Dryvit Systems, Inc. petitions for review, pursuant to Federal Rule of Civil Procedure 23(f), of the district court's certification of a plain- tiff class of homeowners who allege that Dryvit's stucco siding prod- uct, "Fastrak System 4000" (Fastrak), was negligently designed and that Dryvit failed to warn purchasers of the product's alleged defects. This Circuit has not yet had occasion to articulate the standard gov- erning our consideration of petitions under Rule 23(f). We adopt a five-factor "sliding scale" test to guide the consideration of such peti- tions, grant the petition, and hold that certification of this class is inappropriate at this time because it is likely that individual issues necessary to adjudicate Dryvit's liability will predominate over class issues in a manner inconsistent with the dictates of Federal Rule of Civil Procedure 23(b)(3). We thus grant Dryvit's petition for review and vacate the district court's class certification order.

Fastrak is a stucco siding product that is installed on the exterior of a building in order to provide a water-resistant shield that protects the building from weather. Fastrak consists of several component parts: (a) a sheathing comprised of Georgia-Pacific Corporation's "Dens-Glass Gold" product; (b) Dryvit non-cementatious base coat (NCB); (c) reinforcing mesh; and (d) Dryvit finish coat. Dryvit sells the various components of the system to independent distributors and provides instructions and specifications to applicators, who buy the components from various suppliers and produce the final product. Applicators are usually hired by general contractors to install Fastrak on the exterior of a new building; responsibility for installing Fastrak and integrating it with other building components, such as windows, doors, flashing, and the roof is shared between applicators, general contractors, and subcontractors. Dryvit asserts that Fastrak can fail for various reasons related to the conduct of third parties, such as applica- tor error, use of incorrect materials, or improperly sealing Fastrak to other components of a building, such as windows or the roof.

Fastrak's design is based upon the premise that a properly face- sealed barrier can prevent the intrusion of water behind the siding product, eliminating the need for drainage between the siding and the wall of a structure. The plaintiff class argues that a face-sealed barrier of this sort cannot be installed correctly in the field and will inevitably lack water resistance and ability to resist applied loads.

On July 25, 1999, plaintiff Robert E. Lienhart and others (collec- tively "Lienhart") filed a complaint in the Superior Court for Wake County, North Carolina against defendant Dryvit, for themselves and on behalf of a class comprised of all entities in North Carolina whose buildings were constructed after January 1, 1992 using Fastrak. Lien- hart alleged that Dryvit had committed negligence, gross negligence, negligent misrepresentation, fraud, and unfair and deceptive trade practices. Lienhart alleged that Fastrak caused damage to their struc- tures by "trapping" water within each structure's walls. On July 22, 1999, Dryvit removed this action to the United States District Court for the Eastern District of North Carolina based on diversity of citi- zenship. On December 18, 2000, the district court certified the pro- posed class. Certification was limited to the issues of (1) whether Dryvit's product was defectively designed; and (2) whether Dryvit had breached a duty to warn homeowners of the hazards inherent in the use of its product.

Dryvit timely filed a Petition for Review pursuant to Rule 23(f) with this Court.1

This Circuit has not yet had occasion to consider the standards for granting a petition to appeal class certification under Federal Rule of Civil Procedure 23(f), a new provision added in 1998 to Rule 23, the rule governing class actions.

A court of appeals may in its discretion permit an appeal from an order of a district court granting or denying class action certification under this rule if application is made to it within ten days after entry of the order. An appeal does not stay proceedings in the district court unless the district judge or the court of appeals so orders.

Fed. R. Civ. P. 23(f).

The advisory committee note to Rule 23(f) states that "the court of appeals is given unfettered discretion whether to permit" a Rule 23(f) appeal, "akin to the discretion exercised by the Supreme Court in act- ing on a petition for certiorari." Fed. R. Civ. P. 23(f) advisory com- mittee note. Noting that "[t]he courts of appeals will develop standards for granting review that reflect the changing areas of uncer- tainty in class litigation," the advisory committee note states that "[p]ermission to appeal may be granted or denied on the basis of any consideration that the court of appeals finds persuasive." Id. The note discusses three circumstances in which such review may be especially appropriate: where an order denying certification effectively ends the litigation because the individual claims are uneconomical to litigate, so that no review will in practice be available from a final judgment; where an order granting certification effectively ends the litigation because it produces irresistible pressure on the defendant to settle; and "when the certification decision turns on a novel or unsettled question of law." Id.

The first federal court of appeals decision addressing the standards for granting a Rule 23(f) petition was the Seventh Circuit's opinion in Blair v. Equifax Check Serv., Inc., 181 F.3d 832 (7th Cir. 1999). The Seventh Circuit noted that in light of the discretionary nature of Rule 23(f), it would be inappropriate to adopt a bright-line approach which would unduly constrain the court's discretion. Id. at 834. How- ever, the court discussed and elaborated upon some circumstances which may justify the grant of a Rule 23(f) petition. First, the court stated that "when denial of class status seems likely to be fatal, and when the plaintiff has a solid argument in opposition to the district court's decision," review is warranted. Id. In this area, however, the Seventh Circuit cautioned that courts entertaining such "death knell" Rule 23(f) claims must be "wary lest the mind hear a bell that is not tolling," noting that even where individual claims would ordinarily be too small to litigate individually, law firms with substantial resources may pursue an individual claim to final judgment in the hope of win- ning a favorable ruling on class certification on appeal from final judgment. Id. Second, the court stated that review may be appropriate where a grant of class certification creates irresistible pressure on the defendant to settle, especially where the plaintiff's probability of suc- cess on the merits is slight. Id. Noting that "some plaintiffs or even some district judges may be tempted to use the class device to wring settlements from defendants whose legal positions are justified but unpopular," the court suggested that the "interaction of procedure with the merits justifies an earlier appellate look."2 Id. Third, the Seventh Circuit held that whether hearing an appeal would help to "facilitate the development of the law" by clearing up a significant unresolved issue also is relevant. Id. at 835. The court reasoned that one of the purposes of Rule 23(f) was to promote the development of case law regarding certain fundamental class action issues which tend to be under-litigated because so many class actions are settled or oth- erwise resolved in ways that preclude eventual appellate consideration of procedural issues. The court noted that when the basis for granting review is to contribute to the development of the law, "it is less important to show that the district judge's decision is shaky," as affir- mances as well as reversals may serve to illuminate important ques- tions. Id. Thus, the court granted the Rule 23(f) petition before it to resolve issues related to the propriety of multiple, overlapping, con- currently litigated class actions, but it ultimately affirmed the district court's certification order. Id. at 838-39.

The First Circuit, in Waste Management Holdings, Inc. v. Mow- bray, 208 F.3d 288 (1st Cir. 2000), adopted several variations on the Seventh Circuit's approach in Blair. The court noted that it "ordinar- ily will grant leave to appeal when a Rule 23(f) application falls into one of the three [Blair] categories," but noted its ongoing discretion in this area. Id. at 294. The court also restricted the third Blair cate- gory to those instances in which "an appeal will permit the resolution of an unsettled legal issue that is important to the particular litigation as well as important in itself and likely to escape effective review if left hanging until the end of the case." Id. The stated rationale for this additional restriction was that otherwise it would be too easy for cre- ative lawyers to argue that their case involves a"fundamental" issue, undermining the objective of making Rule 23(f) review "the excep- tion, not the rule." Id.

In Prado-Steiman v. Bush, 221 F.3d 1266 (11th Cir. 2000), the Eleventh Circuit further elaborated upon Rule 23(f) standards. Noting that given the volume of class action litigation in the district courts, routine interlocutory appellate review of class certifications "is simply not practicable," the court announced several guideposts to be used in evaluating Rule 23(f) petitions. Id. at 1273-77. Recognizing that Rule 23(f) was explicitly promulgated to replace the use of mandamus in reviewing manifestly erroneous class certifications, the court stated that it would be inappropriate to review such petitions under standards as stringent as those which govern mandamus. Id. at 1275 n.10. Accordingly, the court set forth five factors that are relevant to the appropriateness of granting a Rule 23(f) petition: (1) whether the cer- tification ruling is likely dispositive of the litigation; (2) whether the district court's certification decision contains a substantial weakness; (3) whether the appeal will permit the resolution of an unsettled legal question of general importance; (4) the nature and status of the litiga- tion before the district court (such as the presence of outstanding dis- positive motions and the status of discovery); and (5) the likelihood that future events will make appellate review more or less appropri- ate. Id. at 1274-76.

In explaining the second factor, "substantial weakness in the class certification," the Eleventh Circuit noted that"interlocutory review may be appropriate when it promises to spare the parties and the dis- trict court the expense and burden of litigating the matter to final judgment only to have it inevitably reversed . . . on an appeal." Id. at 1274-75. The Eleventh Circuit held that this factor is to be viewed "on a sliding scale" with the other factors, such that the stronger the case that class certification was a clear abuse of discretion, the weaker a showing is needed on the other factors to gain review. Id. at 1274-75. As to the third factor, the court noted that the impact of the questions at issue to related actions involving the same or similarly-situated par- ties is an appropriate part of the inquiry relative to this prong. Id. at 1275. The Eleventh Circuit explained its fifth factor, the "likelihood that future events may make appellate review more or less appropri- ate," by reference to issues such as the prospect of imminent settle- ment negotiations or the status of an action as the first-litigated of a number of similar claims. Id.

Dryvit reads Prado-Steiman to suggest that a sufficient showing of "substantial weakness," the third factor, alone can constitute grounds for granting a Rule 23(f) motion. Dryvit also embraces Prado- Steiman's somewhat more expansive list of factors capable of sup- porting a grant of a Rule 23(f) petition. Lienhart, on the other hand, argues that Mowbray, with its emphasis on the importance of an unsettled issue of law of general importance, substantially narrowed the Blair approach, and contends that Prado-Steiman further limited Mowbray's approach. (Respondent's Answer, at 16 (asserting that "each successive Court of Appeals that has dealt with Fed. R. Civ. P. 23(f) has limited and narrowed the availability of such an appeal")).

We do not believe that Prado-Steiman limited Mowbray; to the contrary, by adding the weakness of the district court's certification decision as an independent factor supporting review and noting that the impact of a question raised in a Rule 23(f) petition on related liti- gation can favor review, the Prado-Steiman court broadened the bases for a grant of review. See Prado-Steiman , 221 F.3d at 1275 (Rule 23(f) review may be appropriate to "spare the parties and the district court the expense and burden of litigating the matter to final judgment only to have it inevitably reversed"). In light of Rule 23(f)'s purpose to eliminate the unduly restrictive review practices which obtained when mandamus was the only available means to review a class certification prior to final judgment in the absence of a district court's decision to voluntarily certify the issue for immediate review, we do not believe the exceptionally stringent standards for review urged by Lienhart are appropriate. Standards certainly must reflect the limited capacity of appellate courts to consider interlocutory appeals, as well as the institutional advantage possessed by district courts in managing the course of litigation and the judicial diseconomy of per- mitting routine interlocutory appeals. Routine interlocutory review of class certifications is simply not feasible as a practical matter. See id. at 1273 (noting that "there are too many class actions filed each year for federal appeals courts practicably to adjudicate class certification decisions on an interlocutory basis as a matter of course"). However, we must remain cognizant that Rule 23(f) was enacted by the Supreme Court to permit such appeals, pursuant to an express grant of authority by Congress to create appellate jurisdiction over non- final judgments. See 28 U.S.C.A. S 1292(e). In addition to addressing "death knell" situations and promoting the resolution of legal ques- tions of general importance, a careful and sparing use of Rule 23(f) may promote judicial economy by enabling the correction of certain manifestly flawed class certifications prior to trial and final judgment.

Where a district court's certification decision is manifestly errone- ous and virtually certain to be reversed on appeal, the issues involved need not be of general importance, nor must the certification decision constitute a "death knell" for the litigation. Such a rule would waste, rather than conserve, judicial resources, because self-evidently defec- tive classes would proceed through trial to final judgment, only to face certain decertification on appeal and a requirement that the pro- cess begin again from square one. The weakness of the district court's certification, viewed in terms of the likelihood of reversal under an abuse of discretion standard, operates on a "sliding scale" in conjunc- tion with the other factors enumerated by the Eleventh Circuit in Prado-Steiman. In extreme cases, where decertification is a functional certainty, the weakness of the certification order may alone suffice to permit the Court of Appeals to grant review. Conversely, when the weakness of the district court's certification is less substantial (or insubstantial), a commensurately stronger showing on the other fac- tors is necessary to support review. We thus adopt the Eleventh Cir- cuit's five-factor Prado-Steiman test for judging the appropriateness of granting a petition for review under Rule 23(f), and hold that the "substantial weakness" prong operates on a sliding scale to determine the strength of the necessary showing regarding the other Prado- Steiman factors. See Prado-Steiman, 221 F.3d at 1274-76.

Having adopted, with some elaboration, the five-factor Prado- Steiman test for determining when to grant a Rule 23(f) petition, we now apply this test to the Rule 23(f) petition before us. The grant of oral argument in this case for the purpose of addressing the proper Rule 23(f) standard -- an issue of first impression in this Circuit -- has of necessity led to an extensive examination of the merits of class certification. For the reasons set forth below, and after extensively reviewing the merits of the district court's certification, we conclude that, on the current state of the factual record, the proposed class certi- fication is manifestly erroneous and is certain to be vacated on appeal from final judgment. Thus, under the sliding-scale approach we have adopted, review is appropriate without regard to the other factors in order to avoid the judicial diseconomy entailed in allowing a class certification which is clearly infirm to be litigated to final judgment only to face vacatur on appeal. Therefore, we grant the petition for review under Rule 23(f).3

A district court "has broad discretion in deciding whether to certify a class, but that discretion must be exercised within the framework of Rule 23." In re American Med. Sys., Inc., 75 F.3d 1069, 1079 (6th Cir. 1996). Actions such as this one, brought under Federal Rule of Civil Procedure 23(b)(3), may be brought only if the numerosity, commonality, typicality, representativeness, predominance, and supe- riority requirements of both Rule 23(a) and (b)(3) are met. Fed. R. Civ. P. 23. The party seeking class certification bears the burden of proof. Int'l Woodworkers of Am. v. Chesapeake Bay Plywood Corp., 659 F.2d 1259, 1267 (4th Cir. 1981).

In reviewing the Rule 23(a) requirements, we note that numerosity requires that a class be so large that "joinder of all members is impracticable." Fed. R. Civ. P. 23(a)(1). Commonality requires that "there are questions of law or fact common to the class." Fed. R. Civ. P. 23(a)(2). The common questions must be dispositive and over- shadow other issues.4 Stott v. Haworth, 916 F.2d 134, 145 (4th Cir. 1990). Typicality requires that the claims of the named class represen- tatives be typical of those of the class; "a class representative must be part of the class and possess the same interest and suffer the same injury as the class members." General Tel. Co. of Southwest v. Falcon, 457 U.S. 147, 156 (1982) (internal quotation marks omitted). Representativeness requires that the class representatives "will fairly and adequately protect the interests of the class." Fed. R. Civ. P. 23(a)(4). We have stated that the final three requirements of Rule 23(a) "tend to merge, with commonality and typicality `serv[ing] as guideposts for determining whether . . . maintenance of a class action is economical and whether the named plaintiff's claim and the class claims are so interrelated that the interests of the class members will be fairly and adequately protected in their absence.'" Broussard v. Meineke Discount Muffler Shops, Inc., 155 F.3d 331, 337 (4th Cir. 1998) (quoting Falcon, 457 U.S. at 157 n.13).

In contrast to actions under Rule 23(b)(1) and (b)(2), Rule 23(b)(3) actions are "[f]ramed for situations in which class-action treatment is not clearly called for," but "may nevertheless be convenient and desir- able." Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 615 (1997) (internal quotation marks omitted). In addition to the four Rule 23(a) requirements, Rule 23(b)(3) actions such as this one must meet two requirements: predominance and superiority. Predominance requires that "[common] questions of law or fact . . . predominate over any questions affecting only individual members." Fed. R. Civ. P. 23(b)(3). The predominance inquiry "tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation." Amchem, 521 U.S. at 623. Superiority requires that a class action be "superior to other methods for the fair and efficient adjudication of the controversy." Fed. R. Civ. P. 23(b)(3).

In Broussard, we addressed the predominance, commonality and typicality difficulties that arise under Rule 23 when a myriad of plain- tiffs attempt to consolidate legally and factually different claims via the class action device. Addressing a class comprised of various fran- chisees who signed different contracts at different times and had dif- ferent levels of compensable injury, relied on different alleged misrepresentations, and based statute of limitations tolling arguments on different facts, we held that "the class . . . was thus no more than a hodgepodge of factually as well as legally different plaintiffs that should not have been cobbled together for trial." Broussard, 155 F.3d at 343 (internal quotation marks omitted). Relevantly, we noted that while the need for individualized proof of damages does not necessar- ily preclude class certification so long as common issues continue to predominate over individual issues, it is impermissible to determine damages on a classwide basis in order to facilitate class treatment of a case when the governing law requires individualized proof of dam- ages. Id. at 343. This Court has further held that the need for individu- alized proof of damages may defeat predominance where proof of damages is essential to liability. Windham v. American Brands, 565 F.2d 59, 66 (4th Cir. 1977).

Lienhart argues that proof of individual damages is unnecessary because the theory of this case is that publicity surrounding problems with the Fastrak product has caused "stigma damage" in the form of a diminished market value for homes clad with Fastrak, independent of whether these homes have in fact suffered any tangible damage. It also appears, however, that the complaint alleges compensable dam- ages deriving from the actual failure of Fastrak and resultant physical damage to the homes of some plaintiff class members. The district court suggested that damages subclasses might be necessary at a later date.

The district court placed great emphasis on the fact that Lienhart does not assert claims against third parties or allege that the conduct of third parties contributed to the failure of the Fastrak product. In the district court's view, this fact distinguished the instant case from that in In re Stucco Litig., 175 F.R.D. 210 (E.D.N.C. 1997), in which the need to assess the contribution of third-party contractors to the failure of a different stucco siding product was held by the same district court to destroy predominance and preclude class certification. Id. at 215. The district court held that because Lienhart alleged claims only against Dryvit, issues regarding the conduct of third parties could be resolved in subsequent contribution and indemnity actions and thus did not preclude class certification.

The difficulty with this holding is that under North Carolina law, the failure of a product user to follow express and adequate instruc- tions or warnings absolutely precludes liability in a products liability action. N.C. Gen. Stat. S 99B-4(1) (1999). Dryvit thus argues that the alleged failure of applicators and contractors to follow its installation instructions does more than give rise to a subsequent contribution claim; such failure, if proven, would entirely preclude Dryvit's liabil- ity. As a result, Dryvit asserts, this case differs from In re Stucco only in that the third-party conduct issues will arise by means of Dryvit's statutory affirmative defenses, rather than the plaintiff class's direct claims.

N.C. Gen. Stat. S 99B-4(1) "merely codif[ies] the doctrine of con- tributory negligence." Champs Convenience Stores, Inc. v. United Chem. Co., Inc., 406 S.E.2d 856, 860 (N.C. 1991); see also Lee v. Crest Chem. Co., 583 F. Supp. 131 (M.D.N.C. 1984) (concluding that S 99B-4(1) codifies the doctrine of contributory negligence in prod- ucts liability cases). In a closely analogous case to this one, the North Carolina Court of Appeals has held that S 99B-4(1) bars liability on the part of a manufacturer of a roofing system if a contractor improp- erly installed a roofing system, but only if the manufacturer has not "contracted to instruct [the contractor] on installation procedures, and . . . in fact assisted [the contractor] in the installation of the roof." Westover Prods., Inc. v. Gateway Roofing Co., Inc. , 380 S.E.2d 369, 374 (N.C. Ct. App. 1989). Thus, if Dryvit contracted to instruct third parties and assisted those third parties in installing Fastrak, the third parties' failure to follow instructions cannot provide Dryvit with an affirmative defense to liability under North Carolina law. If, on the other hand, a given third party with whom Dryvit did not contract to provide instruction on installation procedures and whom Dryvit did not assist with installation failed to follow Dryvit's instructions, and this failure caused some of the damage suffered by a particular plain- tiff, such failure would provide Dryvit with a complete defense to lia- bility to that plaintiff under North Carolina law. Id.

Thus, if third parties contributed to the failure of Fastrak and Dry- vit did not contract to provide instruction to these parties and assist them in installing Fastrak, Dryvit's claims regarding these parties are not contribution and indemnity claims which may be addressed in a second stage after this litigation, but are instead claims regarding affirmative defenses which, if established, negate Dryvit's liability ab initio. Given this conclusion, even if all plaintiffs suffered some "stigma" damage due to Fastrak's reputation, two problems in certify- ing a class are apparent.

First, if this "stigma" damage coincides with damage from actual failure of the Fastrak product or expected future failure, and if this actual or expected failure is due in part to applicator or contractor failure to follow instructions, Dryvit cannot be held liable to plaintiffs for whom this is true. The "stigma" injury attributable to a home being clad with Fastrak per se is, in such a case, inseparable from the injury attributable to a home being clad with improperly installed Fastrak which, as a result of improper installation, either has failed or will fail earlier or more severely than would otherwise be the case. Under the law of contributory negligence, such negligence need not cause all of the alleged injury; if contributory negligence causes any of a plain- tiff's injury, recovery is barred. See McGill v. French, 424 S.E.2d 108, 113 (N.C. 1993) (stating that to establish contributory negli- gence, "the defendant must show that the plaintiff's injuries were proximately caused by his own negligence," defined as "the injury complained of"). Indeed, it is this principle that distinguishes contrib- utory from comparative negligence. This principle cannot be avoided by artificially segmenting the injury into "stigma" damage caused entirely by Dryvit and other damage caused by applicators and con- tractors. Here, the "injury complained of," on the face of the com- plaint, appears to encompass both stigma damage and actual physical damage caused by the failure of Fastrak in some installations. As a result, defining the parameters of a class for whom Dryvit's liability can be determined based on classwide proof may require the district court to probe deeply into the individualized details of applicator and contractor behavior with respect to each Fastrak installation, because Dryvit cannot be adjudged liable to class members whose injuries were caused in part by applicators and contractors whom Dryvit nei- ther contracted to train nor assisted. If such an individualized inquiry is needed to determine the membership of a workable class, it is clear that common issues do not predominate over individual issues as required by Rule 23(b)(3). As in Windham v. American Brands, Inc., 565 F.2d 59, 66 (4th Cir. 1977), individualized proof of damages -- in this case, individualized proof of damage causation -- would be essential to liability, destroying predominance. The functional equiva- lent of a full-blown trial on damages causation for each putative class member would be required to determine to which individuals Dryvit is liable as a result of the class suit's result. We do not believe that such a class can meet the prerequisites of Rule 23(b)(3).

Second, even for plaintiffs whose damage is wholly of the "stigma" variety, the contributory negligence of applicators and contractors may be a partial cause of any "stigma" attached to the Fastrak prod- uct. In other words, Fastrak's market reputation may be poor in part due to a tendency of applicators and contractors to install the product in a manner incompatible with Dryvit's instructions. While classwide proof would appear to be possible on this issue, and thus, it does not per se destroy predominance, the district court's judgment that this class action is feasible from a manageability standpoint depends in part on the district court's belief that it will be possible to avoid the issues of applicator and contractor behavior entirely during the liabil- ity phase. Our conclusion regarding the application of N.C. Gen. Stat. S 99B-4 casts considerable doubt on this assumption.

We thus conclude that class certification in this case is not feasible under Rule 23 if Dryvit did not contract with applicators and contrac- tors to instruct these parties regarding the proper installation of Fastrak and did not assist these parties in installing the product, and therefore, N.C. Gen. Stat. S 99B-4 applies to bar Dryvit's liability to any class member whose injuries were caused in part by applicator and contractor failure to follow instructions. The current state of the record is not entirely clear regarding the details of Dryvit's relation- ship with contractors and applicators, and thus, the propriety of the district court's class certification turns on a factual determination best made by the district court in the first instance. As a result, we vacate the district court's class certification and remand with instructions to reconsider the propriety of class certification in light of the fact that if N.C. Gen. Stat. S 99B-4 applies on the facts of this case, it poses major -- and perhaps insuperable -- obstacles to the feasibility of certifying this proposed class.