dissenting.
I respectfully dissent from that portion of the court’s opinion that affirms the district court’s judgment in favor of Tobacco on the HMOs’ damages claims. The court affirms based on its conclusion that the HMOs failed to make out a submissible case on the issue of damage calculation, despite the fact that the district court’s analysis reached only the element of causation. I would reverse the judgment on the HMOs’ damages claims because, even without Dr. Harris’s Daubert barred testimony, the HMOs have presented sufficient evidence to create a genuine issue of material fact on causation to survive- a motion for summary judgment on that issue.
First, I believe the district court erred in concluding that the Minnesota Supreme Court’s opinion “militates against” the availability of a presumption to establish causation in the consumer protection claims. 188 F.Supp.2d at 1127. In footnote 11 of its opinion answering the certified questions, the Minnesota Supreme Court suggested that the nature of the proof required to establish the element of reliance in a Minnesota consumer protection action is similar to the kind of proof required to establish a violation of the Lanham Act. The court explicitly cited cases in which plaintiffs were entitled to presumptions of consumer confusion where defendants engaged in certain intentional conduct. Nevertheless, the district court declined to consider whether similar presumptions are available to the HMOs to prove reliance where, for the purposes of this summary judgment motion, Tobacco has admitted to engaging in intentional, deceptive behavior and to spending large sums of money on advertising.
In interpreting the Minnesota Court’s statement that “in a case such as this, it will be necessary to prove reliance on those statements or conduct to satisfy the causation requirement,” 621 N.W.2d at 13, the district court correctly noted that “Plaintiffs bear the burden of proving that a causal nexus or link exists between Defendants’ conduct alleged to have violated the consumer protection laws and Plaintiffs’ claimed damages.” 188 F.Supp.2d at 1127. Nevertheless, the district court then erroneously concluded that the HMOs cannot rely upon a presumption or inference as a method of proving causation. In doing so, I conclude that the district court conflated the burden of proof with methods of proof and with the burden of production. It is clear under Minnesota law that a plaintiff may meet his initial burden of production and offer proof in the form of a rebuttable presumption. See Minn. R. Evid. 301 cmt. (“A presumption is a procedural device that satisfies the burden of producing evidence.”); see also Fed. R.Evid. 301 (“[A] presumption imposes on the party against whom it is directed the burden of going forward with evidence to rebut or meet the presumption, but does not shift to such party the burden of proof in the sense of the risk of nonpersuasion, which remains throughout the trial upon the party on whom it was originally cast.”). An unrebutted presumption is sufficient to raise a material question of fact concerning reliance, and therefore causation, so as-to defeat a motion for summary judgment.
Second, the district court mistakenly concluded that the HMOs’ circumstantial evidence of causation was not sufficiently “tethered” to their participant population to meet their burden of production on a motion for summary judgment. The Minnesota Supreme Court held that circumstantial evidence is a permissible method of proving causation in consumer protection cases. See 621 N.W.2d at 14 (“[T]he showing of reliance that must be made to prove a causal nexus need not include direct evidence of reliance by indi*765vidual consumers of defendants’ products. Rather, the causal nexus and its reliance component may be established by other direct or circumstantial evidence that the district court determines is relevant and probative as to the relationship between the claimed damages and the alleged prohibited conduct. ... [W]e reject the view [ ] that our misrepresentation in sales laws require proof of individual reliance in all actions seeking damages.”). Similarly, plaintiffs may rely on circumstantial evidence to establish an inference of causation in an antitrust case. See St. Louis Convention and Visitors Comm’n v. Nat’l Football League, 154 F.3d 851, 863 (8th Cir.1998) (applying federal antitrust law); Minn. Twins P’ship v. State ex rel. Hatch, 592 N.W.2d 847, 851 (Minn.1999) (noting that Minnesota’s antitrust statute is to be interpreted consistently with federal courts’ construction of federal antitrust law), cert. denied, 528 U.S. 1013, 120 S.Ct. 517, 145 L.Ed.2d 400 (1999).
Nevertheless, the district court concluded that the HMOs’ evidence, including surveys of citizens in other states, was not “relevant and probative.” In general, the district court was unwilling to infer that “people in Massachusetts and people in Minnesota would similarly perceive and then act on the lie [Tobacco’s misrepresentations].” 188 F.Supp.2d at 1130. Without some evidence from Tobacco that these surveys are not representative of the HMOs’ participant population, and given the nationwide scope of Tobacco’s advertising campaigns, I believe this is precisely the type of reasonable inference that plaintiffs are entitled to on a motion for summary judgment. To require more flies in the face of the Minnesota Supreme Court’s conclusion that proof of individual rebanee is unnecessary.
Because the district court recognized that “a more lenient standard of proof with regard to the amount of damages is available ... after [Plaintiffs] have shown that they have” suffered injury caused by Defendants’ wrongdoing, id. at 1128, I read the district court’s opinion as granting summary judgment on the issues of causation and injury in fact alone. The district court did not discuss the HMOs’ evidence as to damage calculation, and it appeared not to reach the question of whether the damage calculations that are in the record, if any, would be sufficient under the more lenient standard of proof which applies once causation has been established.
Our court’s opinion appears to meld the district court’s causation analysis into a new damage calculation analysis. The beginning of the opinion affirms on the basis of a failed damage calculation, but the analysis cites to the district court’s discussion of causation. Section A sets out the basis for affirmance and announces the standard of proof on the issue of damage calculation, but section B starts with the district court’s finding that the HMOs “have only one expert, Dr. Harris, who purports to provide the necessary causal link between Defendants’ alleged misconduct and Plaintiffs’ claimed damages.” Supra at 758 (emphasis added). The opinion goes on to discuss Dr. Harris’s theories and the district court’s Daubert ruling as to their admissibility to prove causation. After somewhat reluctantly affirming on the Daubert ruling, section B concludes by stating that “[b]ecause the HMOs do not have any other evidence demonstrating the amount of damages caused by Tobacco’s alleged misconduct, a factfinder in this case would have to act arbitrarily to set a damage amount.” Id. at 761 (emphasis added). However, by my reading, the district court did not reach the question of whether there was evidence as to the amount of damages because it granted summary judgment only after it deter*766mined that the HMOs’ theory of causation was flawed.
The district court’s statement that Dr. Harris was the only expert purporting to present evidence of causation does not necessarily allow this court to conclude that he is also the only expert calculating damages. Although we may affirm on any basis supported by the record, we should not appear to rely on district court findings where none were made. This court may give deference to the district court’s view that Dr. Harris’s reports were too speculative to prove causation, but there is no similar element of deference available as to their admissibility for proving the amount of damages where the district court never reached that determination and where a more lenient standard of proof applies once causation is established.
I believe the last paragraph of Section III of our court’s opinion demonstrates the inconsistency of its damages approach. In remanding the claims for in-junctive relief, the court appears to conclude that causation and injury in fact have been.established. Supra at 763-764 (“Indeed, we believe that the record contains a mountain of evidence tending to show that advertising generally causes people to begin smoking and causes current smokers to smoke more, which increases costs for the HMOs. If one concedes that a portion of the advertising was fraudulent, which Tobacco has done for the purposes of this motion, a reasonable person could infer that that fraudulent portion caused a part of those costs, even if the HMOs’ participants differed slightly from the populations used to study the effect of advertising generally on the prevalence of smoking. In other words, although the evidence in the case is, as we have said, insufficient to allow a factfinder to arrive at a reasonable estimate of the extent of harm caused, we hold that it was sufficient to raise an inference that harm has in fact been caused.”). These statements are inconsistent with the opinion’s earlier statement that the only causal link between Tobacco’s misconduct and the HMOs’ damages is Dr. Harris, whose testimony was properly excluded under Daubert. While I concur in the court’s decision to reverse and remand the judgment on the claims for injunctive relief, I find the procedural grounds mentioned by the court in the first and second paragraphs of Section III to be sufficient to reach this end.
For the reasons explained above, I would reverse the summary judgment in favor of Tobacco on the claims for damages and injunctive relief, and remand to the district court for further proceedings.