 23’s elements, the Court finds that class certification is appropriate. The only remaining issue is Defendant RoekTenn’s arguments as to why a class action cannot proceed against it.

C. Class Certification—Defendant RockTenn’s Argument

RockTenn’s position in this ease is unique. (The Court uses “RockTenn” for ease of reference, although RockTenn was formally known as Smurfit-Stone Container Corporation during its bankruptcy proceedings). Unlike any other Defendant, RockTenn filed for bankruptcy prior to this lawsuit, which resulted in a bankruptcy discharge order dated June 30, 2010 that released RockTenn from any claims predating the discharge order. Thus, according to RockTenn, Plaintiffs’ must independently establish a separate, RockTenn-specific class with a period starting no earlier than June 30, 2010.

In all the briefing on this issue, both parties rely on the exact same source material to support their argument: Judge Milton Sha-dur’s statements during a hearing on November 24, 2010. According to RockTenn, Judge Shadur made it crystal clear that Plaintiffs would need to define and seek certification on a wholly separate class applicable just to RockTenn. Plaintiffs, on the other hand, argue that Judge Shadur could not have stated in any plainer terms that such a separation was unnecessary. The Court can resolve this issue easily by simply referring to what Judge Shadur said, in light of ease law that supports those statements.

Neither party really explains the context for the hearing before Judge Shadur, but from reading the entire transcript of that hearing, it appears that RockTenn asked the bankruptcy court to enjoin this Court from hearing a claim against RockTenn. Judge Shadur set the hearing to infoi’m RockTenn of his position on RoekTenn’s motion and to provide an avenue for communicating his opinion to the bankruptcy court without having to get into a “tug-of-war” with that court. Judge Shadur informed RockTenn that he had read the complaint and its allegations and found that Plaintiffs were only seeking to hold RockTenn liable for its post-discharge conduct. Understandably, Judge Shadur found it inappropriate for RockTenn to ask the bankruptcy court to enjoin this Court from entertaining a cause of action that is founded on postdischarge conduct over which the bankruptcy court has no jurisdiction. Judge Shadur stated that he found it “flat-out misleading [for RockTenn] to characterize the lawsuit before [the Court] as seeking relief from [RockTenn] that is at odds with its discharge in bankruptcy.” [Hr’g Tr. 6, Nov. 24, 2010, ECF No. 108.].

Judge Shadur acknowledged that the complaint’s allegations “speak ... of [Rock-Tenn’s] pre[-]diseharge conduct,” but he noted the key difference between basing liability on pre-discharge conduct and relying in part on pre-discharge conduct as evidence supporting a post-discharge conspiracy. RoekTenn’s argument, in other words, “conflates evidence with claims.” [Id. at 7.] RockTenn is correct that the Court promised to hold Plaintiffs to their word that their complaint is about imposing liability based on post-discharge conduct. But, contrary to both parties’ assertions, Judge Sha-dur did not mention or address any issues related to the proper size or scope of the class to be certified. That is the issue currently before the Court.

The issue presented by Plaintiffs’ complaint is novel. Plaintiffs have provided evidence that RockTenn participated in the alleged conspiracy post-bankruptcy. If a fact-finder found this to be true, RockTenn’s bankruptcy discharge order would not protect it, since that order applies only to pre-discharge conduct. Plaintiffs have also sued RockTenn’s alleged co-conspirators, and co-eonspirators are joint and severally liable “for all ¡damages caused by the [entire] conspiracy.” Paper Sys. Inc. v. Nippon Paper Indus. Co., 281 F.3d 629, 632 (7th Cir.2002) (emphasis added). No problem so far. But, the conspiracy for which Plaintiffs seek to hold RoekTenn’s co-conspirators liable predates RockTenn’s discharge from bankruptcy, and that pre-discharge conspiracy included RockTenn. The question, then, is whether RockTenn, based on its post-discharge conduct, can be jointly and severally liable for its co-conspirators’ damages when the basis for the co-conspirators’ liability is pre-discharge conduct that includes RockTenn. RockTenn argues that to hold it jointly liable is to punish it for pre-discharge conduct. Plaintiffs argue that such a result is the consequence of the general joint and several liability principles in the antitrust context.

Neither party has cited a case directly on point, nor can the Court find one. RockTenn cites cases that preclude class certification against defendants that emerged from bankruptcy, but in those cases liability was premised on solely pre-discharge conduct. See, In re Travel Agent Comm’n Antitrust Litig., 583 F.3d 896, 901 (6th Cir.2009). Plaintiffs cite cases involving joint and several liability, but none of the defendants in those cases had filed for bankruptcy. See, e.g., Paper Sys. Inc., 281 F.3d at 632. Moreover, the Court notes that important principles weigh on both sides of the issue. On the one hand, bankruptcy is meant to absolve a debtor of liability for its pre-discharge conduct. In re Ruben, 774 F.3d 1138, 1141 (7th Cir.2014) (“A principal goal of bankruptcy is to provide the debtor with ... a fresh start.”) (internal quotation marks omitted). On the other hand, joint and several liability is a “vital instrument for maximizing deterrence.” Paper Sys. Inc., 281 F.3d at 633.

In the absence of controlling, binding authority, the Court finds that holding RockTenn jointly and severally liable does not violate the bankruptcy discharge order for several reasons. Plaintiffs alleged that RockTenn joined (or more correctly, rejoined) the conspiracy post-bankruptcy. Generally speaking, “a co-conspirator who joins a conspiracy with knowledge of what has gone on before and with an intent to pursue the same objectives may, in the antitrust context, be charged with the preceding acts of its eo-eonspirators.” Havoco of Am., Ltd. v. Shell Oil Co., 626 F.2d 549, 554 (7th Cir.1980). Although a debtor discharged in bankruptcy is indeed provided a clean slate, what the debtor does with that slate matters. Assuming that RockTenn did, in fact, re-join the conspiracy as alleged, RockTenn certainly would have had knowledge of what has gone on before and presumably would have had the intent to pursue the same objectives. See, id., Seen properly, RockTenn would not be “held liable” for its pre-discharge conduct. Liability would be premised solely on its post-discharge conduct. It only appears that it is being held liable for pre-discharge conduct because of the joint and several liability rule, which makes RoekTenn on the hook for its eo-eonspirators’ actions. But it is the co-conspirators’ liability that is based on pre-discharge conduct, not RoekTenn’s.

The distinction may be a fine one, but it is important. If, for example, Rock-Tenn is ultimately correct on the merits, ie., its post-discharge conduct does not give rise to an antitrust violation, RoekTenn will be absolved of all liability, despite its participation in the pre-discharge conspiracy. The Court has made clear that it will hold Plaintiffs to their burden of proof that RoekTenn’s postdischarge conduct gives rise to liability. But, once liability is established, the general rule of joint and several liability applies. It would be a windfall to defendants to allow them to join a conspiracy post-bankruptcy, with the perfect knowledge and intent to continue causing damages to vast numbers of consumers, and then refuse to enforce joint and several liability while the remaining co-conspirators are all subject to such liability. If RoekTenn did in fact choose to rejoin the alleged conspiracy, it cannot be heard to complain that it may be on the hook for all the damages the conspiracy caused.

Admittedly, this creates potential problems for trial. For example, what is a jury to do with evidence of both pre- and post-discharge conduct in determining RockTenn’s liability? For one, the Court may give a limiting instruction that explains to the jury that, in order to determine whether RoekTenn violated antitrust laws, it must only consider Rock-Tenn’s post-discharge conduct and determine whether that conduct violated the law. These problems, however, can be addressed later if the case goes to trial. For now, the Court agrees with Plaintiffs that this case can proceed as a single class action.

IV. CONCLUSION

For the reasons stated herein, Defendants’ Motion to Strike [ECF No. 845] is granted in part and denied in part. Plaintiffs’ Class Certification Motion [ECF No. 657] is granted. Named Plaintiffs are hereby designated as class representatives, and Michael J. Freed of Freed Kanner London & Millen LLC and Daniel J. Mogin of The Mogin Law Firm, P.C., are hereby designated as Co-Lead Counsel.

IT IS SO ORDERED.