ABOUSSIE, Justice.
Ford Motor Company, Inc., Leif Johnson Ford, Inc., and Fred Capedeville (collectively, “Ford”) complain in this interlocutory appeal that the trial court incorrectly certified appellees’ suit as a class action. We will modify the order, and affirm it as modified.
The Trial Court Action
Apellees purchased various Ford models on which the paint peeled:
Appellee Car
Sale date First noticed peeling after sale
Barry Sheldon 1987 Ranger January 1987 18-22 months
B.J. Sanders 1989 Ranger September 1989 9 months
John Porter
1988 F-150 March 1988 18 months
Matthew Reuter 1989 F-150 December 1988 4 years
William Dobbs 1990 F-150 February 1990 4 years
Margaret Dunayer 1991 F-150 December 1990 17-18 months
James Beasley 1990 Bronco January 1990 4 years
Appellees claim the cars peeled because the paint process was defective, that Ford knew of the defect, and that Ford nevertheless persisted in using the process and selling the affected cars to Texas consumers. Appellees seek to recover for breach of the implied warranty of merchantability and violations of the deceptive trade practices act. The alleged DTPA violations include representing that goods have characteristics they do not; representing that goods are of a particular standard, quality, or grade when they are not; and failing to disclose information about goods known at the time of the transactions and intending by that nondisclosure to induce transactions that would not have occurred but for the nondisclosure. Appellees contend Ford’s conduct was an unconscionable action or course of action.
The trial court certified the ease as a class action on behalf of the following class of plaintiffs:
All persons who purchased a new 1987-1993 Ford F-Series Truck, 1987-1993 Ford Bronco, 1987-1989 Ford Bronco II, 1987-1992 Ford Ranger or 1987-1989 Ford Mustang in Texas on or after March 8, 1988 which was painted with high build eleetrocoat or medium build eleetrocoat and no spray primer and who suffered past and/or future damage as a result of peeling or flaking paint on these vehicles caused by a defective paint process (i.e., high build eleetrocoat or medium built eleetrocoat and no spray primer) excluding persons who purchased vehicles pursuant to a fleet account or fleet identification number; and
All persons who purchased a new 1984r-1988 Ford F-Series Truck, 1984-1988 Ford Bronco, 1984-1988 Ford Bronco II, 1984-1988 Ford Ranger or 1984-1988 Ford Mustang in Texas prior to March 8, 1988 which was painted with high build electro-coat or medium build eleetrocoat and no spray primer and who paid Ford or a Ford dealership for a paint repair to their vehicle to repair peeling or flaking paint caused by a defective paint process (i.e., high build eleetrocoat or medium build eleetrocoat and no spray primer), excluding persons who purchased vehicles pursuant to a fleet account or fleet identification number.
Though finding issues common to the class, the court acknowledged that the multiplicity of potential causes of peeling meant that the case likely would have a phase of individual
trials following the class-wide resolution of the common issues.
The Interlocutory Appeal
Ford contends that the trial court erred by certifying the class action because the proposed trial structure violates Texas’s general strictures against bifurcated trials and because the suit lacks the prerequisites for class certification. We review the trial court’s decision whether to certify a class for an abuse of discretion.
RSR Corp. v. Hayes,
673 S.W.2d 928, 930 (Tex.App.—Dallas 1984, writ dism’d). A trial court does not abuse its discretion by basing its decisions on conflicting evidence, but does abuse its discretion by failing to properly apply the law to undisputed facts.
Forsyth v. Lake LBJ Inv. Corp.,
903 S.W.2d 146, 149 (Tex.App.—Austin 1995, writ dism’d w.o.j.);
RSR,
673 S.W.2d at 930.
1. Bifurcated trial
By points of error four and five, Ford complains that the trial court set up an im-permissibly bifurcated, piecemeal trial process.
Ford contends that Texas courts’ traditional aversion to piecemeal trials overcomes or modifies the class-action rule that, when appropriate, “an action may be brought or maintained as a class action with respect to particular issues.” Tex.R. Civ. P. 42(d)(1). Texas courts’ aversion to bifurcated trials predates the class-action procedural rule. Ford cites cases dating back sixty years illustrating the principle.
See Phoenix Assurance Co. of London v. Stobaugh,
127 Tex. 308, 94 S.W.2d 428, 430 (1936) (insurance coverage case: whether building was total loss not severable issue from whether plaintiff covered by policy);
see also Transportation Ins. Co. v. Morid,
879 S.W.2d 10, 30 (Tex.1994) (exception to non-bifurcation rule: amount of punitive damages tried separately from rest of action);
Otis Elevator Co. v. Bedre,
776 S.W.2d 152, 153 (Tex.1989) (appellate court cannot remand negligence separately from contributory negligence and damages);
Waples-Platter Co. v. Commercial Standard Ins. Co.,
156 Tex. 234, 294 S.W.2d 375, 377 (Tex.1956) (appellate court cannot remand damages only; court cannot require piecemeal trial because liability and damages are elements of indivisible cause of action). These cases were not class actions under the rules of civil procedure, however.
The language of Rule 42(d) indicates that it creates an exception to the general prohibition of bifurcation. Allowing trial of separate “issues” rather than “claims” — a conscious choice, because Rule 42 refers to “claims” elsewhere — indicates that the rulemakers envisioned that something less than an entire cause of action might be tried as a class action. Unlike former Texas Rule of Appellate Procedure 81(b)(1), Rule 42 does not require that the particular class issues be “clearly separable without unfairness to the parties.”
See Otis Elevator,
776 S.W.2d at 153 (interpreting former Tex.R.App. P. 81(b)(1)).
For cases meeting its other requirements, Rule 42 requires only that the class trial of particular issues be “appropriate.” Tex.R. Civ. P. 42(d).
Ford also argues that the term “issues” in Rule 42 should be interpreted like “issue” in Texas Rule of Civil Procedure 174(b). Rule 174(b) provides, “The court in furtherance of convenience or to avoid prejudice may order a separate trial of any claim, cross-claim, counterclaim, or third-party claim, or of any separate issue or of any number of claims, cross-claims, counterclaims, third-party claims, or issues.” The supreme court held that Rule 174(b) did not allow a trial court to try damages separately from liability in a personal injury case.
Iley v. Hughes,
158
Tex. 362, 311 S.W.2d 648, 651 (1958). The jury had found liability and medical expenses, but deadlocked on general and exemplary damages; the trial court granted a mistrial and empaneled a new jury for the damages issues only.
Id.
at 649. The supreme court held that, though the text of the rule would seem to permit such, Texas’s aversion to piecemeal trials required a reading of the rule to prohibit it.
Id.
at 651. The court noted, however, that the rule had been invoked to allow separate trials on issues of standing and limitations.
Id.
at 650. The court wrote that Texas courts dislike piecemeal trials because the public interest, the interests of litigants, and the administration of justice are better served by rules of trial which avoid a multiplicity of suits.
Id.
at 651.
The efficiency interests the
Iley
court cited for unitary trials actually favor the bifurcated trial of the common and individual issues in this case. Though individual trials may be necessary on individual causation and damage issues, the limited class action will allow Ford to consolidate its case regarding the paint process and knowledge about defects into one presentation in the class trial. If the factfinder fails to find that the paint process is defective, Ford (and the court system) will have saved great amounts of time compared to having the estimated 100,-000 individual trials on all the issues in all the claims. Even if the class jury finds the paint process was defective and Ford learned of the defect sometime during the class period, Ford could be exonerated of knowledge of wrongdoing against class members who bought cars before that time. Further, should Ford lose on all issues in the. class action segment of the trial, it and the public will have gained the economies of the unified trial of these common issues.
The question remains whether this separation of liability and damages issues is appropriate and permissible. It does not appear, however, that individual liability and damages trials will be entirely separated. Rather, it appears that the class trial will determine whether the paint process was defective; and, if so, when Ford knew this and if it withheld this knowledge. The individual trials could then focus on other factors that might diminish or eliminate damage awards to individuals, such as individual contributory causation. The trial structure jointly tries common class issues pursuant to Rule 42(d)(1), but, at the individual trial stage, tries each individual’s intertwined liability and damage issues together. We conclude that the trial court did not abuse its discretion in allowing this arrangement. We overrule points four and five.
II. Class-action prerequisites
Ford contends by the remainder of its points of error that the trial court erred by certifying the class because the appellees’ suit lacks the class-action prerequisites. Rule 42 requires that (1) the class be so numerous that joinder of all members is impracticable, (2) questions of law or fact common to the class exist, (3) the claims or defenses of the representative parties be typical of the claims or defenses of the class, and (4) the representative parties must fairly and adequately protect the interests of the class. Rule 42(a). As certified in this ease, the issues common to the class must predominate over individual issues and the class action must be the superior way to handle the suit.
See
Rule 42(b)(4).
A. Commonality
By point of error six, Ford contends the trial court erred by holding that this suit presents questions of law and fact common to the class. This Court has interpreted Rule 42 to require not just common questions but common answers to those questions for all class members.
Wente v. Georgia-Pacific Corp.,
712 S.W.2d 253, 256-57 (Tex.App.—Austin 1986, no writ). The threshold for commonality is not high.
Jenkins v. Raymark Indus., Inc.,
782 F.2d 468, 471 (5th Cir.1986). The commonality test is met when at least one issue’s resolution would affect the claims of all or a significant number of class members.
Forbush v. J.C. Penney Co., 994
F.2d 1101, 1104 (5th Cir.1993).
Ford contends we should follow two cases in which the appellate courts found a lack of commonality defeated certification.
See Wente,
712 S.W.2d at 256-57;
RSR,
673
S.W.2d at 933. In
Wente,
this Court affirmed the trial court’s refusal to certify a class of borrowers in a usury case against a lender; this Court held the refusal was proper because the variety of types of accounts meant that different usury rules would apply, yielding no common answer to whether the lender had treated the purported class members usuriously.
Wente,
712 S.W.2d at 256-57. In
RSR,
the appellate court reversed certification of a class of persons who lived within two miles of a lead smelter; the court found no commonality because of the wide range among purported class members in the level of lead exposure, the duration of lead exposure, the interplay of other possible injurious causes, and the susceptibility to lead poisoning.
RSR,
673 S.W.2d at 933.
Ford contends variations in the paint process used similarly destroy the commonality even within the limited issues certified. The trial court found that the following questions were common to the class: (1) whether the paint process used on the cars is defective because it lacks spray primer; (2) whether Ford knew about the defect and withheld that knowledge; (3) whether Ford had a duty to disclose the defect; and (4) how the discovery rule applied. Ford contends that the answers to these questions are not common among class members because the variables in the painting process — types of body materials, kinds of eleetrocoat, colors, different assembly plants, models, and model years— yield numerous combinations such that there are really thousands of painting processes within the class.
Ford’s paint expert testified that different metals interact with elec-trocoats in ways that substantially affect the likelihood that a particular car’s paint will peel. Both sides’ experts agree that different electroeoats resist peeling differently. Appellees’ expert testified that solid-color cars treated with this process, when painted and baked properly, should not peel. He also testified that ears painted with ultraviolet light absorbers are less likely to peel than identical cars without UV-absorbers.
Appellees’ expert, however, wrote in his report that the common defect in the paint system leading to the peeling problem was the omission of spray primer. A witness for Ford testified that any car that had paint applied to eleetrocoat without intervening spray primer was a potential candidate for peeling. Internal Ford documents indicated that the removal of spray primer was the likely cause of the peeling of F-150s and that use of spray primer eliminated the peeling problem. Ford even instituted a free-repaint program that covered only cars painted without spray primer. Ford also reinstated use of spray primer at all plants producing class cars.
We cannot say that the court abused its discretion by concluding that there was a common answer to whether the absence of spray primer made the paint process defective. The defect inquiry (whether the cars were defective because they lacked something necessary for adequacy)
is common to the class; all ears in the class have peeling paint and the jury’s decision whether non-peeling paint is necessary for adequacy will apply to all of them. Even if other factors contribute to the peeling, there is some evidence on which a jury might find that the lack of spray primer is the common cause. The jury could instead conclude that other factors were key; for instance, if cars painted without spray primer do not peel when UV-absorbers are used, then the jury might conclude the absence of UV-absorbers is the culprit rather than the absence of spray primer. In either case, the record suggests there are questions and answers common to the class.
The remaining stated common issues also have common answers. The interplay of other factors contributing to peeling does not bar the existence of a common answer to the question of knowledge; the jury could eon-
elude the other factors complicated the task of isolating the absence of spray primer as the culprit and postponed the reasonable discovery date for Ford. Even if Ford’s discovery date precludes some members of the class from recovering, the answer to the discovery date question is common to the class. If the discovery date found excludes some members of the class, it will only reduce the number of individual causation and damage trials. The existence of other causes of peeling also has no effect on the answer to whether Ford, once it discovered the defect, withheld or had a duty to disclose its knowledge of the defect.
This case is distinct from
Wente
and
RSR
in critical respects. A significant distinction from
Wente
is that the
Wente
court affirmed a trial court’s
refusal
to certify a class. 712 S.W.2d at 259. The abuse of discretion standard is such that the
Wente
court might have affirmed a certification as well. The class definitions in both cases were much broader than that used here, and in neither case did the plaintiffs attempt to isolate particular issues for class determination. In
Wente,
the class certified included many types of accounts to which different usury rules applied, giving rise to a variety of answers among purported class members that destroyed commonality.
Id.
at 257. The plaintiffs in this case avoided that pitfall by specifying the flaw they believe made the paint on their cars peel. The class definition in
RSR
also was much broader than the one in this case. The proposed class included all landowners within two miles of the smelter without regard to whether lead was on their land; flaws in the definition included the failure to account for variations in exposure, susceptibility, injury, and outside factors. 673 S.W.2d at 933. In this case, cars that did not peel are excluded from the class, there is only one type of damage (peeling), and variations in degree of individual injuries and outside factors are left for individual trial. The court did not abuse its discretion by concluding that there are common questions and answers. We overrule point six.
B. Typicality
By points of error seven and nine, Ford contends the trial court erred by holding that the named plaintiffs’ claims are typical of the class and that the named plaintiffs are adequate representatives of the class. The trial court’s certification of the class action to particular issues limits the breadth of our review of these requirements. The certification of a class only as to particular issues necessarily implies that the named plaintiffs’ claims are not typical for every issue in their causes of action. We will assess the representatives’ typicality and adequacy of representation on the class-certified issues.
Within the common issues, the ap-pellees’ claims are typical of the class. The typicality requirement is satisfied when the evidence shows that the claims or defenses of the class representative have the same essential characteristics as those of the class as a whole; the claims need not be identical or perfectly coextensive, just substantially similar.
Cedar Crest Funeral Home, Inc. v. Lashley,
889 S.W.2d 325, 331 (Tex.App.—Dallas 1993, no writ). The appellees and the unnamed class members’ ears were all painted without spray primer and they all peeled. The named appellees allege that the lack of spray primer caused the peeling, that Ford knew of this link, and that Ford withheld the information; the class members must have similar claims to be in the class and recover for the alleged defect. Though Ford complains that each named appellee and class member will likely have different additional contributing causes for the peeling and different damages, those will be tried after the class stage and do not detract from the ap-pellees’ typicality for class trial purposes. The diversity of individual causation and damages issues shows that diversity itself is typical on those issues, strengthening the case for limitation of the class certification to particular issues. We overrule point seven.
Next we must decide whether the representatives will adequately represent the class. We consider these factors in deciding whether the named plaintiffs adequately represent the class:
the adequacy of counsel, potential conflicts of interest, the personal integrity of the plaintiffs, the representative’s knowledge of the litigation and belief in the legitimacy
of the grievance, whether or not the class is unmanageable because of geographical limits and whether the plaintiffs can afford to finance the class action.
See Salvaggio v. Houston Independent School District,
709 S.W.2d 306, 310 (Tex.App.—Houston [14th Dist.] 1986, writ dism’d). The primary issue is whether any antagonism exists between the interests of the named plaintiffs and those of the remainder of the class. Only a conflict that goes to the very subject matter of the litigation will defeat a party’s claim of representative status.
Dresser Indus., Inc. v. Snell,
847 S.W.2d 367, 373 (Tex.App.—El Paso 1993, no writ). The trial court found the adequacy of representation was not in substantial dispute. The court noted that the representatives were knowledgeable, informed, participating, cooperative, and showed no interests antagonistic to unnamed class members. Ford has not told us why it believes the representation will be inadequate by these measures and nothing in the record persuades us these findings are unfounded. We overrule point nine.
C. Maintainability of class action
Ford complains by two points of error that the class action is not maintainable. By point eight, Ford complains that common issues do not predominate over individual issues. By point ten, Ford contends that the class action is not superior to other trial methods.
Ford contends the trial court erred by refusing to establish the trial structure and schedule upon certification. The court certified the class understanding that thousands of individual trials might be necessary on liability and damage issues. The results of the class trial could drastically affect the number of claims that survive to the individual stage and could reveal viable subclasses. The court’s decision to acknowledge the need for an individual phase but to wait until that picture is clearer to make specific plans does not constitute an abuse of discretion. We are reviewing the certification, including the predominance and superiority requirements, understanding that there may be 100,000 class members.
The test for predominance is not whether the common issues outnumber the individual issues, but whether common or individual issues will be the object of most of the efforts of the litigants and the court.
Angeles/Quinoco Sec. Corp. v. Collison,
841 S.W.2d 511, 515 (Tex.App.—Houston [14th Dist.] 1992, no writ). Differences in damage recoveries by class plaintiffs or in defenses unique to some class plaintiffs do not destroy the class.
Id.
at 516. The common issues in this case overshadow all individual issues; if the class jury finds the absence of spray primer is not a defect, or that Ford did not know or withhold the information, the trial ends without reaching the individual issues. Even if the class jury finds that Ford knew the paint process was defective, the finding of the date Ford acquired knowledge could extinguish the claims of thousands of class members. “In cases where it appears that the common issues may predominate over the individual issues, the most efficient approach for the trial court is to allow class certification at the present time subject to a motion by the defendants after the case has developed to dissolve the class on the grounds that the common questions are not predominant at trial.”
Life Ins. Co. of Southwest v. Brister,
722 S.W.2d 764, 775 (Tex.App.—Fort Worth 1986, no writ). We conclude the trial court did not abuse its discretion in finding at this stage that class issues predominate. We overrule point eight.
Ford argues that the need for individual trials destroys the normal economies, and thus the superiority, of a class action. Ford contends 100,000 individual trials lasting one hour each would occupy the trial court eight hours a day, 365 days a year for more than thirty-four years. Ford’s formula magnifies the importance of trying all common issues together. If the common issues instead were tried in each of the individual trials, the individual trials would each be that much longer. If the individual trial of the common issues added just one hour to the individual trials (and the appellate record indicates one hour is a conservative estimate), the trial court would be dedicated daily to these trials for sixty-eight years. The trial court’s plan, while potentially
lengthy, is shorter and by this measure is superior to the individual trial of all issues.
The class action is also superior because it resolves all claims. As the calculations in the previous paragraph show, individual trials of all issues save time compared to the class only if individual litigants are dissuaded from bringing their claims as individuals. Class actions can afford individuals the chance to seek recovery of damages for which an individual suit might otherwise be economically unfeasible; for that reason, the class action is superior because it permits claimants to obtain a resolution on the merits by means of an adversarial hearing despite the imbalance in the parties’ economic positions. Ford gains the advantage of facing and possibly defeating all claimants in one action. Appel-lees bear the burden of informing the whole class of plaintiffs and face the disadvantage of having only one opportunity for successful resolution. That appellees insist, in view of these obstacles, on the certification of a class further supports the trial court’s decision.
See Jenkins v. Raymark Indus., Inc.,
782 F.2d 468, 474 (5th Cir.1986). We cannot say that the court’s plan is an abuse of discretion. We overrule point ten.
D. Structure of class definition
Ford complains by three points of error about the structure of the class definition. By points of error one and two, Ford complains that class members are not ascertainable without a finding on the merits regarding the cause of the paint peeling or flaking. Ford argues that appellees therefore currently cannot meet the numerosity requirement because the membership in the class is zero. Ford also contends that the causation requirement impermissibly creates a fail-safe class because the class is either full of people to whom Ford is liable or empty; Ford complains that, if the jury fails to find that the paint process is defective, the judgment in Ford’s favor will bind no one because it will bind only the empty class.
The Supreme Court, considering a different procedural issue, interpreted Rule 42’s federal antecedent and held that the class certification procedure did not authorize a preliminary inquiry into the merits to determine whether the suit could be maintained as a class action.
Eisen v. Carlisle & Jacquelin,
417 U.S. 156, 177-78, 94 S.Ct. 2140, 2152-53, 40 L.Ed.2d 732 (1974). The trial court had held a preliminary inquiry into the merits (as in an injunctive relief proceeding) rather than a trial. The Supreme Court held that examining the merits to determine class certification conferred the benefits of class certification on one who had not shown entitlement.
Id.
at 177, 94 S.Ct. at 2152. The court was also concerned that the determination in a preliminary class inquiry would occur without the traditional safeguards of trial and could prejudice the defendants at actual trial.
Id.
at 178, 94 S.Ct. at 2152-53.
Though
Eisen
is proeedurally distinct from this case, its reasoning is persuasive. The concern over losing the safeguards of trial does not apply here, but the concern of granting class status to persons who have not demonstrated entitlement to the class action resonates even more strongly here where full trial (rather than a preliminary hearing) of some issues would be held before class membership could be ascertained.
See also Armstrong v. Chicago Park Dist.,
117 F.R.D. 623 (N.D.Ill.1987);
Hagen v. City of Winnemucca,
108 F.R.D. 61 (D.Nev.1985). The
Armstrong
and
Hagen
trial courts both rejected attempts to define classes of persons discriminated against by city actions or regulations.
Armstrong,
117 F.R.D. at 626-27 (proposed class of persons denied promotion by discrimination self-defeating because of “prohibition” on inquiring into merits at certification stage);
Hagen,
108 F.R.D. at 63 (proposed class of persons denied constitutional rights by prostitution policy wrongfully requires merits determination for certification). The
Hagen
trial court reformed the proposed class definition to include all persons permitted to work as prostitutes.
Id.
at 64.
We conclude that the class definitions in the trial court’s order violate Rule 42 by allowing the named plaintiffs to proceed in a class action before showing that a class exists. Though seeking to limit the class was a valid goal, certifying a memberless class was an abuse of discretion.
We need not reverse the certification entirely, however. We address Ford’s objection to the liability clause in the definitions by inserting the phrase “who allege the peeling or flaking was” before the merits-determination clause “caused by a defective paint process.” The class of people who purchased the specified peeling cars and who
allege
that the lack of spray primer caused the peeling or flaking is ascertainable without a decision on the merits. Notice can be sent to all persons who purchased the listed cars. Whether the paint on a particular car peeled is susceptible of objective proof and within the buyer’s knowledge. Those persons can choose whether to opt out of the class. Whether the buyer makes the allegation is not an inquiry into a state of mind, but objectively verifiable and within the buyer’s knowledge. This class is not empty — Ford estimated some 100,000 individuals in its brief; whatever the minimum number required, 100,000 clearly meets the numerosity requirement. We therefore sustain in part and overrule in part points of error one and two. Because the membership of this reformed class is not dependent on a merits determination, the reformed class is not a fail-safe class, and we overrule point three as moot.
CONCLUSION
We modify the class certification order and revise the class definitions in accordance with our opinion as follows:
All persons who purchased a new 1987-1993 Ford F-Series Truck, 1987-1993 Ford Bronco, 1987-1989 Ford Bronco II, 1987-1992 Ford Ranger or 1987-1989 Ford Mustang in Texas on or after March 8, 1988 which was painted with high build eleetrocoat or medium build eleetrocoat and no spray primer and who suffered past and/or future damage as a result of peeling or flaking paint on these vehicles
who allege the peeling or flaking was
caused by a defective paint process (i.e., high build eleetrocoat or medium built eleetrocoat and no spray primer) excluding persons who purchased vehicles pursuant to a fleet account or fleet identification number; and All persons who purchased a new 1984-1988 Ford F-Series Truck, 1984-1988 Ford Bronco, 1984-1988 Ford Bronco II, 1984-1988 Ford Ranger or 1984-1988 Ford Mustang in Texas prior to March 8, 1988 which was painted with high build electro-coat or medium build eleetrocoat and no spray primer and who paid Ford or a Ford dealership for a paint repair to.their vehicle to repair peeling or flaking paint
who allege the peeling or flaking was
caused by a defective paint process (i.e., high build eleetrocoat or medium build eleetrocoat and no spray primer), excluding persons who purchased vehicles pursuant to a fleet account or fleet identification number.
(Emphasis added in accordance with this opinion.) These modifications do not substantially alter the class membership; rather, they serve to make class membership ascertainable at the outset of the suit. Because the class membership is essentially that on which the trial court made its findings and conclusions, we need not remand the case for new findings or reexamine the existing findings. These modifications do not limit the trial court’s ability and responsibility under Rule 42(c)(1) to alter, amend, or withdraw the class certification order as the case progresses.
With the stated modifications, we affirm the certification order.