Karen McLAUGHLIN, Jane Amodeo, David Tuttleman, Susan Bailey, Barbara Bishop, Trevor Campbell, Fergal Furlong, David Rogers, Barbara Schwab, Patricia Scocozza, and Jim Sherman, Plaintiffs-Appellees, v. AMERICAN TOBACCO COMPANY, Altria Group, Inc., Philip Morris USA Inc., Lorillard Tobacco Co, British American Tobacco Limited, Liggett Group, Inc., B.A.T. Industries P.L.C., and R.J. Reynolds Tobacco Co., Defendants-Appellants.
While redressing injuries caused by the cigarette industry is “one of the most troubling ․ problems facing our Nation today,” FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 125, 120 S.Ct. 1291, 146 L.Ed.2d 121 (2000), not every wrong can have a legal remedy, cf. Pearl v. City of Long Beach, 296 F.3d 76, 89 (2d Cir.2002), at least not without causing collateral damage to the fabric of our laws. Plaintiffs' putative class action suffers from an insurmountable deficit of collective legal or factual questions. Their claims are brought as based in fraud under the Racketeer Influenced and Corrupt Organizations Act (RICO), 18 U.S.C. §§ 1961-1968, but under RICO, each plaintiff must prove reliance, injury, and damages. Moreover, some undetermined number of plaintiffs' claims are time-barred. Rule 23 is not a one-way ratchet, empowering a judge to conform the law to the proof. We therefore reverse the order of the district court and decertify the class.
Plaintiffs, a group of smokers allegedly deceived-by defendants' marketing and branding-into believing that “light” cigarettes (“Lights”) were healthier than “full-flavored” cigarettes, sought and were granted class certification. Schwab v. Philip Morris USA, Inc., 449 F.Supp.2d 992 (E.D.N.Y.2006) (Jack B. Weinstein, Judge ). Plaintiffs' suit is brought under RICO, with mail and wire fraud as the necessary predicate acts. See 18 U.S.C. § 1962(c) (forbidding “any person employed by or associated with any enterprise engaged in, or the activities of which affect, interstate or foreign commerce, to conduct or participate, directly or indirectly, in the conduct of such enterprise's affairs through a pattern of racketeering activity”); see also id. § 1961(1) (providing that mail and wire fraud constitute racketeering activity); cf. id. § 1341 (mail fraud statute); id. § 1343 (wire fraud statute). The gravamen of plaintiffs' complaint is that defendants' implicit representation that Lights were healthier led them to buy Lights in greater quantity than they otherwise would have and at an artificially high price, resulting in plaintiffs' overpayment for cigarettes. Plaintiffs allege claims arising from their purchase of Lights from 1971, when defendants first introduced Lights, until the date on which trial commences.2
We pause in our narrative briefly to explain the history of Lights, as that history bears on plaintiffs' claims. In 1955, the Federal Trade Commission (FTC) adopted the “Cigarette Advertising Guides,” which proscribed “any implicit or explicit health claims in cigarette advertising ․ [except claims] that a cigarette was ‘low in nicotine or tars' provided it ha[d] ‘been established by competent scientific proof ․ that the claim [wa]s true, and if true, that such difference or differences [we]re significant.’ ” United States v. Philip Morris USA, Inc., 449 F.Supp.2d 1, 432 (D.D.C.2006).
Several years later, in 1967, the FTC introduced the “Cambridge Filter Method” for calculating tar and nicotine yield. The Cambridge Filter Method, however, which relies upon a machine to test the tar and nicotine content of cigarettes, is quite unreliable. Most smokers who smoke Lights obtain just as much tar and nicotine as they would if they smoked full-flavored cigarettes, principally by “compensating”-that is, either by inhaling more smoke per cigarette (e.g., by covering ventilation holes, drawing more deeply with each puff, etc.) or by buying more cigarettes, Schwab, 449 F.Supp.2d at 1094, neither of which a machine is capable of doing. Cigarette manufacturers have apparently been aware of this phenomenon for some time. See Philip Morris, 449 F.Supp.2d at 438; Aspinall v. Philip Morris Cos., 442 Mass. 381, 813 N.E.2d 476, 481 n. 9 (Mass.2004). But some smokers continued at least until 2000 to believe that Lights were healthier than full-flavored cigarettes. As the district court noted, citing a 1977 Brown & Williamson Internal Marketing Study, “[a]lmost all smokers agree that the primary reason for the increasing acceptance of [Lights] is based on the health reassurance they seem to offer.” Schwab, 449 F.Supp.2d at 1095 (internal quotation marks and citation omitted).
In 2001, however, the National Cancer Institute published a report, “Monograph 13,” that “review[ed] evidence on the FTC method for measuring tar and nicotine yields and the disease risks of machine-measured low-tar cigarettes.” J.A. at 855. The stated objective of the report was “to determine whether the evidence taken as a whole shows that the cumulative effect of engineering changes in cigarette design over the last 50 years has reduced disease risks in smokers.” Id. Monograph 13 discussed the introduction and marketing of low-yield cigarettes, the growing use of these cigarettes, and the practice of compensatory smoking. Ultimately, it concluded that there was “no convincing evidence that changes in cigarette design between 1950 and the mid 1980s have resulted in an important decrease in the disease burden caused by cigarette use either for smokers as a group or for the whole population.” Id. at 992. The publication of Monograph 13 sparked both this suit, filed in May 2004, and a parallel civil RICO action brought by the federal government.3
Plaintiffs seek $800 billion in economic damages (trebled) stemming from their purchases of Lights. On September 25, 2006, the district court certified their proposed class of Lights smokers. On November 16, 2006, this court stayed the proceedings below and granted defendants leave to take an interlocutory appeal under Federal Rule of Civil Procedure 23(f). We now reverse the district court's class certification order and decertify the class.
We review the district court's certification order for abuse of discretion. See Moore v. PaineWebber, Inc., 306 F.3d 1247, 1252 (2d Cir.2002). We will “exercise even greater deference when the district court has certified a class than when it has declined to do so.” Marisol A. by Forbes v. Giuliani, 126 F.3d 372, 375 (2d Cir.1997). However, as we recently made clear, “a district judge may not certify a class without making a ruling that each Rule 23 requirement is met ․ [and] all ․ evidence must be assessed as with any other threshold issue,” whether or not any such assessment also bears on the merits of the case. Miles v. Merrill Lynch & Co. (In re Initial Pub. Offerings Sec. Litig.), 471 F.3d 24, 27 (2d Cir.2006) [hereinafter In re IPO ] (emphasis added).
Rule 23(a) requires that a class action possess four familiar features: (1) numerosity; (2) commonality; (3) typicality; and (4) adequacy of representation. If those criteria are met, the district court must next determine whether the class can be maintained under any one of the three subdivisions of Rule 23(b). With respect to class actions for money damages sought under Rule 23(b)(3), the district court must also find that “questions of law or fact common to class members predominate over any questions affecting only individual members,” and that the class action “is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed.R.Civ.P. 23(b)(3).4 The primary issue in this case, to which we now turn, is whether the requirement that common questions predominate has been met. Because we answer this question in the negative, we need not address whether a class action is a superior method of adjudicating plaintiffs' claims.
I. Elements of a Civil RICO Claim and the Predominance Requirement
Section 1964(c) of Title 18 (“civil RICO”) gives private citizens a cause of action under RICO by providing that “[a]ny person injured in his business or property by reason of a violation of [RICO's substantive provisions] may sue therefor in any appropriate United States district court and shall recover threefold the damages he sustains and the cost of the suit, including a reasonable attorney's fee.” 18 U.S.C. § 1964(c). To fulfill the requirement that the injury occur “by reason of” a defendant's action, a plaintiff must show “that the defendant's violation not only was a ‘but for’ cause of his injury, but was the proximate cause as well.” Holmes v. Sec. Investor Prot. Corp., 503 U.S. 258, 268, 112 S.Ct. 1311, 117 L.Ed.2d 532 (1992); see also Commercial Cleaning Servs., L.L.C. v. Colin Serv. Sys., Inc., 271 F.3d 374, 380 (2d Cir.2001) (“RICO's use of the clause ‘by reason of’ has been held to limit standing to those plaintiffs who allege that the asserted RICO violation was the legal, or proximate, cause of their injury, as well as a logical, or ‘but for,’ cause.”). “But for” causation is also known as “transaction causation,” or “reliance,” while proximate causation is often referred to as “loss causation.” See, e.g., Moore v. PaineWebber, Inc., 189 F.3d 165, 169-70 (2d Cir.1999); Powers v. British Vita, P.L.C., 57 F.3d 176, 189-90 (2d Cir.1995); see also Dura Pharms., Inc. v. Broudo, 544 U.S. 336, 341, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005) (noting that reliance is “often referred to ․ as ‘transaction causation’ ”). Thus, a plaintiff asserting a civil RICO claim must be able to support allegations of (1) a RICO violation, (2) injury, and (3) transaction and loss causation. First Nationwide Bank v. Gelt Funding Corp., 27 F.3d 763, 769 (2d Cir.1994). In this case, to prevail in their argument for class certification, plaintiffs must establish that the issues of injury and causation do not defeat the predominance requirement of Rule 23(b)(3). For the reasons that follow, we find that plaintiffs have failed to meet this burden.
In cases such as this one when mail or wire fraud is the predicate act for a civil RICO claim, the transaction or “but for” causation element requires the plaintiff to demonstrate that he relied on the defendant's misrepresentation. See Caviness v. Derand Res. Corp., 983 F.2d 1295, 1305 (4th Cir.1993) (“[A] claim under RICO requires both reliance and damage proximately caused by the violation.”); cf. ATSI Commc'ns, Inc. v. Shaar Fund, Ltd., 493 F.3d 87, 106-07 (2d Cir.2007) (noting, in the securities fraud context, that “[a] plaintiff is required to prove both transaction causation (also known as reliance) and loss causation. Transaction causation only requires allegations that but for the claimed misrepresentations or omissions, the plaintiff would not have entered into the detrimental securities transaction.” (internal quotation marks and citations omitted)). In this case, the plaintiffs argue, and the district court agreed, that they can prove reliance on a class-wide basis, using generalized proof, because defendants conducted a national marketing campaign for Lights and therefore represented that Lights were healthier than full-flavored cigarettes in a “consistent, singular, uniform” fashion. Appellees' Br. at 15. In other words, plaintiffs underscore that defendants' fraud resulted from a common course of conduct and therefore argue that common issues predominate. To support their argument, plaintiffs invoke our discussion in Moore, in which we stated that “[f]raud actions must ․ be separated into two categories: fraud claims based on uniform misrepresentations made to all members of the class and fraud claims based on individualized misrepresentations. The former are appropriate subjects for class certification because the standardized misrepresentations may be established by generalized proof.” 306 F.3d at 1253.
But proof of misrepresentation-even widespread and uniform misrepresentation-only satisfies half of the equation; the other half, reliance on the misrepresentation, cannot be the subject of general proof. Individualized proof is needed to overcome the possibility that a member of the purported class purchased Lights for some reason other than the belief that Lights were a healthier alternative-for example, if a Lights smoker was unaware of that representation, preferred the taste of Lights, or chose Lights as an expression of personal style. See id. at 1255 (“In order to establish [defendant's] liability, each plaintiff must prove that he or she personally received a material misrepresentation, and that his or her reliance on this misrepresentation was the proximate cause of his or her loss.”); cf. Gunnells v. Healthplan Servs., Inc., 348 F.3d 417, 435 (4th Cir.2003) ( “[T]he reliance element of ․ fraud and negligent misrepresentation claims is not readily susceptible to class-wide proof; rather, proof of reasonable reliance ․ depends upon a fact-intensive inquiry into what information each plaintiff actually had.” (omissions in original) (alterations, internal quotation marks, and citation omitted)); Simon v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 482 F.2d 880, 882 (5th Cir.1973) (noting that if written misrepresentations do not actually reach members of the purported class, “they are no more valid a basis for a class action than dissimilar oral representations,” which cannot sustain a class action). We took account of this idea in Moore, when we recognized that, “[o]n the other hand, although having some common core, a fraud case may be unsuited for treatment as a class action if there was material variation ․ in the kinds or degrees of reliance by the persons to whom they were addressed.” 306 F.3d at 1253 (quoting Fed.R.Civ.P. 23(b)(3) Advisory Committee Notes).
Plaintiffs and the district court suggest that defendants distorted the body of public information and that, in purchasing Lights, plaintiffs relied upon the public's general sense that Lights were healthier than full-flavored cigarettes, whether or not individual plaintiffs were actually aware of defendants' alleged misrepresentation. Cf. Falise v. Am. Tobacco Co., 94 F.Supp.2d 316, 335 (E.D.N.Y.2000) (“Where ․ the fraudulent scheme is targeted broadly at a large proportion of the American public[,] the requisite showing of reliance is less demanding. Such sophisticated, broad-based fraudulent schemes by their very nature are likely to be designed to distort the entire body of public knowledge․”). Their argument invokes the fraud-on-the-market presumption set forth in Basic Inc. v. Levinson, 485 U.S. 224, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988), which concerned fraud claims in the securities context.5 “The fraud-on-the market doctrine ․ creates a rebuttable presumption that (1) misrepresentations by an issuer affect the price of securities traded in the open market, and (2) investors rely on the market price of securities as an accurate measure of their intrinsic value.” Hevesi v. Citigroup Inc., 366 F.3d 70, 77 (2d Cir.2004). Thus, a plaintiff alleging securities fraud may establish reliance simply by virtue of the defendant's public dissemination of misleading information. See Basic, 485 U.S. at 241-42, 108 S.Ct. 978 (noting that because the price of stock in an efficient market reflects all publicly available information, “[m]isleading statements will ․ defraud purchasers of stock even if the purchasers do not directly rely on the misstatements”).
We do not think that the Basic presumption, or the district court's variation of it, applies in this case; we cannot assume that, regardless of whether individual smokers were aware of defendants' misrepresentation, the market at large internalized the misrepresentation to such an extent that all plaintiffs can be said to have relied on it. Basic involved an efficient market-the market in securities traded on the New York Stock Exchange-capable of rapidly assimilating public information into stock prices, see id. at 247, 249 n. 29, 108 S.Ct. 978 (describing the securities market as “impersonal, well-developed,” and “information-hungry”); the market for consumer goods, however, is anything but efficient, cf. Sikes v. Teleline, Inc., 281 F.3d 1350, 1364 (11th Cir.2002) (“[E]ach individual plaintiff is the only person with information about the content of the advertisement upon which he relied.”). Indeed, the fact that the publication of Monograph 13 produced no change in either the sales or the price of Lights shows just how unresponsive the consumer market in Light cigarettes is to the advent of new information. See In re IPO, 471 F.3d at 43 (“Plaintiffs' own allegations as to how slow the market was to correct the alleged price inflation despite what they also allege was widespread knowledge of the scheme indicate the very antithesis of an efficient market.”). As we stated in In re IPO, “[w]ithout the Basic presumption, individual questions of reliance would predominate over common questions.” Id.; see also Gunnells, 348 F.3d at 435 (noting that Basic's presumption of actual reliance was based on the efficiency of capital markets, which did not apply to plaintiffs' purchase of health care plans, and that therefore actual reliance could not be presumed and individualized inquiry was required).
We need not go so far as to adopt the Fifth Circuit's blanket rule that “a fraud class action cannot be certified when individual reliance will be an issue,” Castano v. Am. Tobacco Co., 84 F.3d 734, 745 (5th Cir.1996), as some fraud actions do appear within the contemplation of Rule 23's drafters, see Fed.R.Civ.P. 23(b)(3) Advisory Committee Notes (“[A] fraud perpetrated on numerous persons by the use of similar misrepresentations may be an appealing situation for a class action, and it may remain so despite the need, if liability is found, for separate determination of the damages suffered by individuals within the class.”). But in this case, reliance is too individualized to admit of common proof.6
Plaintiffs suggest that regardless of whether reliance is susceptible to aggregate proof under Moore, they should be entitled to a presumption of reliance in light of the market shift in brand preferences (from nonfiltered to filtered to low tar cigarettes) that resulted from defendants' marketing of Lights. See Appellees' Br. at 24-25. While proof of reliance by circumstantial evidence may be sufficient under certain conditions,7 cf. Sikes, 281 F.3d at 1362 n. 32, it is insufficient here. Just as the Ninth Circuit explained in Poulos v. Caesars World, Inc. that people choose to gamble for any number of reasons, each plaintiff in this case could have elected to purchase light cigarettes for any number of reasons, including a preference for the taste and a feeling that smoking Lights was “cool.” See 379 F.3d 654, 665-66 (9th Cir.2004) (“[G]ambling is not a context in which we can assume that potential class members are always similarly situated. Gamblers do not share a common universe of knowledge and expectations-one motivation does not ‘fit all.’ ․ Thus, to prove proximate causation in this case, an individualized showing of reliance is required.”); Davies v. Philip Morris U.S.A., Inc., No. 04-2-08174-2 SEA, 2006 WL 1600067, at *3 (Wash.Super.Ct. May 26, 2006) (denying plaintiffs' motion for class certification because “[i]nescapably individual differences cannot be concealed in a throng․ Was the consumer motivated by health-related reasons, or ․ by taste, peer influence, price, habit, or simply personal preference? ․ [T]here are numerous non-health-related reasons that people buy light cigarettes.” (internal quotation marks and citation omitted)). Indeed, the fact that the market did not shift away from light cigarettes after the publication of Monograph 13 is compelling evidence that plaintiffs had other, non-health-related reasons for purchasing Lights. Three of the six named plaintiffs even continued to purchase Lights after filing the complaint in this case, suggesting the influence of some other motivation.
Moreover, we are not blind to the indeterminate likelihood that, even before the publication of Monograph 13, some members of plaintiffs' desired class were aware that Lights are not, in fact, healthier than full-flavored cigarettes, and they therefore could not have relied on defendants' marketing in deciding to purchase Lights. Cf. Sandwich Chef of Tex., Inc. v. Reliance Nat'l Indem. Ins. Co., 319 F.3d 205, 220 (5th Cir.2003) (“Knowledge that invoices charged unlawful rates, but did so according to a prior agreement between the insurer and the policyholder, would eliminate reliance and break the chain of causation.”); Moore v. Am. Fed'n of Television & Radio Artists, 216 F.3d 1236, 1243 (11th Cir.2000) (“Some found no error in the statements and accepted them at face value; some relied on their agents to monitor the statements, and they likewise found no error; others spotted what they perceived to be errors and complained to AFTRA․”). Thus, differences in plaintiffs' knowledge and levels of awareness also defeat the presumption of reliance.
2. Loss Causation
A plaintiff alleging a violation of civil RICO must also establish loss causation, meaning that the defendant's misrepresentations caused the plaintiff “to suffer economic loss.” Moore, 189 F.3d at 170; see also Anza v. Ideal Steel Supply Corp., 547 U.S. 451, 126 S.Ct. 1991, 1998, 164 L.Ed.2d 720 (2006) (“When a court evaluates a RICO claim for proximate causation, the central question it must ask is whether the alleged violation led directly to the plaintiff's injuries.”). “Furthermore, when factors other than the defendant's fraud are an intervening direct cause of a plaintiff's injury, that same injury cannot be said to have occurred by reason of the defendant's actions.” First Nationwide Bank, 27 F.3d at 769. In this case, plaintiffs' theory is that they suffered an economic loss because they were overcharged for Lights. Plaintiffs argue that defendants' misrepresentation that Lights were healthier led to an increased market demand for light cigarettes, which drove up the price of Lights. Thus, plaintiffs contend that they paid more for Lights than they otherwise would have had the truth been known. As with reliance, plaintiffs claim that they can establish loss causation on a class-wide basis.
This argument fails because the issue of loss causation, much like the issue of reliance, cannot be resolved by way of generalized proof. As we noted above, individuals may have relied on defendants' misrepresentation to varying degrees in deciding to purchase Lights; some may have relied completely, some in part, and some not at all. Thus, establishing the first link in the causal chain-that defendants' misrepresentation caused an increase in market demand-would require individualized proof, as any number of other factors could have led to this increase. If smokers purchased more light cigarettes and drove up demand for reasons unrelated to defendants' misrepresentation, plaintiffs could not show that their economic injury was directly caused by defendants' fraud. Cf. Anza, 126 S.Ct. at 1997 (“There is ․ a second discontinuity between the RICO violation and the asserted injury. [Plaintiff's] lost sales could have resulted from factors other than [defendant's] alleged acts of fraud. Businesses lose and gain customers for many reasons, and it would require a complex assessment to establish what portion of [plaintiff's] lost sales were the product of [defendant's] decreased prices.”).
Given the lack of an appreciable drop in the demand or price of light cigarettes after the truth about Lights was revealed in Monograph 13, plaintiffs' argument that defendants' misrepresentation caused the market to shift and the price of Lights to be inflated fails as a matter of law. We have stated that “[t]he key reasons for requiring direct causation include avoiding unworkable difficulties in ascertaining what amount of the plaintiff's injury was caused by the defendant's wrongful action as opposed to other external factors.” First Nationwide Bank, 27 F.3d at 770. Here, because factors other than defendants' misrepresentation may have intervened and affected the demand and price of Lights, and because determining the portion of plaintiffs' injury attributable to defendants' wrongdoing would require an individualized inquiry, plaintiffs cannot establish loss causation on a class-wide basis.
Plaintiffs also argue that the requisite injury to “business or property” is susceptible to class-wide proof. See 18 U.S.C. § 1964(c); Sedima, S.P.R.L. v. Imrex Co., 473 U.S. 479, 496, 105 S.Ct. 3275, 87 L.Ed.2d 346 (1985) (stating that a RICO plaintiff “only has standing if, and can only recover to the extent that, he has been injured in his business or property by the conduct constituting the violation”). In this case, proof of injury, or whether plaintiffs have been harmed, is bound up in proof of damages, or by how much plaintiffs have been harmed. Only by showing that plaintiffs paid more for light cigarettes than they would have but for defendants' misrepresentation can plaintiffs establish the requisite injury under civil RICO. Cf. Cordes & Co. Fin. Servs., Inc. v. A.G. Edwards & Sons, Inc., 502 F.3d 91, 107 (2d Cir.2007) (“If the fee paid were higher than the but-for fee, then the plaintiff suffered an injury-in-fact. In this case, the extent of the difference between the but-for fee and the actual fee paid is relevant to the question of damages, but it is from a comparison between the two that the court would be asked to decide the question of injury-in-fact.”). Plaintiffs have advanced two theories to support their claim of injury and how the “but for” price of Lights (and thus the resulting damages) might be calculated: the loss of value theory and the price impact model. However, because neither of these theories is plausible as a matter of law, because both would lead to an impermissible fluid recovery, and because the acceptable measure of injury-out-of-pocket damages-would require individualized proof, class-wide issues cannot be said to predominate.
A plaintiff asserting a claim under 18 U.S.C. § 1964(c) must allege actual, quantifiable injury. In First Nationwide Bank, a civil RICO case that, like the instant case, centered on the issues of causation and injury, we held that the assumption of additional risk of loss due to undersecured loans was insufficient to support an allegation of injury under RICO. See 27 F.3d at 767-68. We further stated that “[t]he general rule of fraud damages is that the defrauded plaintiff may recover out-of-pocket losses caused by the fraud.” Id. at 768; see also Commercial Union Assurance Co. v. Milken, 17 F.3d 608, 612 (2d Cir.1994) (“[D]amages as compensation under RICO § 1964(c) for injury to property must, under the familiar rule of law, place [plaintiffs] in the same position they would have been in but for the illegal conduct.”).
In this case, out-of-pocket losses cannot be shown by common evidence because they constitute an inherently individual inquiry: individual smokers would have incurred different losses depending on what they would have opted to do, but for defendants' misrepresentation. For example, smokers who would have purchased full-flavored cigarettes instead of Lights had they known that Lights were not healthier would have suffered no injury because Lights have always been priced the same as full-flavored cigarettes. By contrast, those who would have quit smoking altogether could recover their expenses in purchasing Lights. And those who would have continued to smoke, but in greater moderation, could recover something in between. Thus, on the issue of out-of-pocket loss, individual questions predominate; plaintiffs cannot meet their burden of showing that injury is amenable to common proof.
Plaintiffs, no doubt recognizing the above difficulties with certifying a class claiming out-of-pocket losses, offer two other theories of recovery, but neither is cognizable under RICO. “In re IPO makes clear that courts may resolve contested factual issues where necessary to decide on class certification, and when a claim cannot succeed as a matter of law, the Court should not certify a class on that issue.” Velez v. Novartis Pharms. Corp., 244 F.R.D. 243, 257 (S.D.N.Y.2007). Thus, plaintiffs' alternative theories cannot support their argument for class certification.
1. Loss of Value
The “loss of value” model purports to measure the difference between the price plaintiffs paid for light cigarettes as represented by defendants and the (presumably lower) price they would have paid (but for defendants' misrepresentation) had they known the truth-that Lights are not healthier than full-flavored cigarettes. Schwab, 449 F.Supp.2d at 1163. Plaintiffs' expert “estimated a loss in value of 77.7 percent if the ‘light’ cigarettes sold by defendants were as harmful as regular cigarettes.” Id. at 1057-58.
But the loss of value model is designed to award plaintiffs damages based on the benefit of their bargain. Such damages are generally unavailable in RICO suits. See Commercial Union, 17 F.3d at 612; 2 McLaughlin on Class Actions § 8:16, at 8-102 (3d ed. Dec. 2006 update). This is a sensible rule, and not derived from our “loss causation” cases, as the district court suggested. Schwab, 449 F.Supp.2d at 1065. Rather, the rule of general unavailability follows from the text of RICO, which compensates only for injury to “business or property.” 18 U.S.C. § 1964(c).
Indeed, plaintiffs have not explained how a party's “expectation” can constitute “business or property.” See Heinhold v. Perlstein, 651 F.Supp. 1410, 1412 (E.D.Pa.1987) (holding that a plaintiff does not have standing to sue under RICO when “the only property to which a plaintiff alleges injury is an expectation interest that would not have existed but for the alleged RICO violation”); cf. Oscar v. Univ. Students Coop. Ass'n, 965 F.2d 783, 785 (9th Cir.1992) (en banc) (“[A] showing of ‘injury’ requires proof of concrete financial loss, and not mere ‘injury to a valuable intangible property interest.’ ” (emphasis added) (citation omitted)). While we need not and do not decide whether expectancy damages are ever available under RICO, cf. Fleischhauer v. Feltner, 879 F.2d 1290, 1300 (6th Cir.1989), in cases that sound in fraud in the inducement, they plainly are not.8 Here, the cigarette packs that plaintiffs purchased all bore the Surgeon General's warning that cigarettes may cause cancer and other diseases. Defendants' misrepresentation could in no way have reduced the value of the cigarettes that plaintiffs actually purchased; they simply could have induced plaintiffs to buy Lights instead of full-flavored cigarettes. See Appellants' Reply Br. at 19 (“[P]laintiffs' alleged lost expectation of a safer cigarette arose only because of the alleged fraud.”).
Moreover, even if benefit of the bargain damages could be awarded, there is no reasonable means of calculating them in this case. Cf. Fleischhauer, 879 F.2d at 1300 (“[T]here was no realistic evidence presented as to a reasonable value or estimate of lost profits or of the ‘bargain’ based on analogy, experience, or practice.”). We are asked to conceptualize the impossible-a healthy cigarette-and then to imagine what a consumer might have paid for such a thing. Indeed, Dr. Jeffrey Harris, proponent of the loss of value theory, asked survey respondents to “make binary comparisons between a ‘genuine’ light cigarette that reduced risk ․ and a ‘misrepresented’ light cigarette that was no less harmful than conventional cigarettes.” Schwab, 449 F.Supp.2d at 1164. He concluded that “virtually all respondents reported a non-zero loss in value.” Id. While the district court is quite correct that “damages need not [usually] be demonstrated with precision,” id. at 1065, plaintiffs' theory is pure speculation, as the survey response to Dr. Harris exemplifies.
Thus, plaintiffs' legal theory fails at its inception and, even if it did not, plaintiffs have not made the requisite showing under In re IPO that they could, at trial, marshal facts sufficient to permit them to rely upon it. See In re IPO, 471 F.3d at 40 (asserting that, in determining whether the Rule 23 requirements for class certification have been met, it is appropriate for the judge to “resolve[ ] underlying factual disputes,” and that “as to these disputes, the judge must be persuaded that the fact at issue has been established”); id. at 41 (noting that even when there is overlap between a Rule 23 requirement and a merits issue, “the district judge must receive enough evidence, by affidavits, documents, or testimony, to be satisfied that each Rule 23 requirement has been met”).
2. Price Impact
Plaintiffs also assert a damages theory based on an estimate of the “price impact” that a disclosure that Lights were not safer than full-flavored cigarettes would have had on the market. Using multiple regression analysis, plaintiffs seek to show the amount by which defendants would have had to reduce their prices to account for the concomitant reduced demand. Even if that amount could be proven by common evidence, as with the loss of value model, plaintiffs have failed as a matter of law to adduce sufficient facts to show that the price impact model is a tenable measure of harm. Cf. supra Part I.B.1 (discussing In re IPO ).
Indeed, plaintiffs have not come forward with any meaningful means of estimating how the market has changed or might change in the future in response to fluctuations in the demand for light cigarettes. For instance, as we have already noted, Lights have always been priced the same as full-flavored cigarettes. Furthermore, if plaintiffs' theory of a health-driven preference for Lights were correct, one would have expected demand to drop following the publication of Monograph 13 as people returned to smoking regular cigarettes or quit smoking altogether and, correspondingly, prices to fall. But nothing of the sort happened; the market did not shift appreciably following the publication of Monograph 13. Cf. In re Burlington Coat Factory Sec. Litig., 114 F.3d 1410, 1425 (3d Cir.1997) (Alito, J.) (“[B]ecause the July 29 disclosure had no effect on BCF's price, it follows that the information disclosed on September 20 was immaterial as a matter of law.”).
The price impact model exemplifies the kind of vague inquiry into damages that the Supreme Court forbade in Anza. In that case, the plaintiff (a steel products vendor) sued a competitor, alleging that the competitor's practice of tax evasion had permitted it to charge lower prices than the plaintiff. The Supreme Court concluded that the plaintiff had failed adequately to plead a RICO violation because the injury it had suffered to its business was too remote from the predicate racketeering acts. See Anza, 126 S.Ct. at 1997. Specifically, the Court stated that “[t]he cause of [plaintiff's] asserted harms ․ is a set of actions (offering lower prices) entirely distinct from the alleged RICO violation (defrauding the State),” and noted that the “attenuation” was “clear.” Id. Importantly, the Court discussed
the speculative nature of the proceedings that would follow if [plaintiff] were permitted to maintain its claim. A court considering the claim would need to begin by calculating the portion of [defendant's] price drop attributable to the alleged pattern of racketeering activity. It next would have to calculate the portion of [plaintiff's] lost sales attributable to the relevant part of the price drop.
Id. at 1998. Similarly, in this case, a court considering plaintiffs' price impact model would have to engage in a series of speculative calculations to ascertain whether, and in what amount, plaintiffs suffered a loss.
Plaintiffs in this case argue that “[u]nlike in Anza ․ the class members here are the ‘immediate victims' of the RICO violation, and the causal chain is no more ‘attenuated’ than in any case in which an economist calculates an overcharge resulting from a defendant's unlawful activities.” Appellees' Br. at 43. But Anza spoke not only to the remoteness of the action that had allegedly caused the plaintiff's harm, but also to the possibility that damages could have resulted from factors unrelated to the defendant's alleged acts of fraud. See 126 S.Ct. at 1997 (“Businesses lose and gain customers for many reasons, and it would require a complex assessment to establish what portion of [plaintiff's] lost sales were the product of [defendant's] decreased prices.”). And indeed, here, as plaintiffs' expert concedes, a number of exogenous variables bear on cigarette price, including “rates of cigarette consumption, income levels of smokers, population, taxes, advertising expenditures, production costs, and plaintiffs' knowledge of health risks.” Schwab, 449 F.Supp.2d at 1058.
Thus, plaintiffs cannot show injury due to overall price impact on a class-wide basis and thereby satisfy Rule 23's predominance requirement because their price impact theory, like their loss of value theory, fails as a matter of law. Under In re IPO, plaintiffs must produce persuasive facts at trial that will enable them to prove injury to business or property under RICO. They have failed to persuade us that they can do so.
II. Calculation of Damages
The district court concluded that plaintiffs could prove collective damages on a class-wide basis, and individual plaintiffs would then claim shares of this fund:
First, defendant's aggregate liability is determined in a single, class-wide adjudication and paid into a class fund. Second, “individual class members are afforded an opportunity to collect their individual shares,” usually through a simplified proof of claim procedure.9 Third, any residue remaining after individual claims have been paid is distributed to the class' benefit under cy pres or other doctrines.
Id. at 1254 (citations omitted). But such “fluid recovery” has been forbidden in this circuit since Eisen v. Carlisle & Jacquelin, 479 F.2d 1005, 1008 (2d Cir.1973) (“[N]o ‘fluid recovery’ procedures are authorized by the text or by any reasonable interpretation of amended Rule 23.”), vacated on other grounds, 417 U.S. 156, 94 S.Ct. 2140, 40 L.Ed.2d 732 (1974). And while the fact that damages may have to be ascertained on an individual basis is not, standing alone, sufficient to defeat class certification, see Wal-Mart Stores, Inc. v. Visa U.S.A. Inc. (In re Visa Check/MasterMoney Antitrust Litig.), 280 F.3d 124, 140 (2d Cir.2001) (“The predominance requirement calls only for predominance, not exclusivity, of common questions.” (internal quotation marks and citation omitted)); 6 Alba Conte & Herbert B. Newberg, Newberg on Class Actions § 18:27 (4th ed. 2002) (“A particularly significant aspect of the Rule 23(b)(3) approach is the recognition that individual damages questions do not preclude a Rule 23(b)(3) class action when the issue of liability is common to the class.”), it is nonetheless a factor that we must consider in deciding whether issues susceptible to generalized proof “outweigh” individual issues.
We reject plaintiffs' proposed distribution of any recovery they might receive because it offends both the Rules Enabling Act and the Due Process Clause. The distribution method at issue would involve an initial estimate of the percentage of class members who were defrauded (and who therefore have valid claims). The total amount of damages suffered would then be calculated based on this estimate (and, presumably, on an estimate of the average loss for each plaintiff). But such an aggregate determination is likely to result in an astronomical damages figure that does not accurately reflect the number of plaintiffs actually injured by defendants and that bears little or no relationship to the amount of economic harm actually caused by defendants. This kind of disconnect offends the Rules Enabling Act, which provides that federal rules of procedure, such as Rule 23, cannot be used to “abridge, enlarge, or modify any substantive right.” 28 U.S.C. § 2072(b).
Roughly estimating the gross damages to the class as a whole and only subsequently allowing for the processing of individual claims would inevitably alter defendants' substantive right to pay damages reflective of their actual liability. See, e.g., In re Hotel Tel. Charges, 500 F.2d 86, 90 (9th Cir.1974) (rejecting a fluid recovery argument because “allowing gross damages by treating unsubstantiated claims of class members collectively significantly alters substantive rights,” in violation of the Rules Enabling Act); Eisen, 479 F.2d at 1019 (“[P]ossible recoveries run into astronomical amount [and] generate more leverage and pressure on defendants to settle․”); Schwab, 449 F.Supp.2d at 1272 (“A question under the Rules Enabling Act is posed by the danger of overcompensation inherent in the plaintiff's fluid distribution plan. It is possible that some claimants will benefit from the plaintiff class' recovery despite the fact that they did not rely on defendants' alleged misrepresentations regarding ‘light’ cigarettes and were not, therefore, injured in their business or property by defendants' actions.”). We disagree with the district court's conclusion that “[t]he risk of ․ overcompensation can be limited by requiring proof through claim forms from claimants concerning the extent of their reliance during the distribution stage.” Schwab, 449 F.Supp.2d at 1272. Given that any residue would be distributed to the class's benefit on the basis of cy pres principles rather than returned to defendants, defendants would still be paying the inflated total estimated amount of damages arrived at under the first step of the fluid recovery analysis. See id. at 1254. Thus, even if defendants were able to avoid overcompensating individual plaintiffs, they would still be overpaying in the aggregate.
Moreover, in this case, the district court determined that “evidence of the percentage of the class which was defrauded and the amount of economic damages it suffered appears to be quite weak.” Id. at 1021. It further concluded that “determin[ing] the impact of the fraud on the size of the market and its nature for damage purposes is a daunting enterprise even with the many proffered experts holding up their statistical lanterns to help in the search for the truth.” Id. Nevertheless, the district court believed that “the proof of acts of defendants and the various experts' opinions permit[ ] a finding of damages to the class with sufficient precision to allow a jury award.” Id. at 1137. For the reasons stated above, we disagree, and we further note our skepticism that if statistical experts cannot with accuracy estimate the relevant figures, a jury could do so based on the testimony of those experts.
The district court's distribution scheme also raises serious due process concerns. As we explained in Eisen,
if the ‘class as a whole’ is or can be substituted for the individual members of the class as claimants, then the number of claims filed is of no consequence and the amount found to be due will be enormous․ Even if amended Rule 23 could be read so as to permit any such fantastic procedure, the courts would have to reject it as an unconstitutional violation of the requirement of due process of law.
479 F.2d at 1018. When fluid recovery is used to permit the mass aggregation of claims, the right of defendants to challenge the allegations of individual plaintiffs is lost, resulting in a due process violation. The Third Circuit properly observed in Newton v. Merrill Lynch, Pierce, Fenner & Smith, Inc. that “actual injury cannot be presumed, and defendants have the right to raise individual defenses against each class member.” 259 F.3d 154, 191-92 (3d Cir.2001); see also 2 McLaughlin on Class Actions § 8:16, at 8-95 (3d ed. Dec.2006 update) (“Courts have repeatedly rejected the use of fluid recovery as a substitute for individualized proof when the class pursues claims that require proof of actual damages.”). To be sure, this does not mean that defendants are “constitutionally entitled to compel a parade of individual plaintiffs to establish damages.” In re Antibiotic Antitrust Actions, 333 F.Supp. 278, 289 (S.D.N.Y.1971). However, when fluid recovery is used, as here, to mask the prevalence of individual issues, it is an impermissible affront to defendants' due process rights. Cf. Six (6) Mexican Workers v. Ariz. Citrus Growers, 904 F.2d 1301, 1305-06 (9th Cir.1990) (dismissing concerns about fluid recovery because, in the case before it, “[t]he district court did not use fluid recovery to avoid individual proof of damages”).
III. The Statute of Limitations Defense
As with the difficulty in calculating damages, the presence of individual defenses does not by its terms preclude class certification. Augustin v. Jablonski (In re Nassau County Strip Search Cases), 461 F.3d 219, 225 (2d Cir.2006). But in this case, there is no doubt that a substantial number of class members were on notice of defendants' alleged fraud before the class period. Such a finding counsels in favor of vacating the district court's class certification order.10 See Waste Mgmt. Holdings, Inc. v. Mowbray, 208 F.3d 288, 295 (1st Cir.2000) (“[A]ffirmative defenses should be considered in making class certification decisions.”).
The statute of limitations for a civil RICO claim is four years. Agency Holding Corp. v. Malley-Duff & Assocs., Inc., 483 U.S. 143, 156, 107 S.Ct. 2759, 97 L.Ed.2d 121 (1987). The statute begins to run when the plaintiff discovers-or should reasonably have discovered-the alleged injury. Rotella v. Wood, 528 U.S. 549, 553-54, 120 S.Ct. 1075, 145 L.Ed.2d 1047 (2000). As the district court noted, “a troubling critical problem for plaintiffs is that some members of the class almost certainly were aware long before 2000 that ‘light’ cigarettes were not appreciably safer for them than regular cigarettes.” Schwab, 449 F.Supp.2d at 1069. The district court and plaintiffs rely upon the argument that “the tobacco companies made tremendous efforts to keep the truth about ‘light’ cigarettes from smokers and so should be estopped from arguing that the smokers learned the truth anyhow.” Id. at 1075. They point us to the Fifth Circuit's decision in Bratcher v. National Standard Life Insurance Co. (In re Monumental Life Insurance Co.), which concluded that “[t]o hold that each class member must be deposed as to precisely when, if at all, he learned of defendants' practices would be tantamount to adopting a per se rule that civil rights cases involving deception or concealment cannot be certified outside a two- or three-year period.” 365 F.3d 408, 420 (5th Cir.2004); see also Winoff Indus., Inc. v. Stone Container Corp. (In re Linerboard Antitrust Litig.), 305 F.3d 145, 162-63 (3d Cir.2002); Mowbray, 208 F.3d at 296.
The Supreme Court's recent decision in Ledbetter v. Goodyear Tire & Rubber Co. casts doubt upon In re Monumental. 550 U.S. 618, 127 S.Ct. 2162, 2177, 167 L.Ed.2d 982 (2007) (holding that statute of limitations barred suit for wage discrimination despite difficulty in discovering such discrimination). But even assuming that In re Monumental remains sound, we are not persuaded that it is on point. In that case, the court determined that “a presumption of unawareness by the plaintiff class is warranted.” In re Monumental, 365 F.3d at 420 n. 22; see also id. at 420 (“Of the thirteen representative plaintiffs, defendants point to only one, Jo Ella Brown, whose claim may have expired because of actual knowledge of defendants' practices.”). Such a presumption is unwarranted in this case.
First, as defendants note, two class representatives in this case appear to have understood the phenomenon of compensation-and its attendant risks-prior to May 2000 (four years before the complaint was filed). Appellants' Br. at 37 n. 13. Second, the minimal impact that the publication of Monograph 13 had on the market for Lights suggests that Monograph 13 may have been a reinterpretation of existing studies, as defendants argue, cf. Schwab, 449 F.Supp.2d at 1074, rather than a ground-breaking new study, as plaintiffs would have it. Third, plaintiffs' own attorneys filed several similar lawsuits prior to May 2000. Id. at 1071. Finally, and most importantly, plaintiffs have offered no reliable means of collectively determining how many class members' claims are time-barred. Id. at 1069.
* * *
In sum, because we find that numerous issues in this case are not susceptible to generalized proof but would require a more individualized inquiry, we conclude that the predominance requirement of Rule 23 has not been satisfied. We recognize that a court may employ Rule 23(c)(4) to certify a class as to common issues that do exist, “regardless of whether the claim as a whole satisfies Rule 23(b)(3)'s predominance requirement.” In re Nassau County Strip Search Cases, 461 F.3d at 227. Nevertheless, in this case, given the number of questions that would remain for individual adjudication, issue certification would not “reduce the range of issues in dispute and promote judicial economy.” Robinson v. Metro-N. Commuter R.R., 267 F.3d 147, 168 (2d Cir.2001). Certifying, for example, the issue of defendants' scheme to defraud, would not materially advance the litigation because it would not dispose of larger issues such as reliance, injury, and damages. See id. at 167 n. 12. We therefore decline plaintiffs' request for issue certification.
For the foregoing reasons, we Reverse the judgment of the district court and order the class Decertified.
1. For an exhaustive description of this case, we refer the interested reader to the district court's class-certification opinion below.
2. We assume for purposes of this decision that defendants represented that Lights were “healthier” than full-flavored cigarettes, rather than that Lights were simply lower in tar and nicotine. Cf. Schwab, 449 F.Supp.2d at 1127. Indeed, it makes little sense to argue that defendants' tar- and nicotine-content representations were untrue or deceptive. Cf. Brown v. Brown & Williamson Tobacco Corp., 479 F.3d 383, 392 (5th Cir.2007) (noting that “the Manufacturers are essentially forbidden from making any representations as to the tar and nicotine levels in their marketing about tar that are not based on the FTC method,” and concluding that “[t]he terms ‘light’ and ‘lowered tar and nicotine’ cannot, therefore, be inherently deceptive or untrue”). However, we note that this assumption tends to undermine the importance plaintiffs attach to the National Cancer Institute's publication of “Monograph 13,” a report that described the phenomenon of “compensation,” J.A. at 855, and in particular the practice of drawing more smoke from individual cigarettes; compensation potentially makes deceptive defendants' contention that Lights are low in tar and nicotine, but it does not truly speak to whether or not they are healthier.
3. The District Court for the District of Columbia recently found cigarette manufacturers liable for violating RICO, concluding that there was “overwhelming evidence” of defendants' intentional use of deceptive brand descriptors to induce smokers to purchase Lights. Philip Morris, 449 F.Supp.2d at 27.
4. All parties concede that this suit is not susceptible to certification under Rule 23(b)(1) or (2).
5. This is so despite plaintiffs' contention that they “are not advocating the same ‘fraud-on-the-market’ presumption applicable in a securities case.” Appellees' Br. at 25 n. 19; see id. at 4 (“Based on a belief in the truth of the representation, the market shifted.”).
6. Plaintiffs' effort to produce class-wide proof of reliance reinforces the difficulty of coming up with such proof. Plaintiffs' expert, Dr. John R. Hauser, claimed that 90.1% of those who smoked Lights chose to do so because of Lights' alleged health benefits. Schwab, 449 F.Supp.2d at 1167-68. But Dr. Hauser came to this conclusion on the basis of a method that determined whether, all things being equal, consumers prefer a safer cigarette to a less safe cigarette. And as plaintiffs conceded at oral argument, no one who understood this question would prefer a more dangerous product to a safer one.
7. For instance, payment may constitute circumstantial proof of reliance upon a financial representation. See, e.g., Westways World Travel, Inc. v. AMR Corp., 218 F.R.D. 223, 238 (C.D.Cal.2003) (“Class-wide reliance and injury may be established by virtue of payments class members made to American.”); Chisolm v. TranSouth Fin. Corp., 194 F.R.D. 538, 561 (E.D.Va.2000) (presuming that plaintiffs who paid deficiency judgments “made payments in reliance upon the assurance that the process of repossession, sale and all subsequent steps were taken in conformity with the law”). But a financial transaction does not usually implicate the same type or degree of personal idiosyncratic choice as does a consumer purchase.Klay v. Humana, Inc., 382 F.3d 1241 (11th Cir.2004), is distinguishable on this basis. In Klay, the court concluded that it did “not strain credulity to conclude that each plaintiff, in entering into contracts with the defendants, relied upon the defendants' representations and assumed they would be paid the amounts they were due.” Id. at 1259. But assuming that most individuals are led to believe that they will get paid when they sign a contract calling for payment is very different from assuming that most individuals purchase a consumer good in reliance upon an inference that they draw from its marketing and branding rather than for some other reason.
8. In Heinold, for instance, the defendant misrepresented the value of a diamond ring ultimately purchased by the plaintiff. The plaintiff did not overpay-the diamond ring was presumably worth its purchase price-but he did not make his expected profit. 651 F.Supp. at 1411. As the Sixth Circuit subsequently explained, “the fraud [did] not simultaneously induce plaintiffs to invest and reduce the value of the investment's object.” Fleischhauer, 879 F.2d at 1300 (emphasis added) (citing Heinold ).
9. “Damages would be allocated among class members based on the number of ‘light’ cigarettes purchased by each within the relevant geographic area and time.” Id. at 1252.
10. We speak here only of actual notice, as constructive notice is an issue susceptible to common proof; what a “reasonable person” would have known, and when, can be proven on a class-wide basis. Cf. Lanza v. Merrill Lynch & Co. (In re Merrill Lynch Ltd. P'ships Litig.), 154 F.3d 56, 60 (2d Cir.1998).
JOHN M. WALKER, JR., Circuit Judge: