WIENER and COSTA, Circuit Judges, joined by STEWART, Chief Judge, and DAVIS, SMITH, DENNIS, PRADO, ELROD, SOUTHWICK, GRAVES, HIGGINSON, Circuit Judges:

The Plaintiffs-Appellees brought a civil action under the Racketeer Influenced and Corrupt Organizations Act ("RICO"), 18 U.S.C. §§ 1961-68, alleging that Stream Energy, through its multi-level marketing program, Ignite, as well as a number of other defendants, (collectively the "Defendants") operated a fraudulent pyramid scheme. The Plaintiffs allege that the fraud has caused them financial losses. The district court certified a class of plaintiffs (the "Plaintiffs"), comprising those who lost money participating as Independent Associates ("IAs") in Ignite's program. We now review that certification en banc.

I.

Stream Energy sells gas and electricity to customers in Texas, Georgia, Pennsylvania, Maryland, New Jersey, New York, and the District of Columbia. Ignite is the marketing arm of Stream. Although Stream sells energy to customers, it is not a public utility that directly produces energy by owning the energy-producing infrastructure. Instead, it acts more as a middleman, reselling gas and electricity in deregulated energy markets that it buys from actual utilities. According to the Plaintiffs, Stream has realized only small profits on its energy sales, despite its large revenues, because Stream sells energy just above, or sometimes even at, its costs.

Rather than making meaningful profits through its sales, the Plaintiffs contend that Stream is set up like a classic pyramid scheme to make almost all of its money through the recruitment of salespeople. According to the Plaintiffs, it works like this: Stream's marketing arm, Ignite, operates a multi-level marketing program in which IAs (1) sell energy to customers, and (2) recruit other individuals to join as IAs who in turn sell energy to customers and recruit individuals to join as IAs. Under the IA program, Ignite charges individuals for the right to sell Stream services to customers and to recruit IAs. An IA pays Ignite $329 up front for the right to sell Stream energy and to recruit IAs, and also pays an optional recurring fee for a "Homesite" website that the IA can use to promote his or her Stream business.1 The putative class members are those individuals who paid to become IAs and lost money.

For each energy customer recruited, Ignite pays the IA a small percentage of that customer's bill as a commission, known as "Residual Income" or "Monthly Energy Income" ("MEI"). According to the Plaintiffs, however, the far more lucrative opportunities come from the recruitment of other IAs. Ignite pays IAs "Leadership Income" for recruiting other IAs. When an IA recruits another IA, he or she receives income from both (1) energy sales by that IA and his downline IAs, and (2) recruitment of other IAs by that IA and his downline IAs.

An IA's success depends primarily on recruiting a "downline" of other IAs who, in turn, recruit other IAs and customers into the Ignite program. As an IA recruits more IAs, he proceeds up a ladder of Ignite leadership positions. All IAs start out as "Directors," the lowest level of Ignite leadership. By recruiting more IAs, an IA can move up three additional leadership levels, first to "Managing Director," then to "Senior Director," and finally to "Executive Director." By building a downline, the IA also receives MEI for customers whom the downline IAs recruit to join Stream, along with bonuses for the recruitment of IAs both by the first IA and his downline IAs.

Ignite also promotes a "3&10 program." Under this program, Ignite pays an IA a $100 bonus if the IA enrolls four customers in the first 30 days. An IA can substitute purchase of the Homesite for two customers, and can be his or her own first customer, in which case that IA needs to recruit only one other customer to receive this bonus. Ignite offers an additional $100 bonus if the IA can obtain six additional customers within sixty days, and a $100 bonus for the first three new IAs that an IA recruits. If an IA recruits another IA who in turn enrolls four customers in his or her first thirty days, Ignite will pay the first IA a third $100 bonus. If the IA recruits two IAs and those recruits each enroll four customers in their first thirty days, Ignite will pay two more $100 bonuses. Ignite calls this the "3&10 program" because it requires an IA to recruit three new IAs and ten new customers (or seven if the IA purchased the Homesite and enrolls his or herself as a customer).

Over time, Stream's market has become saturated, and the Plaintiffs claim that they have lost money as a result of their participation in the IA program. The Plaintiffs allege that over 86% of individuals who signed up as IAs lost money in fees, collectively losing over $87 million. In contrast, a miniscule number of individuals have made significant sums of money.

This suit was brought by former IAs Juan Ramon Torres and Eugene Robison, who allege that Stream, Ignite, and various individual defendants have violated RICO. They sought to certify a class consisting of those IAs who have lost money as a result of participating in Ignite's program. The Plaintiffs sought certification under different theories.

The first was that the Defendants' common marketing materials were replete with fraudulent misstatements about how lucrative becoming an IA could be, and that — because all class members saw at least one of these statements — the Plaintiffs could show that their injuries arise from a common set of frauds. This theory did not require the Plaintiffs to prove that Ignite is a pyramid scheme; instead, it required only proof of specific misrepresentations.

But they also sought certification under theories that would require the Plaintiffs to prove that Ignite is a pyramid scheme. If they could prove that illegal conduct — and everyone acknowledges that the liability question is common to all class members — then the Plaintiffs contended that they did not need to identify specific misrepresentations on which particular class members relied, as individual reliance is not an element of a RICO claim. Instead, the Plaintiffs contended that RICO's causation requirement could be satisfied by classwide proof that their joining Ignite was a direct and foreseeable result of the Defendants' engaging in a pyramid scheme. Proximate cause could also be shown, they argued, through a common sense inference that they were duped into joining the pyramid scheme based on the representation that Ignite is a legitimate enterprise.

In response, the Defendants asserted primarily that the predominance requirement of Federal Rule of Civil Procedure 23(b)(3) is not met because individual issues of reliance will necessarily lead to an individualized causation inquiry under RICO. They also disagreed with the Plaintiffs' arguments that reliance is not a required element under RICO.

The district court rejected class certification on the Plaintiffs' theory that depends on specific misrepresentations, concluding that whether the Plaintiffs relied on the array of alleged misrepresentations would require an individualized inquiry. But the court found that class certification was appropriate as to the Plaintiffs' other theories that depend on common proof of a pyramid scheme. It held that first-party reliance is not an element of a RICO claim predicated on mail or wire fraud, and common proof could establish the proximate cause that is required. Although it focused primarily on the argument that a jury could logically infer that class members joined Ignite based on the implicit representation that it is a legal multi-level marketing program, it also recognized a more direct theory for proving proximate causation on a classwide basis: under the discussion of RICO causation in Bridge v. Phoenix Bond & Indem. Co.,2 it is enough to show that a "foreseeable and natural consequence" of the allegedly unlawful pyramid scheme is "that the vast majority of the unwitting IAs would lose money."3

The Defendants then filed a petition for interlocutory review with this court under Federal Rule of Civil Procedure 23(f), and a motion to stay proceedings pending resolution of that petition. The district court declined to stay the proceedings, at which time the Defendants filed a motion to stay with this court. This court granted a stay and granted the petition for review in March 2014. The panel majority agreed with the Defendants that individual issues of causation will predominate at trial and reversed the district court's class certification. We then granted the Plaintiffs' petition for rehearing en banc.

II.

The narrow issue in this case is whether the Plaintiffs may prove RICO causation through common proof such that individualized issues will not predominate at trial. The import of this inquiry is whether class certification is appropriate under Federal Rule of Civil Procedure 23(b)(3). We emphasize at the outset, and the Defendants conceded at the district court,4 that whether Ignite's multi-level marketing program is a fraudulent pyramid scheme is a merits issue subject to common proof. The Defendants might well prove that Ignite is a legal multi-level marketing program. That question, however, is left to be resolved in the first instance at the district court.

A.

We review a district court's certification of a class for abuse of discretion, but if the court's error is a matter of law, the court necessarily abuses its discretion.5 Our review is deferential "in recognition of the essentially factual basis of the certification inquiry and of the district court's inherent power to manage and control pending litigation."6

To obtain class certification, the party seeking it must initially comply with Federal Rule of Civil Procedure 23. That party must first satisfy Rule 23(a)'s requirements of numerosity, commonality, typicality, and adequacy of representation.7 If successful, that party must next satisfy the provisions of one of Rule 23(b)'s three subsections.8 Here, the Plaintiffs rely on subsection (3), "which requires that questions of law or fact common to the class predominate over questions affecting only individual members, and that a class action is superior to other available methods for the fair and efficient adjudication of the controversy."9 "The Rule 23(b)(3) predominance inquiry tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation."10 The Plaintiffs have the burden of showing that these requirements are met.11

The Defendants do not dispute the district court's Rule 23(a) determination and contend only that it erred in finding Rule 23(b)(3)'s predominance requirement met. "Considering whether `questions of law or fact common to class members predominate' begins, of course, with the elements of the underlying cause of action."12

B.

RICO makes it unlawful to conduct or participate in an enterprise's affairs "through a pattern of racketeering."13 To bring a RICO claim, a plaintiff must prove: "(1) the identification of a person, who, (2) through a pattern of racketeering activity, (3) uses or invests income derived therefrom to acquire an interest in or to operate an enterprise engaged in interstate commerce, or acquires, maintains an interest in, or controls such an enterprise."14 The second element, the pattern element, requires "at least two predicate acts of racketeering activity."15 Here, the putative class members advance two patterns of racketeering activity: (1) mail fraud in violation of 18 U.S.C. § 1341 and (2) wire fraud in violation of 18 U.S.C. § 1343.

RICO affords a private right of action only to a plaintiff who can show that he or she has been injured "by reason of" a violation of RICO's criminal prohibitions.16 The Supreme Court requires plaintiffs to establish both but-for cause and "proximate cause in order to show injury `by reason of' a RICO violation."17 Proximate cause "should be evaluated in light of its common-law foundations [and] ... requires `some direct relation between the injury asserted and the injurious conduct alleged.'"18 "When a court evaluates a RICO claim for proximate cause, the central question it must ask is whether the alleged violation led directly to the plaintiff's injuries."19

The Defendants' challenge to predominance rests on their belief that this causation element will require individualized proof. But that premise, and thus much of their opposition to class certification, is at odds with recent decisions from the Supreme Court and this court emphasizing that RICO claims predicated on mail and wire fraud do not require first-party reliance to establish that the injuries were proximately caused by the fraud.20

As the Supreme Court put it in Bridge v. Phoenix Bond & Indemnity Co.: "[A] person can be injured `by reason of' a pattern of mail fraud even if he has not relied on any misrepresentations."21 The Court explained that "[p]roof that the plaintiff relied on the defendant's misrepresentations may in some cases be sufficient to establish proximate cause, but there is no sound reason to conclude that such proof is always necessary."22 It further recognized that "the absence of first-party reliance may in some cases tend to show that an injury was not sufficiently direct to satisfy § 1964(c)'s proximate-cause requirement, but it is not in and of itself dispositive."23 At bottom, "the fact that proof of reliance is often used to prove an element of the plaintiff's cause of action, such as the element of causation, does not transform reliance into an element of the cause of action."24 Indeed, "[u]sing the mail to execute or attempt to execute a scheme to defraud is indictable as mail fraud, and hence a predicate act of racketeering under RICO, even if no one relied on any misrepresentation."25

We applied Bridge in St. Germain v. Howard, explaining that "no reliance requirement exists for civil causes of action under RICO for victims of mail fraud."26 We relied on the same principle in Allstate Ins. Co. v. Plambeck, noting again that "[i]n cases predicated on mail or wire fraud, reliance is not necessary."27 That case involved a group of telemarketing companies, chiropractic clinics, and law offices that convinced not-at-fault car accident victims to obtain chiropractic services so as to receive settlement payments from insurance companies. Allstate alleged that this group of defendants was liable under RICO's civil fraud statute for racketeering activity involving mail and wire fraud. After a trial, the jury returned a verdict in Allstate's favor. As to RICO causation, the district court instructed the jury that "proximate cause was present if `the injury or damage was either a direct result or a reasonably probable consequence of the act.'"28 The defendants appealed, challenging the jury's causation determination based on the absence of evidence that Allstate relied on the misrepresentations. We affirmed the verdict, holding that Allstate proved proximate cause because it was a foreseeable victim, and not one "wronged by the caprice of chance": "The objective of the enterprise was to collect from the insurance companies; the entire structure of the system ... shows that Allstate's paying up was not just incidental but was the object of the collaboration."29

Other circuits have adopted similar definitions of proximate causation under RICO. For example, the Sixth Circuit considers whether a direct relationship between the injury and alleged conduct exists, whether the plaintiff's injury is a foreseeable consequence of the alleged conduct, and whether the casual connection between the injury and alleged conduct is logical and not speculative.30 The Seventh Circuit looks simply to the "probability of a harm attributable to the defendant's wrongful act."31 The First Circuit, relying on the Supreme Court's discussion in Holmes v. Securities Investor Protection Corp.,32 looks to the directness between the injury and alleged conduct with reference to "three functional factors": (1) concerns about proving damages from attenuated injuries, (2) preventing multiple recoveries, and (3) whether societal interest in deterring the alleged conduct is served by the case.33 The Fourth Circuit has also held in an unpublished decision that "Bridge's holding eliminates the requirement that a plaintiff prove reliance in order to prove a violation of RICO predicated on mail fraud" in all contexts, not just third-party reliance cases.34

As will be shown below, this understanding of the causation requirement for fraud-based RICO claims — that such claims, unlike most common law fraud claims, do not require proof of first-party reliance — largely dooms the Defendants' attempt to identify individual issues of causation sufficient to preclude a finding of predominance.

C.

Under Bridge, the most straightforward way of demonstrating reliance in a classwide manner is the Plaintiffs' foreseeability argument.35 This just requires showing that the Plaintiffs' losses were caused "by reason of" the Defendants' operation of a fraudulent scheme.

That showing could flow directly from a jury's finding that the Defendants are operating a pyramid scheme as opposed to a lawful multi-level marketing program. Pyramid schemes are "inherently fraudulent" and are per se mail fraud, a RICO predicate act.36 And, by design, a pyramid scheme's fraud inheres in its concealment of the deceptive nature of the "robbing Peter to pay Paul" payment structure.37 In fact, the Defendants' CEO characterized this payment structure in an internal document as a "pyramid" in which "[t]here are Peters here to rob for the purpose of paying Paul." SRE.26. The Federal Trade Commission has recognized that a pyramid scheme harms its participants "by virtue of the very nature of the plan as opposed to any dishonest machinations of its perpetrators."38 Likewise, the Ninth Circuit recognizes that "[o]peration of a pyramid scheme constitutes fraud for purposes of ... various RICO predicate acts."39

The Federal Trade Commission instructs that a pyramid scheme is characterized by payments by participants in exchange for "(1) the right to sell a product and (2) the right to receive in return for recruiting other participants into the program rewards which are unrelated to the sale of the product to ultimate users."40 The fraud lies in the concealment of the inevitable collapse that results from the scheme's structure because "[t]he promise of lucrative rewards for recruiting others tends to induce participants to focus on the recruitment side of the business at the expense of their retail marketing efforts, making it unlikely that meaningful opportunities for retail sales will occur."41 That structure, which focuses on recruitment of people, not products, inevitably causes the scheme to collapse when participants run out of individuals to recruit and there are no more new recruits to pay those higher up the pyramid. But "[n]o clear line separates illegal pyramid schemes from legitimate multilevel marketing programs."42 Indeed, "the very reason for [their] per se illegality ... is their inherent deceptiveness and the fact that the futility of the plan is not apparent to the consumer participant."43

Because pyramid schemes are per se mail fraud, which include inherent concealment about the deceptive payment scheme, one who participates in a pyramid scheme can be harmed "by reason of" the fraud regardless of whether he or she relied on a misrepresentation about the scheme. "An inherently fraudulent pyramid scheme ... would fall within the[] broad definitions of fraud" under RICO even if no misrepresentations occur.44 Participants are then harmed by the fraud involved in pyramid schemes not because of any misrepresentations, but because the ultimate collapse of the scheme, and thus harm to participants, is a direct and foreseeable consequence of such structure.

Here, the Plaintiffs allege that the Defendants operated a fraudulent pyramid scheme, which has caused them financial losses. There can be no question that the Plaintiffs are both the direct and foreseeable victims of the alleged fraud. By definition, a pyramid scheme operates by taking money from downline recruits, like the Plaintiffs, who will never recoup their payments, and funneling the money to those at the top of the pyramid. Such schemes depend on "there [being] Peters ... to rob for the purpose of paying Paul." Those who lose money in a pyramid scheme necessarily do so "by reason of" the fraud because the fraud is necessary to temporarily sustain the scheme, and ultimately causes the scheme's collapse. And, those who profit from a fraudulent pyramid scheme make money only by virtue of the participation of downline investors, like the Plaintiffs, who lose money.

The Plaintiffs are necessary to the scheme and are the direct victims of the scheme. Equally clear is that the Plaintiffs are the foreseeable victims of the alleged fraud: "Pyramid schemes are destined to collapse, and the most recent entrants to lose their money."45

Whether the Plaintiffs relied on a misrepresentation about the scheme is thus not determinative of whether the Plaintiffs can prove proximate causation under Bridge. As was true in that case, the class members here can prove injury "`by reason of' a pattern of mail fraud even if [they have] not relied on any misrepresentations."46 The participants' injuries arise from the scheme's payment structure, and the inherent concealment of the inevitableness of those injuries.

Further, although a class member's knowledge that Ignite is an illegal pyramid scheme could serve as an intervening cause that would break the chain of causation,47 the Defendants, as will be discussed more below, have offered no evidence that any putative class member knew Ignite was an illegal pyramid scheme before joining as an IA. The district court expressly found that the record contained no such evidence, and we find no error in that determination.

Moreover, the directness of the Plaintiffs' alleged injuries obviates any concerns that might exist in cases with attenuated injuries. As in Bridge, "there are no independent factors that account for [the Plaintiffs'] injury, there is no risk of duplicative recoveries by plaintiffs removed at different levels of injury from the violation, and no more immediate victim is better situated to sue."48

The Plaintiffs' claims under this foreseeability theory of proving causation will rise or fall on common evidence. The facts necessary to prove that the Defendants operated a fraudulent pyramid scheme will also suffice to show under Bridge that the fraud caused the Plaintiffs' injuries. Accordingly, under this theory of causation, individualized issues of causation will not predominate.

D.

We will also address the inference-based theory of causation that was the focus of the panel opinions. We find that this is a separate basis on which to affirm the certification ruling.

Under this theory, the Plaintiffs argue that Ignite's holding itself out as a legitimate multi-level marketing program, when in fact it was a fraudulent pyramid scheme, gives rise to a reasonable inference that that misrepresentation induced their paying to join as IAs and caused their losses. This, the Plaintiffs assert, is because (1) it may be rationally assumed that a precondition for joining Ignite was that it was a legal business opportunity, and (2) the Defendants have offered no evidence of any putative class member who joined or would have joined knowing Ignite was a fraudulent pyramid scheme, in which the majority of participants are bound to lose money.

We note initially that the Defendants do not challenge whether Ignite represented itself to be a legal multi-level marketing program or whether this question is common to the class. They do not do so for good reason: by operating its program, Ignite has and continues to hold itself out as a legal multi-level marketing program. The Federal Trade Commission's persuasive precedent recognizes that pyramid schemes make "the inevitably deceptive representation (conveyed by their mere existence) that any individual can recoup his or her investment by means of inducing others to invest."49 Pyramid schemes are inherently deceptive because their very structure conceals the fact that those at the bottom of the pyramid will be unable to recoup their investment. Accordingly, we conclude that the misrepresentation at issue here — that Ignite is a legal multi-level marketing program — is subject to common proof and is not even disputed.

We turn next to the question whether the Plaintiffs may employ a common inference of reliance based on that alleged misrepresentation. The Defendants concede that a common inference of reliance is appropriate in some cases. They urge us to adopt a rule requiring that, to invoke an inference of reliance in a fraud case, the Plaintiffs must establish that no rational actor would have participated had they known of the misrepresentation. Other circuits, however, have not applied such a narrow rule. Instead, they have permitted inferences of reliance when it follows logically from the nature of the scheme, and there is common, circumstantial evidence that class members relied on the fraud.

In Klay v. Humana, Inc.,50 the Eleventh Circuit upheld the certification of a class of physicians claiming that health maintenance organizations (HMOs) misrepresented that they would pay them for medically necessary services, but instead underpaid them.51 The Eleventh Circuit affirmed the class certification based on a common inference of reliance on those misrepresentations, explaining that "[a] jury could quite reasonably infer that guarantees concerning physician pay — the very consideration upon which those agreements are based — go to the heart of these agreements, and that doctors based their assent upon them."52 Similarly, in In re U.S. Foodservice Inc. Pricing Litigation,53 the Second Circuit held that customers who were allegedly overbilled by a food distributor's inflated invoices scheme could be certified as a class.54 It reasoned that "customers who pay the amount specified in an inflated invoice would not have done so absent reliance upon the invoice's implicit representation that the invoiced amount was honestly owed."55 Conspicuously absent from both the Eleventh and Second Circuits' decisions was any requirement that the plaintiffs prove that no other rational explanation existed for their behavior other than reliance.56

Given the unfavorable holdings of the courts' decisions in Klay and U.S. Foodservice, it is unsurprising that the Defendants relegated these opinions to a footnote in their en banc briefing. Instead, they urge this court to rely on the Tenth Circuit's recent opinion in CGC Holding Co. v. Broad & Cassel.57 That court approved a common inference of reliance to certify a class when a class of borrowers alleged that a group of lenders fraudulently extracted nonrefundable loan commitment fees from the borrowers for loans that the lenders never intended to provide.58 It explained that:

Although the Tenth Circuit approved the theory of inferred reliance after concluding that no rational actor would join the scheme had he or she known of the fraud, we do not read its opinion as limiting an inference of reliance to that situation. That court's opinion says only that the absence of another rational explanation for the plaintiffs' behavior is sufficient to infer reliance — it does not say it is a necessary condition. And tellingly, the Tenth Circuit cited the district court's opinion in this case approvingly.60

Turning to the facts of this case, we conclude that if the Plaintiffs prove that Ignite is a fraudulent pyramid scheme, they may use a common inference of reliance to prove proximate causation under RICO. A jury may reasonably infer that, in deciding to pay to become IAs, the Plaintiffs relied on Ignite's implicit representation that it is a legal multi-level marketing program, when it is in fact a fraudulent pyramid scheme. Two points support this conclusion.

First, it is reasonable to infer that individuals do not knowingly join pyramid schemes because (1) pyramid schemes are inherently deceptive and operate only by concealing their fraudulent nature; and (2) knowingly joining a pyramid scheme requires the individual to choose to become either a victim or a fraudster. Both points support a reasonable inference that the class members would not have knowingly joined a fraudulent pyramid scheme.

Whether a multi-level marketing program is fraudulent or legitimate depends on its internal structure. And such information is not readily apparent or interpreted. "[T]he very reason for the per se illegality of [such] schemes is their inherent deceptiveness and the fact that the `futility' of the plan is not `apparent to the consumer participant.'"61 If a scheme's illegality were apparent, the scheme would not work. After all, the whole point of a pyramid scheme is to dupe unwitting investors into joining. The sheer improbability that more than a handful of class members (and even a handful seems unlikely) would be able to recognize that Ignite was a fraudulent pyramid scheme before joining as IAs supports the reasonableness of the Plaintiffs' inference of reliance.62

Second, the record is devoid of evidence that a single putative class member joined as an IA despite having knowledge of the fraud. Even after the close of discovery and the commencement of summary judgment motions before the district court, the Defendants produced no evidence that a single class member even knew of the fraud or would have paid to become an IA knowing of the fraud. Faced with this vacuum of evidence, the district court correctly concluded that individual issues of reliance will not predominate at trial.

The Defendants protest, however, that our pointing to the absence of evidence supporting their defense somehow improperly shifts the burden of proof to them. Not so. The Defendants, while advocating a narrower rule, have now conceded in their en banc brief that the absence of contrary evidence would support class certification based on an inference of reliance: "To be sure, in cases where a plaintiff has demonstrated that nobody would want the opportunity the defendant is offering, then class certification could be appropriate — absent contrary evidence." The district court was tasked with determining how a trial would proceed. That court did not simply presume that individual issues of reliance would not predominate; rather, it specifically made this conclusion based on its determination that the Plaintiffs' case could be made with common evidence. And, in the absence of any evidence showing that individuals joined the pyramid scheme knowingly — the district court correctly ruled that individual issues of reliance will not predominate.63

Neither now nor before the district court have the Defendants even attempted to bear this burden of rebutting the Plaintiffs' evidence of reliance.64 On appeal, they do not even contest the district court's factual finding, which we review only deferentially for an abuse of discretion. Had the Defendants presented evidence that could rebut the Plaintiffs' common inference of reliance on an individualized basis, we and the district court might have concluded that individual issues of reliance would predominate at trial. In the total absence of such evidence, however, we have no evidentiary basis to conclude that the district court abused its discretion in holding otherwise.

Rather than pointing to evidence, the Defendants rely on speculation alone that a hypothetical class member could have joined as an IA despite knowing of the fraud. But such sheer speculation as to the improbable motivations of an undefined, but likely minute number of class members does not cause individual issues of reliance to predominate. Our inquiry looks to how the trial will proceed;65 trials are grounded in evidence, not extra-record attorney speculation. As our sister circuit recognized, "if bald speculation that some class members might have knowledge of a misrepresentation were enough to forestall certification, then no fraud allegations of this sort (no matter how uniform the misrepresentation, purposeful the concealment, or evident plaintiffs' common reliance) could proceed on a class basis."66 And mere conjecture that some class members may have acted with knowledge of the misrepresentation seems particularly inappropriate here as anyone who joins a pyramid scheme hoping to become one of the few winners sitting at the top of the pyramid would become liable as a knowing participant.

For these reasons, our result in the instant case is not inconsistent with Sandwich Chef of Texas, Inc. v. Reliance National Indemnity Insurance.67 There, insureds alleged that insurers charged premiums in excess of approved rates, then misrepresented the correctness of the premiums charged.68 We rejected class certification because the insureds could not prove proximate causation through common proof. Unlike the Defendants in the instant case, the insurers in Sandwich Chef not only contended that the insureds "were aware that [the insurance] carriers were charging them more than the filed rates," but also "introduced evidence that ... class members individually negotiated with insurers regarding workers' compensation and insurance premiums."69 Thus, "[k]nowledge that invoices charged unlawful rates, ... according to a prior agreement between the insurer and the policyholder, would eliminate reliance and break the chain of causation."70 Here, the Defendants have put forth no such evidence.71

None of this is to say that if the Plaintiffs prove that Ignite is a fraudulent pyramid scheme, they must necessarily prevail at trial if this inference-theory is advanced. The inference of reliance to which the Plaintiffs are contingently entitled is simply the common mechanism by which they seek to prove their affirmative case. The jury may or may not make this inference in the Plaintiffs' favor: "[T]he trier of fact is not required to accept the inference; it is merely permitted to utilize it as common evidence to establish the class's prima facie claims under RICO."72 And the district court may revisit its decision and choose to decertify the class should the Defendants eventually produce individualized rebuttal evidence causing their individualized defense to predominate.

But the focus must remain on the predominance inquiry. We thus recognize that even if conjecture alone is sufficient to establish that a few class members might have knowingly joined a fraudulent pyramid scheme, this will not necessarily cause individualized issues of reliance to predominate at trial. In the context of the fraud-on-the-market theory, the Supreme Court's recent pronouncement in Halliburton Co. v. Erica P. John Fund, Inc. is highly instructive:

This reasoning applies with equal weight here.74 Evidence indicating that a few class members decided to take the risk of being a winner in an illegal pyramid scheme does not automatically rebut the inference of reliance for the overwhelming remainder of class members or mean that individual issues concerning the atypical knowing fraudsters will predominate at trial. This is underscored by the fact that the instant class is comprised of only those who lost money participating in Ignite's program.

In sum, we conclude that if the Plaintiffs prove that the Defendants operated a fraudulent pyramid scheme, a jury may reasonably infer from the Plaintiffs' payments to join as IAs that they relied on Ignite's implicit representation of legitimacy, when in fact it was a fraudulent pyramid scheme. Although it is not impossible that some class members might have joined as IAs despite knowledge of the fraud, economic speculation alone as to what could have motivated an individual class member is not enough to defeat class certification. Based on the deception inherent in pyramid schemes and the losing proposition that they present to the vast majority of participants, it is highly unlikely that many — if any — of such class members exist. And more importantly, the district court expressly found no evidence indicating that any putative class member knew of the fraud. Because the Defendants failed to demonstrate that such individualized issues will affect even a single class member at trial, we find no error in the district court's conclusion that individualized issues of causation will not predominate. Accordingly, we affirm the district court's class certification.

III.

The class certification of the district court is AFFIRMED.

E. GRADY JOLLY, Circuit Judge, joined by Edith H. Jones and Edith Brown Clement, Circuit Judges, and joined, as to Parts IB and II, by Priscilla R. Owen, Circuit Judge, dissenting:

The majority concludes that the plaintiffs do not need to make any showing of reliance to establish proximate cause under RICO. Citing the Supreme Court's decision in Bridge v. Phoenix Bond & Indemnity Co., 553 U.S. 639, 128 S.Ct. 2131, 170 L.Ed.2d 1012 (2008), and this circuit's recent decision in Allstate Insurance Co. v. Plambeck, 802 F.3d 665 (5th Cir. 2015), the majority opinion holds that the plaintiffs have satisfied Rule 23's predominance requirement for RICO proximate cause simply because the plaintiffs have made a sufficient showing that Ignite is an illegal pyramid scheme, and that they lost money by investing. The majority thus asserts that the plaintiffs do not need to show that the defendants made any false representation upon which the plaintiffs relied to make their losing investment.

I.

A.

First, the majority errs in its cavalier disregard of evidence of individualized knowledge among the class members. The majority concludes that the plaintiffs have met Rule 23's predominance inquiry with respect to causation under RICO simply because there is evidence suggesting Ignite was a pyramid scheme. In reaching this holding, the majority opinion ignores that, from the outset of their involvement with Ignite, the plaintiffs were provided all the information needed to warn investors of Ignite's likely illegality.

Again, there is no quarrel here with the majority opinion's simple assertion that reliance is not a prerequisite for proving proximate cause under RICO.1 The facts of this case, however, do not allow for such a glossy approach to class certification. The majority's reasoning has force only to the extent that the plaintiff-investors were actually unaware of Ignite's fraudulent structuring. See Bridge, 553 U.S. at 658-59, 128 S.Ct. 2131 (stating that, although first-party reliance is not a formal element of a RICO claim, proximate cause fails where there is evidence that the aggrieved party or an intermediary knew of the fraud, because such knowledge acts as an "intervening cause breaking the chain of causation between petitioners' misrepresentations and respondents' injury"); see also Sandwich Chef of Tex., Inc. v. Reliance Nat'l Indem. Ins. Co., 319 F.3d 205, 218-19 (5th Cir. 2003) (recognizing that knowledge, which is actually a defense to causation, is a relevant consideration when addressing class certification). Moreover, as both parties concede, for the purposes of proximate cause, it does not matter whether Ignite actually is a pyramid scheme. Instead, the relevant inquiry is whether there are class members who understood Ignite was likely to be a pyramid scheme, but invested anyway. If so, the line of causation becomes too tenuous to maintain through common evidence solely based on the contention of Ignite's alleged illegality.

The majority opinion takes for granted that no individualized issues of knowledge exist among the plaintiff class, asserting that "the record is devoid of evidence that a single putative class member joined [Ignite] despite having knowledge of the fraud." It adopts this position notwithstanding that the plaintiffs, by their own admission, were provided the information that Ignite was likely an illegal pyramid scheme. The record shows that the tell-tale signs of an illegal pyramid scheme were disclosed to the plaintiffs in the documents they were provided before signing up for Ignite. Ignite's business plan, published to potential investors, openly preached recruiting additional IAs over selling Ignite's purported product, residential energy.2 Similarly, Ignite's published compensation scheme, which the plaintiffs do not dispute is accurate and was provided to all investors, also bears all the hallmarks of an illegal pyramid scheme. For example, Ignite paid only fifty cents in commission to new IAs per each energy customer they enrolled. In contrast, those IAs that were higher up in the pyramid structure received the bulk of profit resulting from the sale of residential energy. This mark, of course, is a defining trait of a pyramid scheme, but it is also a trait that the plaintiffs themselves assert was made obvious to Ignite's investors from the outset.

In their en banc briefing, the plaintiffs themselves repeatedly urge that anyone could see that the only realistic way to make money as an Ignite IA was to recruit new IAs to work underneath you, and to teach those new IAs to do the same. The plaintiffs emphasize that common sense compels the conclusion that Ignite's business model was illegal from the outset, since the unsustainability of such a scheme is apparent on its face; eventually, there are no more new IAs to recruit. According to the plaintiffs, "[a]ny `energy company' sales program that is `not about becoming an energy salesperson' necessarily collapses; if everyone tries to succeed by `duplicating' a huge class is inevitably left with a loss when the recruits run out." Appellees' Supplemental En Banc Brief at 7. Taking the plaintiffs at their own emphatic word, it follows that the class members who took minimal time to read the investment materials would have developed serious concerns about Ignite's risk and illegality. Still, they invested. The plaintiffs, however, contend, in contradictory fashion, that these overt "buyer beware" warnings were insufficient to put even a single plaintiff on notice that Ignite was actually an illegally structured venture. At the very least, these warnings were sufficient to cause the prudent investor to question Ignite's business structure before blindly investing.

It is true that our caselaw, of course, does not require an investor to comb through the finest details of a defendant's business plan to preserve a later claim for fraud. But it does, however, require that a plaintiff-investor do some minimum amount of research into the nature of an investment opportunity before signing up, losing money, and crying fraud. See Martinez Tapia v. Chase Manhattan Bank, N.A., 149 F.3d 404, 409 (5th Cir. 1998) ("The investor who seeks to blame his investment loss on fraud or misrepresentation must himself exercise due diligence to learn the nature of his investment and associated risks.... [T]he party claiming fraud and/or misrepresentation must exercise due diligence to discover the alleged fraud and cannot close his eyes and simply wait for facts supporting such a claim to come to his attention.").

In addition to the investment documents, a cursory Google search would have led the plaintiffs to a Dallas Morning News article, published during the time frame relevant to class certification, in which an economic expert expressly stated that Ignite was an illegal pyramid scheme, destined to result in a loss of money for most of its investors. Indeed, the plaintiffs themselves refer to this article in their complaint, but still contend that there is no sound basis to conclude that at least part of the class members were aware that Ignite was thought to be an illegal venture, but chose to "take their chances" and sign up anyway.

Standing on its own, the evidence above is enough to undermine the notion that all 200,000-plus members of the putative class were unaware that Ignite had all the indicia of an illegal pyramid scheme. But this is not the extent of the evidence suggesting knowledge of the defendants' fraud, which the plaintiffs now allege was a surprise. In fact, there is significant evidence that Ignite's own promoters, when talking to potential investors, were explicit about the company's dubious structuring. The defendants routinely held large, revival-style recruitment events, where Ignite executives and promoters explained Ignite's business model. Although each recruiter's style differed, there was a common theme in their presentations: Ignite offered potential IAs a great opportunity to make money, albeit through recruiting other IAs instead of through actual sales. Indeed, one promoter, Randy "the Cowboy" Hedge, told a crowd of potential investors that, to scare off the faint of heart, he would sometimes refer to Ignite as a "pyramid" deal. Hedge suggested that he did this because he knew that those people who remained interested in joining Ignite, even after hearing the alarm-sounding descriptive "pyramid" applied to its business model, were chiefly concerned about making money, and not about the details of Ignite's structuring.3

The majority opinion dismisses this evidence of individualized knowledge by deeming it too speculative. Citing a four-decade-old order from the Federal Trade Commission, published when pyramid schemes were still a relatively new form of potential fraud, the majority urges that pyramid schemes are "inherently deceptive," to the extent that unsophisticated consumer-investors could not possibly discern whether Ignite's business model was illegal before joining up. What is implied by this statement is that a multilevel marketing scheme that, at first glance, bears the indicia of legality may, upon deeper investigation, reveal subtleties of its structuring that actually make it an illegal pyramid scheme. Such subtleties, however, are entirely absent from this case. Indeed, as discussed above, all the evidence necessary to conclude Ignite was a pyramid scheme was provided to the class members and they still chose to invest; moreover, at least a number of the plaintiffs were exposed to recruitment pitches that emphasized Ignite's pyramid character. This evidence, even if thought not to be conclusive on whether most plaintiffs knew of the likelihood that Ignite was an illegal pyramid scheme, is far more than "speculative." At the very least, the defendants are entitled to probe these plaintiffs' understanding of the Ignite investor documents and accompanying sales pitches, in an effort to challenge the plaintiffs' supposed naiveté of Ignite's unsustainability. Accordingly, these lingering problems of individualized knowledge among many of the class preclude a finding that, consistent with the meaning and requirements of Rule 23, common issues predominate with respect to proximate cause under RICO. See Sandwich Chef, 319 F.3d at 220.

B.

Next, given that the evidence discussed above raises concerns of individualized knowledge, the majority errs in placing the burden regarding the appropriateness of class certification with the defendants, instead of the plaintiffs. The majority opinion asserts that, even assuming there is record evidence showing an indeterminate number of plaintiffs knew of Ignite's illegality, the record evidence fails to show that individualized issues of knowledge will actually undermine those issues common to the class. The majority opinion points out that knowledge is an affirmative defense, which the defendants must raise and prove at trial. According to the majority, the fact that a "few" plaintiffs might be "picked off" because of issues of individualized knowledge does not defeat class certification, so long as issues common to the class continue to predominate over the "outliers."

There is no questioning that, as a general proposition, a class may be certified even when a few stray issues of individualized knowledge remain among the class's members. It is certainly correct that Rule 23 requires a predominance of common issues, not a uniformity of them. More relevant here, however, is that Rule 23 also requires that the plaintiffs, not the defendants, carry the burden of establishing whether class certification is appropriate under Rule 23; the plaintiffs must do so by showing that individualized inquires will not cast a shadow over those issues common to the entire class. See Wal-Mart Stores v. Dukes, 564 U.S. 338, 131 S.Ct. 2541, 2551, 180 L.Ed.2d 374 (2011). This burden includes showing that a defendant's proffered affirmative defense, if based on individualized issues of knowledge, applies only to an insignificant segment of the putative class. See Gene & Gene LLC v. BioPay LLC, 541 F.3d 318, 329 (5th Cir. 2008). ("An affirmative defense is not per se irrelevant to the predominance inquiry, as the parties seem to believe. We have noted that the predominance of individual issues necessary to decide an affirmative defense may preclude class certification." (internal citation and quotation marks omitted)); Sandwich Chef, 319 F.3d at 220 (stating that Rule 23 requires that the district court's predominance inquiry account for any individual issues of knowledge that will be "components of defendants' defense against RICO fraud.").

The plaintiffs, however, disregard this burden under Rule 23. Importantly, the plaintiffs do not even attempt to show that the defendants' proffered defense of individualized knowledge applies only to an insignificant number of plaintiffs. Instead, they argue that a lack of knowledge may be presumed, because no "rational" individual would ever participate in an illegal pyramid scheme. Again, this theory — which, at different points in this case's history, the plaintiffs have referred to as the "fraud-on-the-market" theory, the "rational economic actor" theory, and the "inferred reliance" theory — is the only basis upon which the district court granted class certification.4 It follows that, if such a theory were accepted in error, individualized issues of knowledge overwhelm those issues common to the class, rendering this class action unfit for certification. The plaintiffs, however, did not even confront the evidence suggesting individualized knowledge; instead, as stated, they chose to seek an inference of reliance — and hence, an inference that all of the plaintiffs lacked knowledge of Ignite's illegality — based solely on an "implicit" misrepresentation, made by virtue of Ignite's mere existence.

As discussed below, the plaintiffs' theory of "inferred reliance" is both logically strained and is belied by the absence of any actual misrepresentation on behalf of the defendants. Ultimately, however, it does not matter whether reliance is required to establish RICO proximate cause; even if no showing of reliance is necessary, superseding issues of individualized knowledge cloud the waters of RICO causation. Accordingly, the plaintiffs have failed to meet their burden, under Rule 23(b)(3), of showing that common issues predominate with respect to RICO's proximate cause element.

II.

Let us now turn to the majority's alternative holding regarding the appropriateness of an inference of reliance in this case. The majority opinion asserts that, assuming reliance on a misrepresentation must be shown to establish RICO causation, the plaintiffs have done so through common evidence. The plaintiffs, however, do not point to any common, specific misrepresentation upon which they relied, much less offer evidence demonstrating reliance. Instead, they seek an "inference" of reliance on an "implicit" misrepresentation. The plaintiffs contend that, simply by seeking to recruit new customers and investors, Ignite falsely held itself out as a legitimate business opportunity. They further assert that, because the legality of Ignite's business structure would have been a bedrock assumption of any reasonable investor, the court may infer that the plaintiffs relied on this implicit misrepresentation when choosing to join Ignite.5 In arguing that an inference is appropriate here, the plaintiffs point to a handful of circuit-level cases that have allowed a class-wide inference of reliance under certain circumstances: CGC Holding Company, LLC v. Broad & Cassel, 773 F.3d 1076 (10th Cir. 2014), In re U.S. Foodservice Inc. Pricing Litigation, 729 F.3d 108 (2d Cir. 2013), and Klay v. Humana, Inc., 382 F.3d 1241 (11th Cir. 2004).

These cases are distinguishable. Both Foodservice and Klay allowed a jury to "infer" reliance when the false representations at issue were straightforward misstatements of an amount owed or paid on a bill or invoice. Those courts concluded that a jury could infer reliance because an individual's payment of a bill or acceptance of a payment was, in effect, an acknowledgment of reliance on the correctness of the amount in the bill or payment. This reliance makes sense, as no rational economic actor would knowingly pay extra for nothing. CGC Holding also involved a scenario where the plaintiffs were purchasing a worthless product; they were applying for loans and paying a nonrefundable fee to the defendants, even though the defendants had already decided that they would eventually deny the plaintiffs' loan applications. Indeed, every single case, cited by the plaintiffs or the district court, where an "inference of reliance" was used to establish RICO causation involves allegations of a palpable, specific misrepresentation.6

The plaintiffs here, however, seek a wholly novel application of the inferred reliance theory. They urge the court to conclude that, as a matter of law: No rational person would ever knowingly invest in a business venture that could be illegal. Such an implausible argument ignores that, even if Ignite was a pyramid scheme, it allowed IAs the chance to make money. By the plaintiffs' own admission, roughly 10-15% of investors made a profit over the time frame relevant to this litigation. Unlike the "something-for-nothing" transactions that served as the basis for an inference of reliance in the other circuit-level decisions, a person could rationally invest in a pyramid scheme with the hope that he or she might profit significantly, notwithstanding knowledge that a majority of participants will likely be losers. As for the majority's altruistic suggestion that an inference of reliance is appropriate because no rational individual would ever knowingly chance defrauding others in an effort to make money for herself, I respectfully suggest that our criminal docket demonstrates the error of this assumption.

There is no attempt here to defend the legality of the defendants' alleged pyramid scheme. The point is that the plaintiffs cannot maintain this class action as it has been structured and presented to the court. The plaintiffs do not allege, much less offer any common evidence, that the defendants misrepresented any aspect of its business structure; nor do they allege that the defendants misrepresented the plaintiffs' likelihood of being able to sign up enough customers or downline recruits to make a profit. One is blind to reality to assume perfunctorily that approximately 200,000 IAs, pitching this scheme to each other and among themselves, were predominantly motivated only by an implicit, unspoken representation that Ignite was a "legal business opportunity." Given the lack of an actual misrepresentation, coupled with the fact that the plaintiffs had all the information necessary to know that Ignite was a risky pyramid scheme, the plaintiffs' theory of reliance is ill-adapted and out of place. Without this inference, the plaintiffs do not offer any common evidence with respect to proximate causation under RICO. Thus, the class should be decertified for failure to meet Rule 23's predominance requirement.

III.

To sum up: the majority opinion allows the plaintiffs to overcome Rule 23's predominance inquiry with respect to RICO causation, even though all of the plaintiffs were provided the information to understand the risk that Ignite was an illegally structured enterprise. Moreover, at least part of the class was warned of the risk of investing in Ignite by the defendants' own promotional representatives. It is impossible rationally to presume that, out of 200,000-plus investors, a significant number of the class were not aware of the precise character of their investment.

The majority opinion dilutes both RICO's causation requirement and Rule 23's predominance requirement to the point that they have little relevance in cases based on allegations of a pyramid scheme. Indeed, if the court finds class certification appropriate here — in a case with over 200,000 putative class members, all of whom learned about Ignite at different times and through different channels of communication, and undoubtedly held different levels of knowledge about the company's business plan — it is difficult to see when individualized issues among class members would preclude certification under Rule 23.

Accordingly, I respectfully dissent.

EDITH H. JONES, Circuit Judge, joined by EDITH BROWN CLEMENT, dissenting.

I am pleased to join Judge Jolly's dissent to the class certification approval in this case. The majority's rules, as Judge Jolly's dissent shows, afford far less scrutiny to class actions in cases involving mere allegations of "illegal pyramid schemes," and are legally ill-founded. I wish to make two observations, lest the reader of the majority opinion believe that Stream Energy is already condemned for operating an illegal pyramid scheme. Courts should not be in the business of writing one-sided opinions that lay a thumb on the scale simply by ignoring proof that does not comport with their conclusions. Thus, a few facts, as opposed to suppositions and allegations, cast doubt on the ease with which the majority condemns Stream's marketing method as illegal.

First, Stream Energy has existed in Texas for more than a decade and has become the fourth largest retail gas and electrical energy provider in this state. Stream is also authorized to sell energy in a half dozen additional jurisdictions. Stream serves over a million Texas customers, in part because it offers energy at competitive prices. Stream characterizes its marketing subsidiary Ignite's business as multilevel marketing, the bare bones of which are sketched in the majority and dissenting opinions. Whatever else may be the case, however, Stream sells a lot of real product to real people at favorable prices and its marketing model has yet to collapse.

Second, the majority never defines an "illegal pyramid scheme." The majority cites two elements described by the FTC over forty years ago: it is characterized by payments by participants in exchange for "(1) the right to sell a product and (2) the right to receive in return for recruiting other participants into the program rewards which are unrelated to the sale of the product to ultimate users." In re Koscot Interplanetary, Inc., 86 F.T.C. 1106 (1975). But the FTC has refused more rigorously to define an illegal pyramid scheme, and the majority opinion admits that "[n]o clear line separates illegal pyramid schemes from legitimate multilevel marketing programs." (citation omitted). Indeed, there are dozens of legitimate, longstanding multilevel marketing companies in the United States (e.g., Avon, Mary Kay Cosmetics, Amway, and Tupperware). The majority thus leaves it to the unfettered and untutored discretion of the district court and jury to decide whether Ignite is an "illegal pyramid scheme." I do not ever recall sending a case to a jury with so little definition of the elements of the offense, much less, for class action purposes, assuming guilt from the enterprise's mere structure, allowing an inference of class-wide reliance and requiring no proof of individual causation.

If this isn't stacking the deck legally, I don't know what is. But I surmise that even plaintiffs' counsel do not really believe Stream runs an "illegal pyramid marketing scheme." Had they truly believed this, they could have invoked the Department of Justice or FTC to assist in shutting Stream down. Instead, they claim to be suing to recover about $329 apiece for over 200,000 IAs who, they assert, lost money on their "investments" with Stream. This amount, nearly $60 million, would be trebled pursuant to RICO, exposing Stream to over $190 million in potential damages, plus contingent attorneys' fees. Since this is far more than Stream is worth, however, the plaintiffs' attorneys must either want to take over the business themselves or simply strong-arm a settlement, leaving the "illegal pyramid scheme" in place until it pays off.

This, I suggest, is the price of lowering the standards for liability and stripping businesses of the ability to know in advance what the law commands. Reckless allegations of undefined illegality, coupled with immense uncertainty as to outcomes, are an affront to the rule of law.

HAYNES, Circuit Judge, dissenting:

The majority opinion allows any group of plaintiffs who have lost money in a multi-level marketing program to automatically obtain class certification by making the simple allegation that the program was in actuality an illegal pyramid scheme. In so doing, it minimizes the fact that many plaintiffs would be unable to show that defendants caused their injuries, and it allows the plaintiffs to skirt their burden of establishing "that the questions of law or fact common to class members predominate over any questions affecting only individual members." FED. R. CIV. P. 23(b)(3).

The Supreme Court has emphasized that for plaintiffs to satisfy the causation requirement of a civil RICO claim, there must be "some direct relation between the injury asserted and the injurious conduct alleged." Bridge v. Phoenix Bond & Indem. Co., 553 U.S. 639, 654, 128 S.Ct. 2131, 170 L.Ed.2d 1012 (2008) (citation omitted). With over 200,000 plaintiffs in this case, there are numerous and disparate motivations behind each plaintiff's decision to participate in Ignite's multi-level marketing program, many of which weaken or sever any chain of causation.

For example, some of the plaintiffs could have been fully aware of the questions surrounding Ignite's legality, but nevertheless decided to participate for the simple reason of making a profit. For these plaintiffs, there would be no "direct relation" between the funds lost and Ignite's actions; the cause of any losses incurred would be based on the plaintiffs' own informed decision to take on a calculated risk that ultimately did not pay off. In other words, these plaintiffs' own assumption of risk "would constitute an intervening cause breaking the chain of causation between" Ignite's actions and these plaintiffs' injuries. Id. at 658, 128 S.Ct. 2131. By affirming the certification of a class that includes this subset of plaintiffs, the majority opinion provides a potential bailout for those who knowingly gambled and lost.

Other plaintiffs could have joined Ignite's program for the sole purpose of selling (or learning the business of selling) energy, which, as Judge Jones's dissenting opinion points out, is an aspect of the business that is indisputably legal. For these plaintiffs, Ignite's structure as a purported pyramid scheme could not have caused their injury, as any losses would be directly related only to an "independent[] factor[]." Id. at 654, 128 S.Ct. 2131 (citation omitted). Specifically, their losses would have been caused by their own inability to sell the energy necessary in order to turn a profit.

Other plaintiffs may have joined Ignite solely to take advantage of Ignite's training courses or networking opportunities, while others could have participated without any intention of making a profit in order to help out a friend or family member who was already a part of the program. For these plaintiffs, it would be impossible for Ignite to have caused any alleged injury, because no injury exists: these plaintiffs obtained exactly what they were hoping to receive by participating in Ignite's program. By affirming the certification of a class that includes these plaintiffs, the majority opinion allows those who have already received the benefit of their bargain with Ignite to potentially recoup the fees paid and effectively receive Ignite's products and services for free. In so doing, the majority opinion undermines one of the purposes of RICO causation, which the Supreme Court has stated is "to obviate the risk of multiple recoveries." Id. (citation omitted).

Plaintiffs could have participated in the program as "a form of escape, a casual endeavor, a hobby, a risk-taking money venture, or scores of other things." Poulos v. Caesars World, Inc., 379 F.3d 654, 668 (9th Cir. 2004) (concluding that class certification was inappropriate in a civil RICO case because the various motivations for gambling precluded common issues from predominating over individual ones). Each plaintiff had subjective and individualized reasons for joining Ignite's multi-level marketing program. As the parties seeking class certification, plaintiffs had the burden to show that — despite each plaintiff's differing motivations and expectations — common questions "predominate over any questions affecting only individual members." FED. R. CIV. P. 23(b)(3); see Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350-51, 131 S.Ct. 2541, 180 L.Ed.2d 374 (2011). This they failed to do. I respectfully dissent.