In Re New Motor Vehicles Canadian Export Antitrust Litigation

DECISION AND ORDER ON DEFENDANTS’ MOTIONS FOR SUMMARY JUDGMENT 1 D. BROCK HORNBY, District Judge. In antitrust law terms, the plaintiff car buyers in this multi-district lawsuit against car manufacturers are known as “indirect purchasers” because they bought their cars from dealers, not directly from the manufacturers that they are suing. I previously granted the manufacturers’ motion to dismiss the plaintiffs’ federal antitrust claim for damages, because the Supreme Court decided in Illinois Brick Co. v. Illinois, 431 U.S. 720, 97 S.Ct. 2061, 52 L.Ed.2d 707 (1977), that indirect purchasers cannot recover federal antitrust damages. I did allow a federal injunctive claim and state law damage claims to proceed, and I certified a federal injunctive class and state damage classes. But the Court of Appeals for the First Circuit vacated my class certification orders. In doing so, it also ruled that the plaintiffs’ claim for federal injunctive relief could not proceed, because the altered relationship between the U.S. and Canadian dollar at the time of the appeal had mooted the need for that relief. In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d 6, 15-16 (1st Cir.2008). What remain, now, are state law claims for damages, based upon nineteen states’ *45 antitrust and consumer protection statutes that permit indirect purchasers to recover. 2 The buyers assert that the manufacturers violated state antitrust or consumer protection statutes by conspiring to restrict the movement of lower priced Canadian vehicles into the U.S. market so as to prevent downward pressure on U.S. new vehicle prices. Individually, the manufacturers deny that they conspired, and collectively they deny that the buyers can prove the necessary impact on ultimate consumer transaction prices to establish liability. As a result, those manufacturers who have not already settled or been dismissed have moved for summary judgment both individually and jointly. After oral argument, I now Grant summary judgment to the remaining manufacturer defendants on their joint motion. While there may be sufficient evidence to go to a jury on whether at least some of the manufacturers entered into an illegal agreement, I choose not to resolve that issue because the plaintiffs are unable to prove causation (sometimes called antitrust impact) by the method of proof they have chosen. Without proof of causation, there can be no liability. As a result, certain other motions, including the plaintiffs’ renewed motion for class certification (Docket Item 896), the individual motions for summary judgment (Docket Items 715, 718, 726, 739, 765), and motions to exclude expert testimony (Docket Items 728, 734) are Moot. Background The plaintiff car buyers claim that from at least 2001 through 2003, the currency exchange rate differential between a strong United States dollar and a weaker Canadian dollar created arbitrage opportunities to sell virtually identical, but lower priced, Canadian cars in the United States. These arbitrage opportunities arose from the difference between the prices at which a broker could buy a vehicle in Canada and resell it in the United States, whether to a dealer or a consumer, even after accounting for the various costs associated with exporting the vehicle to the United States. The plaintiffs assert that if cross-border competition had been unhindered, U.S. new vehicle prices would have declined. The plaintiffs claim that the manufacturers violated state laws by conspiring to maintain or enforce policies to discourage this looming cross-border traffic. 3 Ultimately, the business practices enforced by this conspiracy, the plaintiffs say, had the effect of suppressing the supply of Canadian cars in the United States. According to *46 the plaintiffs, this supply restriction supported artificially elevated prices in the domestic United States automobile market in two key components: the Manufacturer’s Suggested Retail Price (“MSRP”) and the dealer invoice price, 4 both of which are set by each manufacturer, are generally known, and together define the negotiating window for the price that consumers actually pay. 5 The plaintiffs say that the manufacturers accomplished their illegal conspiracy to restrict Canadian exports over the course of several meetings, communications of various sorts, and dissemination of best practices and other information directly and through dealer associations. For their part, the manufacturers say that for years they have had legal export restrictions in place, manufacturer by manufacturer, and without agreement, but that they did not enter into any illegal horizontal agreement among themselves. The plaintiffs respond that when the arbitrage opportunities became substantial in 2000 and thereafter, the vertical legal restraints (manufacturer by manufacturer) would have disappeared if there had been no illegal horizontal agreement. Finally, the parties dispute vehemently whether the plaintiffs can prove causation, i.e., that any horizontal agreement (as opposed to any legal vertical restrictions) actually affected the transaction prices that American consumers paid in buying their cars, by the method of proof that the plaintiffs have chosen in this case. In summary, on the manufacturers’ motions for summary judgment, the significant material questions that are in dispute are: (1) whether there was an illegal agreement; (2) whether the plaintiffs can segregate the effects of individual manufacturers’ unilateral legal restraints from the effect of any illegal agreement; (3) whether any illegal agreement affected the prices consumers paid; (4) whether the plaintiffs can prove their contentions by evidence common across all consumer purchases, the method of proof they have chosen; and (5) whether the plaintiffs have an adequate model for proving damages. The claim for each state is that the defendants’ conduct has violated that particular state’s antitrust statute and/or that state’s consumer protection statute. On all these claims, the plaintiffs seek damages as indirect purchasers (as state law allows). The parties do not distinguish between the antitrust statutes and consumer protection statutes for these purposes, so neither do I. Analysis Under all of the nineteen states’ laws— whether those laws be antitrust statutes or consumer protection statutes — the parties agree that the plaintiffs must prove an agreement among the defendants to restrain cross-border traffic; they must prove causation, i.e., that the agreement caused antitrust (or consumer protection) injury, sometimes called antitrust impact *47 in the antitrust cases 6 ; and they must prove damages. 7 Throughout the analysis that follows, I rely heavily on federal case law because, aside from the availability of damages to indirect purchasers under state law, the parties agree that state law generally follows federal. (1) Agreement Proof of illegal conspiracy does not require proof of an explicit or formal agreement. See United States v. General Motors Corp., 384 U.S. 127, 142-43, 86 5.Ct. 1321, 16 L.Ed.2d 415 (1966) (“[I]t has long been settled that explicit agreement is not a necessary part of a Sherman Act conspiracy-certainly not where ... joint and collaborative action was pervasive in the initiation, execution, and fulfillment of the plan.”); Jones v. City of Chicago, 856 F.2d 985, 992 (7th Cir.1988) (“It is enough if you understand the general objectives of the scheme, accept them, and agree, either explicitly or implicitly, to do your part to further them.”). I do not decide the summary judgment issue of illegal agreement finally, but there is probably enough evidence to reach a jury on whether the manufacturers had an illegal horizontal agreement. That conclusion is easiest for Ford and Chrysler; it is somewhat closer for GM because of disclaimer statements it made; 8 it is closest of all for the Honda and Nissan entities because for them the evidence is almost entirely circumstantial. 9 The evidence is *48 hotly contested for all the manufacturers, but viewing it in a light most favorable to the plaintiffs, a jury could find the following. 10 In 1999 to 2000, new vehicle exports from Canada to the U.S. were experiencing a sustained “spike.” Pis.’ Opposing Statement of Add’l Material Facts [Corrected] (“POSMF”) ¶ 718 (Docket Item 964) (superseding Docket Item 858-2). In June 2000 to September 2000, high-ranking executives at Ford Canada, GM Canada, Honda Canada, Toyota Canada and Chrysler Canada communicated directly with each other concerning unauthorized export sales and what their competitors were doing on the issue. Id. ¶¶ 200-19. Early on, one industry executive asked if another manufacturer would be “interested in working together as a group” because “[i]t was for the betterment of the industry” and “it would be easier if all the manufacturers were common in their approach to the dealers, as it relates to export sales.” Id. ¶ 219. Thereafter, in October of 2000, vehicle manufacturers and industry groups planned to “organiz[e] a meeting of industry stakeholders to work toward consensus on the prohibition of export sales,” id. ¶ 222, and “end[ ] unauthorized export sales,” id. ¶ 225. Between October 2000 and May 2001, executives and/or counsel for GM Canada, Ford Canada, Toyota Canada, Chrysler Canada and CADA held conferences on five separate occasions. At one early meeting, the participants discussed a possible “export sales memorandum of understanding amongst industry stakeholders.” Id. ¶¶ 242, 244. Ultimately, these meetings resulted in a “consensus” among the participating manufacturers “supporting] the concept of trying to keep the vehicles in Canada.” Id. ¶¶ 194-95 (describing May 15, 2001 meeting); see also id. ¶ 227 (Oct. 27, 2000 conference call); id. ¶¶ 237-55 (Nov. 9, 2000 conference call); id. ¶¶ 264-67 (Mar. 14, 2001 meeting); id. ¶¶ 273-83 (May 15, 2001 meeting); id. ¶¶ 322-23 (May 28, 2001 conference call); id. ¶ 335 (Feb. 27, 2002 meeting). In addition, the manufacturers sought and received confirmation from industry groups that they would work together “to eliminate export sales in Canada.” Id. ¶ 235. Throughout the end of 2000 and early in 2001, representatives from industry groups and manufacturers continued to discuss controlling Canadian export sales. Id. ¶¶ 256-58 (“development of a common industry solution to the pervasive problem of export sales”); id. ¶259 (“proposing an informal meeting to explore potential ways of addressing” unauthorized exports of *49 new vehicles from Canada); id. ¶ 263 (Industry executive explained that “he is T00% behind working with you and any other manufacturers on this issue of Export Sales’ ”). The stated objective of a May 2001 meeting was “strategy to solve the industry problem of export sales of new automobiles to the United States.” Id. ¶ 287. Notes from the meeting show numerous “Proposed Actions” including: (1) quantifying the problem by exchanging data concerning their respective export data, id. ¶¶ 289-90; (2) having “CADA ... survey dealers with respect to a list of their ‘Best Practices’ to prevent export sales,” id. ¶ 296; (3) investigating “development of a shared database” of exporters names, id. ¶ 298; and (4) seeking legal advice on “any Competition Act implications of any industry-wide export sales initiatives,” id. ¶295. The participants also discussed a “[national] set of rules developed by [the manufacturers]” concerning “what [Canadian dealers] need to do” to discourage export sales; “CADA and provincial] [associations] would support” such rules imposed on dealers to prevent exports. Id. ¶ 302. After the May 2001 meeting, manufacturer and industry group representatives participated in an “Export Vehicles Conference Call,” where the participants “[r]ecogni[zed] that there is no ‘silver bullet’ that will solve the export vehicle problem and that a number of initiatives will need to be undertaken to try and curb the problem.” Id. ¶¶ 322-23, 325. Beginning in spring 2001, each defendant manufacturer stepped up its export restraints. Id. ¶ 569 (Chrysler Canada provided an enhanced due diligence checklist to its Canadian dealers); id. ¶¶ 567-68, 573 (Chrysler Canada initiated several Dealer audits resulting in substantial chargebacks and developed an online version of its known exporter blacklist); id. ¶ 580 (Chrysler Canada expanded charge-back audits); id. ¶¶ 582, 585 (Chrysler Canada began enforcing the company’s longstanding policy of not honoring warranties on exported vehicles); id. ¶ 603 (Ford Canada enhanced its dealer audit and chargeback process); id. ¶ 631 (GM Canada instituted a harsher chargeback policy); id. ¶¶ 633-37 (GM increased frequency of VIN traces, to create blacklists and provide a basis for chargeback audits of dealers and threatened dealer termination); id. ¶ 639 (GM reduced vehicle allocations); id. ¶ 638 (GM decided again to void warranties on Canadian export vehicles); id. ¶ 659 (Honda Canada retroactively charged back $2,000 Canadian dollars per Canadian vehicle exported to the U.S. and imposed holdback to dealers violating its export policy); id. ¶ 665 (Honda “tighten[ed] up the policy on Canadian vehiele[s]” in order to “control the entry of ‘gray market imports’ ”); id. ¶ 679 (Nissan tightened its export policy by voiding warranties on Canadian export vehicles). From summer 2001 through fall 2002, the manufacturers continued to meet and share information about measures that they were undertaking to staunch Canadian exports. Id. ¶ 335 (Feb. 27, 2002 meeting); id. ¶¶ 342-44 (March 2002 communications between industry representatives “discuss[ing] potential industry-wide solutions”); id. ¶¶ 345-46 (Mar. 19, 2002 meeting discussing “CADA Export Sales Initiatives” and “Pricing/alloeation” issues related to unauthorized export of new vehicles from Canada); id. ¶347 (industry group contacted manufacturers, stating that CADA “is interested in getting ‘the group’ together to discuss this issue and potential solutions to the problem”); id. ¶ 348 (manufacturer responses included: “[A]n export chargeback meeting would be helpful. An industry only 1st round would probably help us develop a position and stratagy [sic]”); id. ¶ 350 (manufac *50 turers meeting “on April 2, 2002 concerning .... how to mitigate vehicle exports to the United States”); id. ¶ 359 (goal of the meeting was to “collectively ... endeavor to make the practice of exporting so difficult that it will become unpalatable to the exporters”); id. ¶¶ 360, 362 (participants discussed an “Industry Strategy” and assigned tasks including: (1) “ ‘Best Practices’ List for dealers”; (2) “Commitment of manufacturer to take on legal cases against known exporters; indemnification of dealers by manufacturers when dealers who refuse to sell to suspected exporters or their mandataries are sued;” (3) “Addendum agreement or clause on existing sales agreement where consumer agrees not to export vehicle purchased;” (4) “Common list of Known Exporters;” and (5) “[D]etermine what regulatory actions could be taken to dissuade export brokers”); id. ¶¶ 198-99 (May 2, 2002 industry letter stated: “We have joined forces with the other manufacturers ... and have met with CADA officials to explore additional deterrents to the export activity”); id. ¶¶395, 397 (In May 2002, CADA and General Motors Canada discuss “GM’s approach to export sales issues”); id. ¶¶ 403-04 (In June 2002, Chrysler Canada met at CADA offices and said it “will continue to work with CADA and the manufacturers’ associations to develop a strategy to reduce the export of Canadian vehicles to the U.S.”); id. ¶405 (In fall 2002, an industry group representative communicated that “the [export sales] issue remains a priority”); id. ¶ 409 (In November 2002, manufacturer and industry organizations held an “Export Sales Meeting”); id. ¶ 413 (A November 2002 letter from industry organization stated that “[a]t an ‘Industry’ meeting on the prohibition of Export Sales, held April 2, 2002, consensus was reached between CADA and manufacturers’ associations to work together on the issue”). In sum, there is probably enough evidence to reach a jury on whether there was a horizontal conspiracy. But I do not resolve that issue finally, because the following decision on impact produces summary judgment for the defendants regardless. (2) Impact/Causation To establish liability for an illegal antitrust conspiracy, the plaintiffs must provide evidence not only that there was an illegal agreement, but also that the agreement actually caused antitrust injury. Sullivan v. Nat’l Football League, 34 F.3d 1091, 1103 (1st Cir.1994). 11 Establishing impact is not the same as proving the quantity of damages, but amounts to proving the fact of damage. See Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 114 n. 9, 89 S.Ct. 1562, 23 L.Ed.2d 129 (1969) (observing that the antitrust plaintiffs “burden of proving the fact of damage under § 4 of the Clayton Act is satisfied by its proof of some damage flowing *51 from the unlawful conspiracy; inquiry beyond this minimum point goes only to the amount and not the fact of damage”). I do not write on a clean slate in analyzing this claim. During the defendants’ successful appeal of my certification decisions, the method of how the plaintiffs would prove impact was a central issue. As a result, the First Circuit opinion announced several important propositions on that subject that I must treat as law of the case, binding on me as a trial court judge and on the parties for the duration of this case. 12 Most importantly, the plaintiffs are required to prove impact by common proof that applies to every member of the putative class: “Plaintiffs cannot make their case without common proof of causation, and they can only prove causation through common means if their novel theory is viable; that viability in turn depends on their ability to establish-whether through mathematical models or further data or other means-the key logical steps behind their theory.” In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 25-26. 13 Thus, the plaintiffs cannot and do not rely upon individual buyers’ statements as to how their new car purchase prices would have been affected, had the negotiating window been lowered because of competing Canadian vehicles in the U.S. market. Moreover, the First Circuit ruled that the plaintiffs, as part of their proof of causation, have the burden of separating out the effects of legal behavior on the part of the car manufacturers (the so-called vertical restraints by which an individual manufacturer may have decided individually to limit cross-border traffic of its own Canadian cars) from the effects of any illegal conspiracy. Id. at 27. (a) What the First Circuit Said The following paragraphs from the First Circuit opinion establish the parameters *52 for the causation issue. I have italicized the important requirements that the First Circuit has said that the plaintiffs must meet and that I must apply. Plaintiffs’ theory of impact on indirect purchasers is both novel and complex. Injury in price-fixing cases is sometimes not difficult to establish. Plaintiffs do not, however, advance such a price-fixing theory. Rather, the plaintiffs’ theory is that the higher prices are the result of a “but-for” world. In step one of plaintiffs’ theory, but for the defendants’ illegal stifling of competition, the manufacturers would have had to set dealer invoice prices and MSRPs lower to avoid losing sales to the lower-priced Canadian cars coming across the border for resale in the United States. In step two, the higher dealer invoice prices and MSRPs enabled by this stifling of competition resulted in injury to consumers in the form of higher retail prices. The first step of plaintiffs’ theory requires demonstrating that the defendants’ actions did result in an increase in dealer invoice prices and MSRPs in the United States. This in turn depends on at least two factors. First, there would have had to be, in this but-for world, a flood of significantly lower-priced Canadian cars coming across the border for resale in the United States during times of arbitrage opportunities, enough cars to cause manufacturers to take steps to protect the American market from this competition by decreasing nationally set prices. As plaintiffs themselves note, without a very large number of cars poised to cross the border, a nationwide impact on the automobile market of the sort required by plaintiffs’ theory is implausible, and the theory collapses. In our view, plaintiffs’ expert Professor Hall had not yet, at the time of class certification, fully answered such potentially relevant questions as how the size of the but-for influx of cars would be established or how large that influx would have to be to affect the national market sufficiently to raise effective dealer invoice prices and MSRPs. Second, the plaintiffs must be able to sort out the effects of any permissible vertical restraints from the effects of the alleged, impermissible horizontal conspiracy .... If plaintiffs do not have viable means for distinguishing between these two sets of effects, they cannot show that it was the horizontal conspiracy that caused the impact on the domestic national market upon which their theory depends. As for the second step of plaintiffs’ theory it must include some means of determining that each member of the class was in fact injured, even if the amount of each individual injury could be determined in a separate proceeding. ... Plaintiffs seem to rely on an inference that any upward pressure on national pricing would necessarily raise the prices actually paid by individual consumers. There is intuitive appeal to this theory, but intuitive appeal is not enough. Even if it is fair to assume that hard bargainers will usually pay prices closer to the dealer invoice price and poor negotiators will usually pay prices closer to the MSRP, a minimal increase in national pricing would not necessarily mean that all consumers would pay more. Too many factors play into an individual negotiation to allow an assumption — at least without further theoretical development — that any price increase or decrease will always have the same magnitude of effect on the final price paid. Even if Professor Hall’s proposed models could determine when MSRPs and dealer invoice *53 prices were affected for which models and to what degree, it is a further question whether it can be presumed that all purchasers of those affected cars paid higher retail prices. Id. at 27-29 (emphasis added, footnotes & citations omitted). 14 (b) Applying What the First Circuit Said on the “First Step” I conclude that the plaintiffs have enough evidence to get to a jury on what the First Circuit described as the two factors of the first step. They can show enough Canadian cars poised on the border to make a difference (i.e., enough volume that would cause manufacturers to reduce prices if the cross-border traffic were not prevented), and they can produce evidence that legal vertical restraints would have disappeared during the period in question, so that only the alleged horizontal agreement remains implicated. The evidence on both is fiercely contested, and indeed the plaintiffs’ expert has changed his opinion on the second factor during this lawsuit, but those are issues for a jury. (i) The Threat of Cross-Border Competition The summary judgment record contains sufficient evidence that the threat of cross-border movement was enough to affect prices. Over the class period, Professor Hall calculates that approximately 1,572,-695 of the defendant automakers’ new vehicles sold at retail in Canada had price gaps greater than export costs. POSMF ¶ 872. There is ample evidence to support a finding that the manufacturers were aware of and concerned about the effect of unrestrained Canadian new vehicle exports on United States prices, dealer relations and profits generally. Id. ¶¶ 717, 719-20, 729, 735-37, 751, 754, 766-67. The evidence on this issue includes the following: In June 2001, an internal Honda document stated: “The export of product through the Canadian dealer network into the United States has reached the critical point and must be stopped.” Id. ¶ 757. Referring to the export problem, in August 2001, Chrysler U.S.’s Vice-President of Sales wrote: “This practice must stop immediately” and that “Teeth must be put into the system to prevent this practice from continuing.” Id. ¶ 721. In March 2002, he wrote: “the flow of Canadian new and nearly new vehicles is now officially ‘out of control.’ ” Id. ¶ 745. During this time, Chrysler management believed that unauthorized Canada-to-U.S. vehicle exports were costing Chrysler $100,000,000 per year. Id. ¶ 734. Later, in July 2002, Ford’s Vice President stated: “I do worry that if all OEM’s put too much pressure on the system so publically [sic], we’ll end up with litigation outside of Canada. We could expedite the ‘Europeanization’ of pricing in North America, if we aren’t careful. That will mean pricing at the lowest common denominator, no matter what we pretend to think.” Id. ¶ 753. The plaintiffs’ position is further supported by the testimony of vehicle exporters, who sold Canadian vehicles to United States authorized dealers as well as dealers of competing brands. Two exporters estimated that they each could have exported at least 1,000 more new vehicles per year from Canada to the United States during the class period were it not for the restraints imposed by the manufacturers. Id. ¶¶ 879-80. *54 It is the threat that the plaintiffs have to prove, not the actual movement of cars. Prof. Hall’s Resp. to Defs.’ Rebuttal Reports ¶¶ 28, 36, 39, Feb. 3, 2008 (“Hall Feb. 3, 2008 Report”) (Ex. 77 to Decl. of Daniel Purcell in Supp. of Honda’s Mot. for Summ. J. (Docket Item 763)) (Docket Items 817-77 through 817-79); Joseph P. Kalt Dep. 199:3-13, Dec. 13, 2007 (Ex. 35 to Defs.’ Reply Statement of Material Facts (Docket Item 961) (superseding Docket Item 927)) (Docket Item 929-5) (agreeing that if trade opens between two markets and there is a threat of imports from a country with lower prices, prices will decrease in the country with higher prices); Carl Shapiro Dep. 87:11-19, Dec. 13, 2007 (Ex. 692 to Pis.’ Opp’n to Defs.’ Mots, for Summ. J. (Docket Item 964) (superseding Docket Item 853)) (Docket Item 893-32) (same); James Levinsohn, Testing the Imports-as-Market-Discipline Hypothesis, 35 J. Int’l Econ. 1, 2 (1999) (“When faced with intensified international competition, domestic industries, which may have reaped oligopoly profits in a protected domestic market, are forced to behave more competitively.”). The defendants’ experts also agree that during the class period the threat was looming. See Report of Prof. Howard P. Marvel ¶ 99, Oct. 26, 2007 (Ex. 41 to Defs.’ Reply Statement of Material Facts) (Docket Item 929-11) (“Ford’s gray market exports in 1999-2002 exceeded those in any prior period.”); Expert Report of Kevin M. Murphy ¶ 36, Oct. 26, 2007 (Ex. 707 to Pis.’ Opp’n to Defs.’ Mots, for Summ. J.) (Docket Item 894-15) (“This decline [of the Canadian dollar] corresponded to an expansion of gray-market exports from Canada, which peaked at roughly 4,000 [Chrysler] cars per month in [early] 2002.”); Expert Report of Janusz A. Ordover ¶ 48, Oct. 26, 2007 (Ex. 45 to Defs.’ Reply Statement of Material Facts) (Docket Item 929-15) (“[B]y early 2001, the U.S.-Canadian exchange rate had become even more favorable to the U.S. dollar, increasing the arbitrage incentive to export cars from Canada to the U.S. on the gray market.”); Expert Report of Prof. Carl Shapiro at 21 & Ex. 8, Oct. 26, 2007 (Ex. 170 to Defs.’ General Motors Corp. and General Motors of Canada’s Statement of Material Undisputed Facts in Supp. of Their Mot. for Summ. J. (Docket Item 960) (superseding Docket Item 721)) (Docket Item 777-17) (“[T]he periods of relatively-high export activity for [General Motors of Canada] vehicles correspond to periods in which the Canadian dollar was its weakest vis-á-vis the U.S. dollar,” which was late 2000 through mid-2002.). (ii) Sorting Out the Vertical Restraints With regard to the second question related to “step one,” the Court of Appeals identified the need for the plaintiffs to sort out the effects of permissible unilateral restraints from the effects of any illegal horizontal conspiracy. The plaintiffs assert that Professor Hall has now conducted an econometric profitability analysis that sorts out the effects of permissible vertical restraints from impermissible horizontal restraints. See Pis.’ Opp’n to Defs.’ Mots, for Summ. J. at 29-31; Expert Report of Robert E. Hall on Impact & Class Damages ¶¶ 78-107, May 10, 2007 (“Hall May 10, 2007 Report”) (Ex. 711 to Pis.’ Opp’n to Defs.’ Mots, for Summ. J.) (Docket Item 894-19). In his May 2007 Report, Professor Hall conducted a profitability analysis of automobile industry data and concluded that without a horizontal conspiracy, the manufacturers would not have enforced any vertical export restraints. 15 *55 Hall May 10, 2007 Report ¶¶ 78-107. Because it would not have been profitable to enforce vertical export restraints unilaterally, Professor Hall concluded that no defendant manufacturer would have done so. Id. ¶ 106. Instead, Professor Hall asserted that the profit-maximizing- response to the export threat would have been to set U.S. prices at lower levels. Professor Hall initially recognized that there were legal unilateral restraints as well as unlawful restraints, and that there would continue to be some legal vertical restraints even if the illegal horizontal agreement were taken away. Robert E. Hall Dep. 68-72, Aug. 29, 2005 (Ex. 642 to Pis.’ Opp’n to Defs.’ Mots, for Summ. J.) (Docket Item 891-29). At that time, Professor Hall proposed that his damages model would differentiate between the lawful vertical restraints and the unlawful restraints. Id. at 72, 75-76. But ultimately he changed his mind: “In my opinion, the appropriate but-for world in this case is one in which carmakers did not [jointly or unilateral!y] impose export restraints.” Hall Feb. 3, 2008 Report ¶¶ 16, 108; see also id. ¶ 11. Certainly the jury might hold his change-of-heart/opinion against him, in deciding whether to accept his final opinion, but that change-of-heart is not enough to grant summary judgment in favor of the defendant manufacturers. An expert is not required to stick with the original opinion when persuaded that it is wrong. The plaintiffs have enough evidence from which the jury could conclude that the change in the economic environment would have eliminated the vertical restraints in the absence of an illegal horizontal agreement. Based on Professor Hall’s analysis, there is a sufficient factual basis in the record to generate a genuine issue of material fact on “step one” of the plaintiffs’ impact theory. (c) Applying What the First Circuit Said on the “Second Step”: Proving that each member of the putative class was in fact injured The defendants assert that the record evidence fails to establish that all members of the putative class were in fact injured. In their summary judgment memorandum the plaintiffs do not contend that they have .a methodology that will demonstrate (by common proof) that all consumers suffered injury from the alleged conspiracy, but rather that “[t]he empirical data in Prof. Hall’s analysis show that-injury flows from MSRP to transaction prices with a high degree of correlation, thus impacting virtually all class members.” 16 Pis.’ Opp’n to Defs.’ Mots, for Summ. J. at 69 (emphasis added). The plaintiffs do not directly address the First Circuit’s requirement that they make a showing that “all consumers would pay more,” In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 29 (emphasis omitted), but rather assert that their burden in proving the element of causation — including the likelihood of pass-through 17 — is one satisfied by a prepon *56 derance of the evidence. Pis.’ Opp’n to Defs.’ Mots, for Summ. J. at 70. Relying on Automatic Radio Mfg. Co. v. Ford Motor Co., 390 F.2d 113 (1st Cir.1968), the plaintiffs assert that antitrust causation does not need to be proven with 100% certainty. Pis.’ Opp’n to Defs.’ Mots, for Summ. J. at 70-71 (“We have long since crossed the bridge of precision of proof of causation and extent of damages in antitrust cases.... [W]e recognized that ... in an antitrust suit, covering as it must many imponderables, rigid standards of precise proof would make a plaintiffs task practically hopeless.” (quoting Automatic Radio Mfg. Co., 390 F.2d at 117)). Although the plaintiffs are correct that all of the nineteen remaining states’ laws require proof of antitrust or consumer protection claims to be established by a preponderance of the evidence, according to the First Circuit the plaintiffs must nevertheless establish that all class members paid a higher price. In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 28-29. I conclude that the plaintiffs do not have enough evidence to reach a jury on the First Circuit’s second step; they cannot show that every member of the putative class was in fact injured by paying a higher transaction price than he or she would have paid without the restraint. 18 The plaintiffs summarize this part of their case as follows: [AJbsent an agreement with competing manufacturers to maintain or enhance export restraints, all defendant manufacturers facing inter-brand and intrabrand competition from Canadian exports would have maximized profits in these circumstances by lowering list prices. POSMF ¶709.... [PJlaintiffs and other automobile purchasers in the U.S. — aggressive and passive bargainers alike — would likely have paid a lower price than they did pay, as transaction prices in the absence of such export restraints would have been lower. POSMF ¶ 885. Economic analysis and empirical economic study indicate that a change in effective list prices (list prices net of rebates) is passed through to transaction prices; here, this means that the elevated MSRPs governing plaintiffs’ transactions resulted in elevated transaction prices. POSMF ¶¶ 885-891. A defense affiant confirms that effective list price changes by one manufacturer are quickly followed by others, with the result of lower transaction prices for all models. POSMF ¶ 886. Pis.’ Opp’n to Defs.’ Mots, for Summ. J. at 20. Because this is the critical part of the plaintiffs’ case, I also quote in full their more detailed argument: New vehicle purchases at retail are individually negotiated between the consumer and dealer, but the starting point for the negotiation is the dealer invoice price and MSRP. POSMF ¶ 14. Defendants agree that list prices are available to consumer purchasers, and frame the individual negotiations. POSMF ¶ 888. Further, it is undisputed that most transaction prices fall between effective *57 dealer invoice price and effective MSRP. POSMF ¶ 889. With dealer invoice prices and MSRPs elevated, the entire distribution of bargaining outcomes between consumers and dealers shifts. POSMF ¶ 890. The question, then, is whether an alteration of the negotiating range that comes with higher list prices translates into a “pass th[r]ough” of a higher price to the consumer. There is substantial evidence from which a jury may conclude that the answer to that question is “yes.” Plaintiffs’ expert economist, utilizing reliable, empirical, economic analysis, has opined that the retail prices actually paid by plaintiffs were elevated by the conspiracy. In addition, a defense affiant confirms that changes to effective list prices pass through to transaction prices. POSMF ¶ 886. Prof. Hall relied on published, peer-reviewed, empirical research that resolves the question of whether a change in effective MSRP is passed through to the consumer transaction price. A National Bureau of Economic Research (“NBER”) paper, first published in 2004, analyzed the outcomes of consumer-dealer price negotiations. The researchers used actual U.S. transaction data, and accounted for factors such as types of vehicles purchased, dealers’ costs, consumer demographic information and the competitive environment facing dealers. POSMF ¶ 891. Established econometric and statistical techniques were used by the researchers, including a difference-in-difference approach and a regression discontinuity approach. The NBER study found that a given change in effective price made through a manufacturer’s consumer cash rebate had the effect of lowering the transaction price by 70-90% of the amount of the cash rebate. POSMF ¶ 891. Defendants submitted a witness declaration with their Joint Motion which confirms that changes to effective list prices pass through to the transaction prices of all models. Nissan’s Director of Pricing Strategy, Duane Leffel, describes a real-world situation when one manufacturer cut U.S. prices, others followed and transaction prices were impacted: A recent example of this occurred in June 2005, when GM announced an “Employee Pricing” incentive program across the United States for all of its model year 2005 vehicles. Ford and Chrysler announced in July 2005 that they would also offer a national incentive program similar to GM’s for their model year 2005 vehicles. These incentive programs resulted in transaction prices that were up to thousands of dollars less than MSRP for all of the models for three competitors ... POSMF at ¶ 886. Id. at 27-28 (emphasis in original). Untutored, I might have found the foregoing sufficient to reach a jury, but remembering the First Circuit’s admonitions (“an inference that any upward pressure on national pricing would necessarily raise the prices actually paid by individual consumers” is only “intuitive appeal;” “intuitive appeal is not enough;” “[t]oo many factors play into an individual negotiation to allow an assumption — at least without further theoretical development — that any price increase or decrease will always have the same magnitude of effect on the final price paid;” and “it is a further question whether it can be presumed that all purchasers of those affected cars paid higher retail prices”), I examine the record citations that the plaintiffs have provided here, so as to assess whether they have produced what the First Circuit said that they must, ie., “that all consumers would pay *58 more.” In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 29 (emphasis in original). Their evidentiary support consists of the New Cars Order in the United Kingdom, the Duane Leffel statement, the National Bureau of Economic Research (“NBER”) paper, 19 and Professor Hall’s analysis and opinions. First, do the plaintiffs have evidence that the illegal agreement was enough to move all list prices (MSRPs and dealer invoice prices) on all models, as they say they do? This question is perhaps a close call. The plaintiffs rely in part on the United Kingdom/Ireland experience where vehicle prices in the United Kingdom at one time were much higher than in Ireland and other European Union countries. Hall May 10, 2007 Report ¶¶ 129, 133. The British government issued a “New Cars Order” in 2000, which effectively loosened then-existing restraints on imports into the United Kingdom of cheaper, identical vehicles poised to enter from Ireland and other European Union countries. But the comparability of that experience is very much in doubt. As the defendants’ experts point out, there were a number of other factors potentially at work in that experience. 20 Moreover, the previous price differences were much greater in the UK/Ireland example. Hall Feb. 3, 2008 Report ¶¶ 189-90 (0.4% in U.S. v. 7.4% in UK). Ultimately, because of the uncertainty, the disagreement boils down to whether the plaintiffs’ expert must prove that the other factors in the New Cars Order did not affect the outcome, or the defendants’ experts must prove that the other factors did affect the outcome. See id. ¶¶ 20, 22. This economists’ argument over the validity of the comparison is doubtful stuff for a jury decision. On the other hand, some of the defendants’ own employees have made statements that show that they contemporaneously feared that U.S. list prices would be affected by Canadian vehicles, POSMF ¶¶738, 753, 801, and perhaps there is barely enough therefore to get to a jury. But I need not finally resolve that controversy because even if a jury could find that the illegal agreement caused all list prices to move, the plaintiffs definitely fail to provide evidence that each transaction sales price was affected by the agreement. They imposed that requirement upon *59 themselves — that every transaction with a retail purchaser was affected and that they would show it by common proof — during the earlier progress of this case, including their appearance before the First Circuit. In his July 28, 2005 report, Professor Hall said: Changes in invoice prices and MSRPs would have a common impact on retail transaction prices — prices that U.S. consumers pay. The MSRP and the invoice price are the anchor points from which consumers and dealers negotiate the final transaction price. Retail transactions typically occur at prices between the invoice price and MSRP. As a result, reducing the invoice price and the MSRP would lead to a downward shift in the distribution of transaction prices. Those consumers who negotiate well would still achieve a price closer to the invoice price than those who negotiate badly, but all would benefit from the shift in the bargaining window.. Moreover, economic theory shows that intermediaries are highly likely to pass on to their customers most of any cost reduction they enjoy. A reduction in the price dealers pay for automobiles, then, will be passed through to consumers. The higher wholesale prices enabled by export restraints led to higher end-purchaser prices. Class members experienced a common impact. Expert Report of Robert E. Hall ¶ 16, July 28, 2005 (“Hall July 28, 2005 Report”) (Ex. A to Defs.’ Joint Mot. to Exclude Opinions & Testimony of Robert E. Hall (Docket Item 959) (superseding Docket Item 728)) (Docket Item 788). For the First Circuit, that was insufficient. As I have quoted from the opinion, the “intuitive appeal” of that “theory” was “not enough.” It “would not necessarily mean that all consumers would pay more.” In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 29 (emphasis in original). The court wanted at least “further theoretical development” before “it can be presumed that all purchasers of those affected cars paid higher retail prices.” 21 Id. So I examine both the evidence and the theory on transaction prices that the plaintiffs have listed in response to the defendants’ joint motion for summary judgment and the First Circuit’s announced description of what is required. 22 (i) Testimony of Duane A. Leffel I start with the plaintiffs’ use of Duane Leffel’s declaration that certain “Big Three” incentive programs resulted in transaction prices that were up to thousands of dollars less than MSRP for all of their models. 23 The plaintiffs quote a *60 statement that was the precursor to Leffel’s statement that Nissan USA did not follow the employee incentive pricing incentives of GM, Ford and Chrysler on that particular occasion. Assuming that as a Nissan employee he has personal knowledge to testify that transaction prices on all models of the Big Three were reduced on that occasion, a jury could not find his statement sufficient to show that every reduction in list prices affects every transaction. There are two reasons: first, Leffel was speaking in generalities, focusing on whether Nissan USA followed competitors’ pricing strategies and, while he referred to all models, he did not purport to say that every buyer was impacted; second, the amount of the incentive he described was obviously high (“thousands of dollars less than MSRP”), Decl. of Duane A. Leffel ¶ 16, Mar. 19, 2008 (Ex. 726 to Pis.’ Opp’n to Defs.’ Mots, for Summ. J.) (Docket Item 894-31), whereas many buyers in the putative class in this case were looking at minimal list price differentials — as low as $3 for the GEO Metro, according to Hall. See Revised Ex. 19 to Hall May 10, 2007 Report, Price Elevation for Models Sold in the U.S. as a Result of Imposition/Tightening of Restraints (Ex. 704 to Pis.’ Opp’n to Defs.’ Mots, for Summ. J.) (Docket Item 894-12). Leffel also pointed out in his Declaration that dealers are not limited to the MSRP and can in fact charge more than that price. Leffel Decl. ¶ 4. (Transaction prices above the MSRP occur from time to time particularly for what are known as “hot cars.” Expert Report of Kevin M. Murphy 1H!74b, 137, Oct. 31, 2007 (Ex. F to Defs.’ Joint Mot. to Exclude Opinions & Testimony of Robert E. Hall) (Docket Item 788-6).) I conclude that the Leffel Declaration is insufficient to support the plaintiffs’ theory that all transaction prices are affected even when the list price alteration is minimal, a primary concern of the First Circuit. 24 (ii) NBER Study I turn next to the NBER paper. This was a study designed to determine if it mattered to ultimate consumer pricing whether a car-pricing cash incentive went directly to a consumer or went instead to a dealer. 25 The NBER study found that, as a result of price negotiations between buyer and dealer, when a given incentive was paid directly to the customer, the customer got 70% to 90% of it (ie., the dealer got part of it by negotiating a higher price, given the incentive), whereas when the incentive was paid to the dealer, the con *61 sumer got only 30% to 40% of it. Meghan Busse, et ah, $1,000 Cash Back: The Pass-Through of Auto Manufacturer Promotion, 96 Am. Econ. Rev. 1253, 1254 (2006). I note first that the First Circuit was aware of and cited this NBER study, In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 21 n. 21, but still concluded that the plaintiffs did not have enough evidence to support their impact case. Second, the plaintiffs apparently are using the paper as independent evidence for their case. They argue in their reply brief on the renewed class certification motion that it indicates that a pass-through rate of 70-90% occurs in 99.7% of all transaction. Pis.’ Reply to Defs.’ Opp’n to Pis.’ Renewed Mot. for Class Certification at 3-4 (Docket Item 964) (superseding Docket Item 950). 26 The defendants reject that statistical analysis. I do not rule on whether the plaintiffs or the defendants are correct on the validity of this interpretation, but only rule that use of the study for that proposition requires more than introduction of the paper and argument to the jury. The lawyers are not able to testify about its significance, but require expert testimony to do so. Third, what the plaintiffs actually have by way of expert use of the NBER paper is this: Hall reiterating in two paragraphs that the study’s conclusion was directed at the proportion of the rebate that was passed through to the transaction price and that the researchers used reliable methods of economic analysis and confirmed their results using multiple approaches. Hall July 28, 2005 Report ¶¶ 56-57. Those two paragraphs do not establish that Hall believes that pass-through happens in every case. (Hi) United Kingdom-Ireland Experience Aside from the Leffel Declaration, the only other new information that the plaintiffs have produced (new since what the record contained on the certification appeal) is the UK experience with the 2000 New Cars Order. But even the plaintiffs’ expert, Professor Hall, testified that the UK experience shows only the effect that the loosening of restraints on certain business practices had on list prices. Hall Feb. 3, 2008 Report ¶ 187. Professor Hall points to nothing in the UK experience with the New Cars Order that reveals what happens to actual transaction prices between dealer and buyer. 27 (iv) Professor Hall Finally, for “further theoretical development,” I turn to what Professor Hall has said in his most recent report, February 3, 2008: 282. In my analysis, I infer that the changes in list prices were passed on to consumers fully in percentage terms. If, for example, list prices change by 5 percent then the consumer transaction prices will also change by 5 percent. Professor Murphy states that dealers do not fully pass all of the savings from a decrease in list prices on to consumers and thus my analysis needs to be adjusted to account for the possibility of an incomplete pass-through. His basis for this argument is an academic paper [the NBER study by Busse et al.] that re *62 ports that 70 to 90 percent of carmakers’ consumer cash-back incentives are passed on to consumers. 283. The paper by Busse and co-authors to which Professor Murphy refers does not consider the extent to which dealers pass changes in list prices on to consumers. Rather, the paper considers the extent to which dealers pass two different types of cash incentives to consumers: consumer cash and dealer cash. The authors find that only 30 to 40 percent of dealer cash incentives are passed on to consumers, while consumers receive 70 to 90 percent of consumer cash incentives. The authors conclude that the difference in pass-through rates is explained by the visibility of the incentives: “customer rebates are well-publicized to customers, while dealer discount promotions are not.” 284. Changes in suggested retail prices are entirely observable — they are posted on each car’s window in the dealer’s lot, while consumer cash-back incentives are not — and thus one would expect that all of the change in list prices would be passed on to consumers. In my opinion, a change in list prices is equivalent to a shift in the entire bargaining window. Decreasing the list price decreases the maximum amount that the dealer typically requests and thus shifts the range over which the dealer and consumer will negotiate. This is consistent with 100 percent pass-through. Professor Murphy states “Professor Hall admits that the literature implies a range of effects of MSRP on transaction prices,” but this is untrue. He cites my May 2007 report and October 2007 deposition, neither of which supports his statement. 285.The most important difference between the findings of Busse and coauthors and my measure of price elevation, with respect to pass-through, is that the changes in consumer cash-back studied in that paper were carmakerspecific or even model-specific, whereas the changes in list prices in my but-for world would have been across all makes and models. 28 Three observations are critical here. First, Professor Hall agrees that the Busse paper “does not consider the extent to which dealers pass changes in list prices on to consumers.” 29 Id. ¶ 283. That conclusion by the plaintiffs’ expert confirms that the paper cannot be independent evidence of common proof of impact on transaction prices. Second, in response to the First Circuit’s expressed dissatisfaction with the sufficiency of the intuitive appeal of the economic theory, Professor Hall has added nothing. He simply “infers” that the entire change in list price is passed on to the car buyer; as the First Circuit looks at it, that is nothing more than additional “intuitive appeal.” Third, Professor Hall does reject a defense expert’s interpretation that the NBER Busse paper shows that only a portion of the list price change is passed on. But Professor Hall has no further support for his conclusion of complete pass-on, only further inference. He states that the NBER study shows that when a check for a given amount is presented to the buyer at the closing of the *63 deal, that price change is not “entirely observable,” whereas the MSRP is “entirely observable” because it is posted on the ■window of the car to be purchased. Professor Hall infers from this that if 70% to 90% is passed on when the consumer receives a check (the Busse study), 100% will be passed on if the MSRP is reduced. That is neither further evidence nor further theoretical development. I conclude, therefore, that the plaintiffs still are unable to prove causation, i.e., impact on all transaction prices. (d) Individual States’Laws My reasoning and conclusion do not differentiate among the nineteen states. California, the easiest case for the plaintiffs’ burden because of a shifting presumption (as I said in my certification order), is no longer in the mix: the plaintiffs elected not to pursue their California class certification in this lawsuit, and I granted the named California plaintiffs’ motion to dismiss the California claims. (Similar claims are being pursued in California state court.) Each of the other nineteen states requires affirmative proof of causation. 30 Maine is the most demanding, as I detailed in my class certification order, but no state dispenses with causation. Since the plaintiffs are unable to prove that every class member paid a higher transaction price by the common evidence they have gathered, I conclude that they fail to meet their burden of proof for each of the nineteen states. Conclusion Under the Supreme Court’s Celotex trilogy, 31 on a summary judgment motion the plaintiffs have the burden of coming forward with evidence on any issue where they bear the burden of proof. The plaintiffs have failed to meet that burden here with respect to proving antitrust impact— their claim that the alleged illegal horizontal agreement actually affected the price that each putative class member paid for his or her car. As a result, the defendants’ joint motion for summary judgment is Granted. The defendants’ individual motions for summary judgment, therefore, are Moot, as are the various motions to exclude expert testimony and the plaintiffs’ renewed motion for class certification. So Ordered. Footnotes [1] Defendants General Motors Corp. and Chrysler LLC (and its subsidiary Chrysler Motors LLC), filed for bankruptcy during the pendency of this case, and an automatic stay has issued. Notice of Bankruptcy: Chrysler LLC (Docket Item 1010); Notice of Bankruptcy: General Motors Corp. (Docket Item 1018). Accordingly, this decision is not applicable to those defendants. Their Canadian counterparts, however, are bound by this Order. See Letter from Joseph J. Tabacco, Jr., Counsel for Plaintiffs (June 29, 2009) (Docket Item 1022); Statement of Non-Settling, Non-Bankrupt Defs. (Docket Item 1023) (Defendants, including General Motors of Canada, Ltd. and Chrysler Canada Inc., “believe that the Court should decide all pending motions as to them, without regard to the bankruptcy filings of Chrysler LLC, Chrysler Motors LLC and General Motors Corporation”). [2] The plaintiffs allege violations of state antitrust laws in Arizona, Kansas, Maine, Michigan, Minnesota, Mississippi, Nevada, New Mexico, North Dakota, South Dakota, Tennessee, Vermont, West Virginia, and Wisconsin. They allege violations of state consumer protection laws in Arkansas, Idaho, Maine, Massachusetts, Nebraska, Nevada, New Hampshire, New Mexico, West Virginia and Vermont. [3] The plaintiffs point to policies such as manufacturers refusing to honor warranties on Canadian cars in the United States and discouraging dealers from installing odometers that measure miles rather than kilometers; mandating “no export” clauses in sales agreements between dealers and consumers and requiring Canadian dealers to conduct due diligence into whether potential customers were likely to export their cars out of Canada; withholding information about safety recalls from exporting customers; imposing disciplinary measures on Canadian dealers who sold to exporting customers; imposing a "chargeback,” a monetary penalty sometimes amounting to thousands of dollars, on the Canadian dealer who sold the car when a Canadian car was discovered in the United States; threatening to withhold inventory of desirable models from offending dealerships; and threatening to terminate dealerships that sold to exporters. The defendants say that all such policies were unilaterally adopted. [4] The MSRP represents the retail price presented to the public. The dealer invoice price represents the manufacturer-determined net wholesale price to dealers. Both the MSRP and the list dealer invoice price are determined annually by manufacturers and apply nationally. Pis.' Opposing Statement of Add’l Material Facts [Corrected] ("POSMF”) ¶¶ 7-10 (Docket Item 964) (superseding Docket Item 853-2). Pursuant to an order issued September 12, 2008 (Docket Item 953), the parties filed paper copies of relevant documerits previously electronically docketed with the court. I have relied on these paper copies in issuing this decision. See Am. Stipulation Re: Hyperlinked Versions and Paper Copies (Docket Item 975). [5] Actual sales prices vary according to individual negotiations between dealers and consumers. Defs.' Statement of Undisputed Material Facts ¶ 40 (Docket Item 961) (superseding Docket Item 730). [6] In my Order on Motion for Class Certification: Exemplar State Damage Classes, I suggested some of the differences in state law proof requirements, In re New Motor Vehicles Canadian Exp. Antitrust Litig., 235 F.R.D. 127 (D.Me.2006), but they do not affect the analysis here. [7] Because of my ruling on causation, I find it unnecessary to reach the issue of whether they have an adequate model to prove damages consumer by consumer. [8] Ford, Chrysler and General Motors do not distinguish between their Canadian and U.S. entities on this argument, so neither do I. Chrysler also argues that the plaintiffs admit that as to a number of restraints, which were discussed at CADC meetings, no joint action was ever taken. Specifically, the Chrysler defendants point out that: the manufacturers never developed or used a uniform protocol due diligence list, nor did Chrysler share its list with other manufacturers, Chrysler Defs.' Statement of Material Undisputed Facts in Supp. of Their Mot. for Summ. J. (“Chrysler SMF”) ¶221 (Docket Item 957) (superseding Docket Item 741-2); Pis.' [Corrected] Opposing Statement of Material Fact — Chrysler Defs. ¶ 221 (Docket Item 964) (superseding Docket Item 875); the manufacturers never agreed to create nor did they create a shared known exporter database, Chrysler SMF ¶ 226; Pis.’ [Corrected] Opposing Statement of Material Fact' — Chrysler Defs. ¶ 226; the manufacturers never developed an industry-wide non-export clause, nor did they uniformly use their own non-export clause in retail sales contracts, Chrysler SMF ¶¶ 239-40; Pis.’ [Corrected] Opposing Statement of Material Fact — Chrysler Defs. ¶¶ 239-40; the attendees did not reach an agreement to indemnify dealers who became engaged in legal action over export sales, Chrysler SMF ¶ 244; Pis.’ [Corrected] Opposing Statement of Material Fact — Chrysler Defs. ¶ 244; the November 11, 2002 meeting attendees did not agree to any other joint initiative regarding specific restraint at that meeting, Chrysler SMF ¶¶ 218, 252, 255; Pis.’ [Corrected] Opposing Statement of Material Fact — Chrysler Defs. ¶¶ 218, 252, 255. But there is nevertheless probably enough for a jury to find at least an informal agreement to restrain Canadian exports. [9] An illegal agreement can be proven by direct evidence, or it can be proven by circumstantial evidence, often necessary because people frequently do not document or voice their illegal agreements. But if the evidence is only of parallel behavior, the Supreme Court adds additional proof requirements to survive summary judgment. "To survive a motion for summary judgment ... a plaintiff seeking damages for a violation of § 1 must present evidence “that tends to exclude the possibility” that the alleged conspirators act *48 ed independently.” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 588, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986) (internal quotation omitted); accord Bell Atl. Corp. v. Twombly, 550 U.S. 544, 553-54, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007); In re Baby Food Antitrust Litig., 166 F.3d 112, 124 (3d Cir.1999). Evidence of "conduct [that is] as consistent with permissible competition as with illegal conspiracy does not, standing alone, support an inference of antitrust conspiracy.” Matsushita, 475 U.S. at 588, 106 S.Ct. 1348; see also Williamson Oil Co. v. Philip Morris USA, 346 F.3d 1287, 1300 (11th Cir.2003) ("Evidence that does not support the existence of a price fixing conspiracy any more strongly than it supports conscious parallelism is insufficient to survive a defendant's summary judgment motion.”). [10] In deciding whether to grant the defendants' motion, I state the facts in the light most favorable to the plaintiffs. The defendants urge me to conclude from the fact that the plaintiffs have chosen not to sue some manufacturers altogether or have dismissed other manufacturers from this action, that those manufacturers were innocent or that any restraints they had were merely unilateral or would not have been pursued absent an illegal horizontal agreement among others. The conclusion does not follow. There are other potential explanations for such strategic choices by the plaintiffs. [11] The parties treat the consumer protection claim as requiring the same elements, and so do I. See Pis.’ Opp’n to Defs.’ Mots, for Summ. J. [Corrected] at 61-62 (Docket Item 964) (superseding Docket Item 853) ("[Plaintiffs agree that the state antitrust and consumer protection laws of the ... 19 states do not permit proof of injury by inference alone.”). The defendants also rely on federal law for their analysis and, although the plaintiffs deal with individual state laws, they do not reject - the relevance or applicability of federal law cases. See Defs.’ Joint Mot. for Summ. J. at 13 (Docket Item 961) (superseding Docket Item 729) ("Under federal law ... each plaintiff must still prove that [the antitrust violation occurred] and that it did in fact cause him injury.... Similarly, under the law of every state at issue here, indirect-purchaser plaintiffs also must show that each member of the state classes suffered a legally cognizable injury.” (emphasis omitted)). I find it most straightforward to use the federal cases; the state cases for the remaining state law claims yield the same outcome. [12] The plaintiffs have not argued that they are entitled to change their strategy. [13] I recognize that the First Circuit made its statements about proof of impact while ruling on and vacating class certification, whereas I am ruling on summary judgment in the absence of a certified class. (At oral argument, there was court-caused confusion on what motions were to be argued. I agreed then to rule first on summary judgment before considering the renewed and fully-briefed class certification motion. Summ. J. Oral Argument Tr. 6:2-19, Mar. 6, 2009 (Docket Item 999).) Strictly speaking, therefore, I am ruling on only the individual claims of the named plaintiffs. But obviously with an eye to the law of the case on certification, the named plaintiffs have chosen not to produce individualized proof that their particular transaction prices were elevated by the conspiracy. See, e.g., POSMF ¶¶ 892-919; Pis.' First Supplemental Resps. to Defs.’ General Motors Corp. and General Motors of Canada, Ltd.’s Am. First Set of Interrogs. # 6 (Ex. 744 to Deck of Todd A. Seaver in Supp. of Pis.’ Opp’n to Defs.’ Mot. for Summ. J. (Docket Item 870)) (Docket Item 870-17). Presumably if they cannot get to a jury on class-wide evidence — i.e., evidence that will also support certification of a class — then pursuing the lawsuit for individual purchasers is not worth the candle because of the huge expense. Moreover, I must be faithful to the spirit of what the First Circuit said that the plaintiffs must do. Therefore, my summary judgment ruling considers only the class-wide impact evidence, and that is how the parties have presented this summary judgment dispute to me. My conclusion, as elaborated in the text of the decision, is that the plaintiffs’ class-wide proof is not sufficient on its own to demonstrate impact on particular individual plaintiffs, here the named plaintiffs. But I do point out that my ruling, in the absence of class certification, disposes of the claims of only the named plaintiffs. Theoretically, other car-buyers could sue the manufacturers unaffected by the law or facts of this decision. Practically speaking, of course, that development is unlikely because the individual amounts at stake are generally very small compared to the cost of litigating, and there may also be statute of limitations problems for such claims. [14] The First Circuit then went on to discuss the so-called Bogosian shortcut, which I had discussed in my 2006 Order, In re New Motor Vehicles Canadian Exp. Antitrust Litig., 235 F.R.D. at 138 n. 35, but observed that "plaintiffs disclaim any intent to rely on the Bogosian [v. Gulf Oil Corp., 561 F.2d 434 (3d Cir.1977)] model.” In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d 6, 29 (1st Cir.2008). [15] Professor Hall reached this conclusion by employing a standard econometric technique known as a "discrete choice model," and specifically the version of discrete choice *55 modeling known as the “nested-logit model.” Expert Report of Robert E. Hall on Impact & Class Damages ¶ 91, May 10, 2007 (“Hall May 10, 2007 Report”) (Ex. 711 to Pis.’ Opp’n to Defs.' Mots, for Summ. J.) (Docket Item 894-19). [16] Even in their class certification motion, the plaintiffs claim only that they can "demonstrate that, by a preponderance of the evidence, the overwhelming majority of class members overpaid for their automobiles due to the inflated negotiating range,” and that they can prove, "with class-wide evidence, that there was widespread injury to class members” Pis.' Renewed Mot. for Class Certification at 5 (Docket Item 964) (superseding Docket Item 898). [17] The plaintiffs assert that the conspiracy elevated the retail or list prices and that these *56 elevated prices moved the actual transaction prices. [18] The plaintiffs’ expert Prof. Hall has testified that his methodology cannot establish injury to every consumer. See, e.g., Robert E. Hall Dep. 1082-83, Oct. 2, 2007 (Ex. 640 to Pis.’ Opp'n to Defs.' Mots, for Summ. J.) (Docket Item 891-27) (admitting "I don’t think I could say with confidence that everybody paid more”); id. at 1116-17 (admitting that his damages calculations would potentially include class members who were not overcharged); id. at 1118 (confirming that he was not contending that every single transaction price was affected by the alleged conspiracy); see also supra note 13. [19] The NBER paper to which they refer is Meghan Busse, et al„ $1,000 Cash Back: The Pass-Through of Auto Manufacturer Promotion, 96 Am. Econ. Rev. 1253 (2006). [20] For example, the relative size of the importing and exporting countries is effectively reversed. In the U.S. experience, the importing market (the United States) is much larger than the exporting market (Canada). See Expert Report of Kevin M. Murphy ¶ 180, Oct. 31, 2007 (“Murphy Oct. 31, 2007 Report”) (Ex. F to Defs.' Joint Mot. to Exclude Opinions & Testimony of Robert E. Hall (Docket Item 959) (superseding Docket Item 728)) (Docket Item 788-6). In addition, Dr. Hall concedes that the export restraint addressed in the Order — manufacturers discouraging dealers from ordering new, authorized vehicles directly from dealers in other countries and selling them as new — would remain legal in the United States even in Dr. Hall’s but— for world. Hall Dep. 1071-72, Oct. 2, 2007. Conversely, numerous practices that the plaintiffs challenge — such as manufacturers' use of chargebacks and warranty restrictions — were not addressed by the Order’s parallel-trade provisions. See Robert E. Hall Dep. 987-88, Oct. 1, 2007 (Ex. 639 to Pis.' Opp’n to Defs.' Mots, for Summ. J.) (Docket Item 891-26). Finally, the defendants’ expert points out that regional price incentives to address localized price pressures are ever-present in the United States market. Murphy Oct. 31, 2007 Report ¶ 247. There are no analogous type of price incentives used in the United Kingdom. Id. Thus, it is problematic whether a nationwide price change in the United Kingdom says anything about whether there would be a nationwide price change in the United States — a much larger and geographically varied nation. [21] Among the reasons that inference alone is not enough, is the fact that car purchases generally are a product of individual negotiation and can often involve other issues such as trade-in allowances, below-market-rate financing, etc. [22] I am not sure whether the plaintiffs mean to argue that these First Circuit statements unfairly shift the burden of proof. See Pis.' Opp’n to Defs.’ Mots, for Summ. J. 70-71 (opposing the argument that because it is possible that a given class member would have paid the same price they fail on their causation claim). The plaintiffs object to the defendants’ suggestion that the plaintiffs’ burden is to “rule out the possibility” that a plaintiff could have paid the same price in the but-for world. See Defs.' Joint Mot. for Summ. J. at 25 n. 30. In any event, I follow the First Circuit statements and I do not believe that they shift the burden of proof, but only describe what the plaintiffs must do, given the method of proof they have chosen for this putative class action. [23] Leffel has been Nissan USA's Director of Pricing Strategy since April 2003, and before that Leffel worked in Nissan’s Pricing Department. The Leffel declaration recounts an instance relating to model year 2005 in which GM advertised an "employee pricing” incentive and cut U.S. prices. POSMF ¶ 886 (citing Decl. of Duane A. Leffel II 16, Mar. 19, *60 2008 (Ex. 726 to Pis.' Opp’n to Defs.’ Mots, for Summ. J.) (Docket Item 894-31)). Leffel explained that Ford and Chrysler followed, and "transaction prices ... for all of the models for [the] three competitors” were reduced. Leffel Decl. ¶ 16. This sworn statement, the plaintiffs assert, together with the NBER paper’s empirical study, makes it more likely than not that the elevated MSRPs during the class period were passed through to the transaction prices paid for all plaintiffs' purchases. In the same statement, however, Leffel makes it clear that in response to the employee pricing incentives employed by other manufacturers "Nissan USA offered some incentives on some models, but on all others it made no changes in price or incentives.” Id. [24] "[A] minimal increase in national pricing would not necessarily mean that all consumers would pay more. Too many factors play into an individual negotiation to allow an assumption — at least without further theoretical development — that any price increase or decrease will always have the same magnitude of effect on the final price paid.” In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d at 29. [25] "From an economist's perspective, these two types of promotions provide an interesting comparison. While the promotion payments are nominally directed to one party or the other, who ultimately receives the benefit of the promotion depends on the outcome of the price negotiation process.” Busse, et al., supra note 19, at 1253. [26] Ordinarily on summary judgment I would not deal with assertions like these that are not part of a statement of material facts but are filed instead in a brief on a different motion. However, the defendants' lawyer referred to them at oral argument. [27] Hall states: "I estimate the effect of the New Cars Order on list prices, not consumer transaction prices.” Prof. Hall's Resp. to Defs.’ Rebuttal Reports ¶ 228, Feb. 3, 2008 (Ex. 77 to Deck of Daniel Purcell in Supp. of Honda’s Mot. for Summ. J. (Docket Item 763)) (Docket Items 817-77 through 817-79). [28] Professor Hall also addresses “Common Impact” in paragraphs 299-307 of his February 3, 2008 Report, but those paragraphs are all addressed to his rejection of defense expert Professor Murphy’s criticisms, not any new analysis or evidence to support Professor Hall’s position. [29] That statement also dooms the plaintiffs’ argument that the NBER study shows that “there is a low likelihood of many transactions occurring without pass-through,” or that price changes "are passed through at extremely high rates.” Pis.’ Opp'n to Defs.’ Mots, for Summ. J. at 68. [30] "[P]laintiffs agree that the state antitrust and consumer protection laws of the other 19 states do not permit proof of injury by inference alone.” Id. at 61-62. [31] The Celotex trilogy of cases consists of Matsushita, 475 U.S. at 577, 106 S.Ct. 1348; Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247-48, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986); and Celotex Corp. v. Catrett, 477 U.S. 317, 323-25, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986).