holding that a trial court must resolve challenges to an expert's qualifications as well as “any challenge to the reliability of information provided by an expert if that information is relevant to establishing any of the Rule 23 requirements for class certification”Summary of this case from In re Blood Reagents Antitrust Litig.
Submitted December 30, 2009.
Decided April 7, 2010.
Appeal from the United States District Court for the Northern District of Illinois, Nan R. Nolan, United States Magistrate Judge.
David B. Johnson, Attorney (submitted), Sidley Austin, Chicago, IL, for Petitioners.
Timothy P. Mahoney, Attorney, Neal, Gerber Eisenberg, Chicago, IL, for Respondents.
Before POSNER, EVANS, and TINDER, Circuit Judges.
American Honda Motor Company and Honda of America Manufacturing (collectively "Honda") seek leave to appeal the district court's grant of class certification pursuant to Federal Rule of Civil Procedure 23(f). Specifically, Honda asks us to resolve whether the district court must conclusively rule on the admissibility of an expert opinion prior to class certification in this case because that opinion is essential to the certification decision. Since this is the type of question that Rule 23(f) was designed to address, and because the district court's analysis was incomplete, we accept the appeal. See Allen v. Int'l Truck Engine Corp., 358 F.3d 469, 470 (7th Cir. 2004) ("The parties' comprehensive submissions show not only that immediate review would promote the development of the law . . . but also that the district court committed an error best handled by a swift" action.).
Plaintiffs are purchasers of Honda's Gold Wing GL1800 motorcycle; they allege that the motorcycle has a design defect that prevents the adequate dampening of "wobble," that is, side-to-side oscillation of the front steering assembly about the steering axis. In other words, they claim that the defect makes the steering assembly shake excessively and they want Honda to fix the problem. Plaintiffs moved for class certification pursuant to Rule 23(b)(3). To demonstrate the predominance of common issues, they relied heavily on a report prepared by Mark Ezra, a motorcycle engineering expert. Ezra's report opined that motorcycles should "by [their] design and manufacture exhibit decay of any steering oscillations sufficiently and rapidly so that the rider neither reacts to nor is frightened by such oscillations." Assuming that human reaction time to wobble is ½ to ¾ of a second, Ezra opined that wobble should decay, or dissipate, to 37% of its original amplitude within ¾ of a second to ensure that riders do not perceive and react to the oscillations. This standard, which Ezra devised himself and characterizes as "reasonable," was published in the June 2004 edition of the Journal of the National Academy of Forensic Engineers. After testing one used GL1800 serviced to factory condition, Ezra concluded that it failed to meet his wobble decay standard. He also concluded that his standard could be achieved in the GL1800 motorcycle by replacing the regular ball bearings in the steering assembly with tapered ones.
Honda moved to strike the report pursuant to Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), arguing that Ezra's wobble decay standard was unreliable because it was not supported by empirical testing, was not developed through a recognized standard-setting procedure, was not generally accepted in the relevant scientific, technical, or professional community, and was not the product of independent research. In the alternative, Honda argued that even if the standard was reliable, Ezra did not reliably apply it to this case because he only tested one motorcycle and did not account for variables that could affect the wobble decay rate.
The district court concluded that it was proper to decide whether the report was admissible prior to certification because "most of Plaintiffs' predominance arguments rest upon the theories advanced by Mr. Ezra." Allen v. Am. Honda Motor Co., 264 F.R.D. 412, 425 (N.D.Ill. 2009). The court then discussed Honda's Daubert arguments. It noted that it was concerned that, among other things, Ezra's wobble decay standard may not be supported by empirical evidence, the standard has not been generally accepted by the engineering community, and Ezra's test sample of one may be inadequate to conclude that the entire fleet of GL1800s is defective. The court then concluded, "Viewing all of the arguments together, the court has definite reservations about the reliability of Mr. Ezra's wobble decay standard. Nevertheless, the court declines to exclude the report in its entirety at this early stage of the proceedings." Id. at 428. The court denied Honda's motion to exclude "without prejudice," id. at 437, and granted Plaintiffs' motion for class certification in part, certifying two classes of individuals who purchased GL1800s.
In Szabo v. Bridgeport Machs., Inc., 249 F.3d 672, 676 (7th Cir. 2001), we held that a district court must make whatever factual and legal inquiries are necessary to ensure that requirements for class certification are satisfied before deciding whether a class should be certified, even if those considerations overlap the merits of the case. And in West v. Prudential Sec., Inc., 282 F.3d 935, 938 (7th Cir. 2002), we held that a plaintiff cannot obtain class certification just by hiring a competent expert. We emphasized, "A district judge may not duck hard questions by observing that each side has some support. . . . Tough questions must be faced and squarely decided, if necessary by holding evidentiary hearings and choosing between competing perspectives." Id. But we have not yet specifically addressed whether a district court must resolve a Daubert challenge prior to ruling on class certification if the testimony challenged is integral to the plaintiffs' satisfaction of Rule 23's requirements.
Plaintiffs argue that we do not need to accept this appeal because district courts in this circuit generally agree that a Daubert challenge must be resolved prior to class certification. This is true, though the courts have relied on different authority, including other district court decisions, to reach their conclusion. See, e.g., Reed v. Advocate Health Care, No. 06 C 3337, 2009 WL 3146999, at *21 (N.D.Ill. Sept.28, 2009); In re Ready-Mixed Concrete Antitrust Litig., 261 F.R.D. 154, 162 (S.D.Ind. 2009); Srail v. Village of Lisle, 249 F.R.D. 544, 557, 560-61 (N.D.Ill. 2008); cf. Szabo, 249 F.3d at 675 (accepting an appeal pursuant to Rule 23(f) in part because the district court relied largely on decisions by other district judges "[a]t critical junctures"). Further, in many of these cases, including the case at hand, this issue was heavily contested. Finally, other appellate courts have not directly addressed whether challenges to experts must be resolved prior to certification. See, e.g., In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 315 n. 13 (3d Cir. 2008) (noting that parties agreed that lower court properly addressed admission of expert testimony); Dukes v. Wal-Mart, Inc., 509 F.3d 1168, 1174, 1179 (9th Cir. 2007) (withdrawing and superceding prior opinion, 474 F.3d 1214, which held that full Daubert examination should not be conducted at class certification stage), reh'g granted, 556 F.3d 919 (9th Cir. 2009); In re Initial Public Offering Sec. Litig., 471 F.3d 24, 42 (2d Cir. 2006) (disavowing suggestion that expert's testimony may establish component of class certification "simply by being not fatally flawed"). Given the uncertainty surrounding the propriety of conducting a Daubert analysis at the class certification stage, and the frequency with which this issue arises, we find the question to be one appropriate for resolution under Rule 23(f). See Blair v. Equifax Check Servs., Inc., 181 F.3d 832, 835 (7th Cir. 1999).
We hold that when an expert's report or testimony is critical to class certification, as it is here, see Allen, 264 F.R.D. at 420 ("Mr. Ezra's wobble decay standard . . . forms the basis of Plaintiffs' theory of defect."), a district court must conclusively rule on any challenge to the expert's qualifications or submissions prior to ruling on a class certification motion. That is, the district court must perform a full Daubert analysis before certifying the class if the situation warrants. If the challenge is to an individual's qualifications, a court must make that determination "by comparing the area in which the witness has superior knowledge, skill, experience, or education with the subject matter of the witness's testimony." Carroll v. Otis Elevator Co., 896 F.2d 210, 212 (7th Cir. 1990). The court must also resolve any challenge to the reliability of information provided by an expert if that information is relevant to establishing any of the Rule 23 requirements for class certification.
Here, the district court started off on the right foot by beginning to undertake what might have become a fairly extensive Daubert analysis. Cf. Naeem v. McKesson Drug Co., 444 F.3d 593, 608 (7th Cir. 2006) (noting that there is no need for a court to perform the Daubert analysis in any particular or mechanical way). It noted its role as "gatekeeper" and its duty to "determine reliability in light of the proposed expert's full range of experience and training as well as the methodology used to arrive at a particular conclusion." Allen, 264 F.R.D. at 423 (quotations omitted). The district court acknowledged Honda's concerns about the reliability of Ezra's testimony and largely agreed with them. It expressed reservations about Ezra's failure to "establish the minimal amplitude required for a rider to detect an oscillation," id. at 426, his failure to "verif[y] whether a lesser or greater percentage of decay would also provide an appropriate margin of safety," id. at 427, the fact that his wobble decay standard was developed "to assist with a lawsuit and was not conceived through the logical flow of independent research," id., the questionable peer-review process that his article underwent, see id., the engineering community's lack of acceptance of his proposed standard, id., and his test sample size of one used GL1800, see id. at 427-28. Yet the district court ultimately declined, without further explanation, "to exclude the report in its entirety at this early stage of the proceedings." Id. at 428.
"We give the court great latitude in determining not only how to measure the reliability of the proposed expert testimony but also whether the testimony is, in fact, reliable, but the court must provide more than just conclusory statements of admissibility to show that it adequately performed the Daubert analysis." United States v. Pansier, 576 F.3d 726, 737 (7th Cir. 2009) (citation omitted). The court's effective statement of admissibility here is not even conclusory; it leaves open the questions of what portions of Ezra's testimony it may have decided (or will decide) to exclude, whether Ezra reliably applied the standard to the facts of the case, and, ultimately, whether Plaintiffs have satisfied Rule 23(b)(3)'s predominance requirement. As a result, the district court never actually reached a conclusion about whether Ezra's expert report was reliable enough to support Plaintiffs' class certification request. Instead it denied Honda's motion to exclude without prejudice and noted that the case was in an "early stage of the proceedings." This was not sufficient. Indeed, it was an abuse of discretion. See Nightingale Home Healthcare, Inc. v. Anodyne Therapy, LLC, 589 F.3d 881, 883 (7th Cir. 2009) ("It is an abuse of discretion not to exercise discretion."); Smith v. Ford Motor Co., 215 F.3d 713, 717 (7th Cir. 2000) ("We review for abuse of discretion the district court's choice of factors to include within [the Daubert] framework as well as its ultimate conclusions regarding the admissibility of expert testimony.").
As we have explained, a district court must make the necessary factual and legal inquiries and decide all relevant contested issues prior to certification. See West, 282 F.3d at 938; Szabo, 249 F.3d at 676. The district court's actions here were more akin to the "provisional" approach that we rejected in Szabo. 249 F.3d at 676. Ezra's testimony is necessary to show that Plaintiffs' claims are capable of resolution on a class-wide basis and that the common defect in the motorcycle predominates over the class members' individual issues. Therefore, by failing to clearly resolve the issue of its admissibility before certifying the class, the district court erred. Cf. West, 282 F.3d at 938 ("A district judge may not duck hard questions by observing that each side has some support. . . . Tough questions must be faced and squarely decided. . . ." (emphasis added)).
The district court was reluctant to exclude Ezra's report "in its entirety at this early stage of the proceedings," Allen, 264 F.R.D. at 428, but our examination of the record reveals that exclusion is the inescapable result when the Daubert analysis is carried to its conclusion. Under the Daubert framework, a district court must determine "whether a given expert is qualified to testify in the case in question and whether his testimony is scientifically reliable." Gayton v. McCoy, 593 F.3d 610, 616 (7th Cir. 2010). Ezra's qualifications are not at issue here; the reliability of his testimony is what Honda contests. But even the most "supremely qualified expert cannot waltz into the courtroom and render opinions unless those opinions are based upon some recognized scientific method and are reliable and relevant under the test set forth by the Supreme Court in Daubert." Clark v. Takata Corp., 192 F.3d 750, 759 n. 5 (7th Cir. 1999).
Daubert sets forth a non-exhaustive list of guideposts to consult in assessing the reliability of expert testimony: (1) whether the scientific theory can be or has been tested; (2) whether the theory has been subjected to peer review and publication; and (3) whether the theory has been generally accepted in the relevant scientific, technical, or professional community. Daubert, 509 U.S. at 593-94, 113 S.Ct. 2786. In addition to these guideposts, the 2000 Advisory Committee's Notes to Rule 702 suggest other benchmarks for gauging expert reliability, including whether the testimony relates to "matters growing naturally and directly out of research they have conducted independent of the litigation, or whether they have developed their opinions expressly for purposes of testifying"; "[w]hether the expert has adequately accounted for obvious alternative explanations"; and "[w]hether the expert is being as careful as he would be in his regular professional work outside his paid litigation consulting." Fed.R.Evid. 702 Advisory Committee's Notes (2000 Amends.); see also Fuesting v. Zimmer, Inc., 421 F.3d 528, 534-35 (7th Cir. 2005), vacated in part on other grounds, 448 F.3d 936 (7th Cir. 2006).
The "theory" here is Ezra's wobble decay standard, and, as the district court thoroughly enumerated, there are many reasons to harbor "definite reservations" about its reliability. Allen, 264 F.R.D. at 428. Ezra originally developed the standard for use in a mid-1980s lawsuit in which he testified as an expert against Honda and subsequently published it in a journal article aimed at forensic engineers who testify as experts on motorcycle instability, see Mark A.M. Ezra, Forensic Engineering Investigation of Motorcycle Instability Induced Crashes, 21 J. Nat'l Acad. Forensic Eng'rs 69, 80-84 (2004) (discussing seven common "attacks by opposing counsel," including "You Did What While Testing the Motorcycle?!", and instructing future experts to "be ready to defend in simple lay terms the standard [they have] proposed and relied upon in evaluating the motorcycle and its reasonableness"). Despite its publication, there is no indication that Ezra's wobble decay standard has been generally accepted by anyone other than Ezra. See Porter v. Whitehall Labs., Inc., 9 F.3d 607, 613 (7th Cir. 1993) ("A known technique that has gained only a minimal following may be viewed with some skepticism."). Indeed, Ezra's article merely "suggested" a standard for wobble decay; it acknowledged that he was "unaware of any governmental, industry or [Society of Automotive Engineers] standards determining acceptable response characteristics for [motorcycles] . . . in . . . wobble mode," Ezra, supra, at 78, and noted that "it is up to the investigating forensic engineer to define a reasonable standard that he may defend in the legal forum before opposing council [sic] or a jury," id. at 79.
Even if we were to assume that Ezra's standard is generally accepted by mere virtue of its publication in a peer-reviewed journal, its reliability remains in question. Ezra has never conducted any rider confidence studies to determine when motorcycle riders perceive wobble, or performed any tests to determine the minimal wobble amplitude at which riders detect oscillation. See Allen, 264 F.R.D. at 426; see also West, 282 F.3d at 939 (expressing skepticism in a model that "has not been verified empirically"). His report merely deemed "reasonable" his proposed standard, relying solely on his own previous (and similarly unsupported) assessment of the same for support. The "principles and methodology" underlying his findings, Winters v. Fru-Con, Inc., 498 F.3d 734, 742 (7th Cir. 2007), then, are questionable at best. And although we do not consider the validity or accuracy of the conclusions Ezra reached, see id., we note that the methodological omissions here render the 37% wobble decay standard Ezra articulated "somewhat speculative," Allen, 264 F.R.D. at 427; see Chapman v. Maytag Corp., 297 F.3d 682, 686-87 (7th Cir. 2002) (noting that courts must consider "whether the testimony has been subjected to the scientific method, ruling out any subjective belief or unsupported speculation").
The methodology underlying the tests Ezra conducted to determine whether the GL1800 met his standard also gives us pause. Ezra tested a single, used 2006 GL1800, ridden by a single test rider, and extrapolated his conclusions to the fleet of GL1800s produced from 2001 to 2008. "Determining the minimum sample size from which reliable extrapolations can be made to the sampled population is tricky," DeKoven v. Plaza Assocs., 599 F.3d 578, ___, at *4 (7th Cir. 2010), but a sample size of one is rarely, if ever, sufficient. That is especially true when deposition testimony from putative class members reveals that different GL1800s handle and manifest the wobble problem differently, and when the handling — and wobble — of motorcycles is known to vary with the rider and road conditions. The small sample size also highlights the constraints litigation placed upon Ezra's methods and professional judgment; Ezra was not being as thorough as he might otherwise be due to Plaintiffs' reluctance to pay for more testing. See Smith, 215 F.3d at 719 ("The trial court must . . . `make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.'" (quoting Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999))).
"`[S]haky' expert testimony may be admissible, assailable by its opponents through cross-examination," Gayton, 593 F.3d at 616, but the testimony proffered here is not merely shaky: it is unreliable. And expert testimony that is not scientifically reliable should not be admitted, even "at this early stage of the proceedings." Allen, 264 F.R.D. at 428. Without Ezra's testimony, Plaintiffs are left with too little to satisfy Rule 23(b)(3)'s predominance prong. The named plaintiffs have presented evidence that their GL1800s wobble, that 49 declarants' GL1800s also wobble, that all GL1800s have the same ball bearing, that over 11,000 aftermarket kits containing tapered ball bearings and marketed to fix the GL1800 wobble have been sold, and that some plaintiffs and declarants fixed their wobble problems by replacing the factory-installed ball bearings with tapered ones. See Pl. Answer to Petition for Leave to Appeal at 20. Without Ezra's testimony, which asserts that the factory-installed ball bearings are responsible for an "increased wobble mode decay time" in all GL1800s, Plaintiffs are not only unable to support their theory that all GL1800 motorcycles use ball bearings that fail to adequately dampen wobble, they are also unable to demonstrate that their wobble claim sufficiently predominates as to warrant class certification under Rule 23(b)(3).
We therefore GRANT Honda's petition for leave to appeal, VACATE the district court's denial of Honda's motion to strike and its order certifying a class, and REMAND for proceedings consistent with this opinion.