Anderson Contracting, Inc. Vs. Dsm Copolymers, Inc.

• Court: Supreme Court of Iowa
• Decided: December 18, 2009
• Docket: 07–0722
• Status: Published

Opinion

IN THE SUPREME COURT OF IOWA

No. 07–0722

Filed December 18, 2009

ANDERSON CONTRACTING, INC.,

Appellee,

vs.

DSM COPOLYMERS, INC.,

Appellant,

BAYER AG; BAYER POLYMERS, L.L.C. n/k/a BAYER MATERIALSCIENCE, L.L.C.; BAYER CORPORATION; CROMPTON CORPORATION, UNIROYAL CHEMICAL CORPORATION, INC., n/k/a CROMPTON MANUFACTURING COMPANY, INC., THE DOW CHEMICAL COMPANY; E.I. DUPONT DE NEMOURS & COMPANY; DUPONT DOW ELASTOMERS, L.L.C.; DSM ELASTOMERS EUROPE B.V.; and EXXON MOBIL CHEMICAL CORPORATION, d/b/a EXXON MOBIL, INC.

Defendants.

Appeal from the Iowa District Court for Polk County, Richard G.

Blane II, Judge.

Defendant manufacturer of synthetic rubber appeals from the

district court’s certification of a class action. AFFIRMED.

Edward W. Remsburg of Ahlers & Cooney, P.C., Des Moines,

Andrew S. Marovitz and Britt M. Miller of Mayer Brown LLP, Chicago, IL,

and Richard J. Favretto, Gary A. Winters, and Andrew A. Nicely of Mayer

Brown LLP, Washington, D.C., for appellant. 2

Joseph R. Gunderson and Jason D. Walke of Gunderson, Sharp &

Walke, LLP, Des Moines, Rex A. Sharp of Gunderson, Sharp & Walke,

LLP, Prairie Village, KS, and Isaac L. Diel of Sharp McQueen, PA,

Overland Park, KS, for appellee. 3

HECHT, Justice.

A manufacturer appeals from the district court’s order certifying a

class in an action alleging price fixing of the market for ethylene

propylene diene monomer (EPDM). We affirm.

I. Factual and Procedural Background.

Anderson Contracting, Inc. (Anderson), an Iowa corporation that

performs roofing work, brought suit against various manufacturers,

marketers, and distributors of EPDM 1 (EPDM manufacturers) for

violations of the Iowa Competition Law.

EPDM is a synthetic rubber composed of ethylene, propylene, and

diene monomers. EPDM is produced in various grades which exhibit

different properties and is then used to make various products. It is

most heavily used in the automobile industry to make weatherstripping,

seals, belts, hoses, and tires. It is also used in roofing compounds,

electrical insulation, garden hoses, golf club grips, and in gaskets and

seals for many household appliances.

Anderson brought suit alleging the EPDM manufacturers

conspired to restrain trade and fix the price of EPDM in violation of the

Iowa antitrust laws. Anderson claims it purchased various items containing EPDM for a higher price than it would have had the

conspiracy not existed and seeks to represent all end purchasers of

products containing EPDM in the state of Iowa.

Anderson moved for class certification in June 2006. A contested

hearing was held on December 1, 2006, and the district court granted

1The only defendant participating in the appeal is DSM Copolymers, Inc. The

other defendants have settled, including Bayer AG; Bayer Polymers, L.L.C., n/k/a Bayer MaterialScience, L.L.C.; Bayer Corporation; Crompton Corporation; Uniroyal Chemical Corporation, Inc. n/k/a Crompton Manufacturing Company, Inc.; The Dow Chemical Company; E.I. Dupont de Nemours & Company; Dupont Dow Elastomers, L.L.C.; DSM Elastomers Europe B.V.; and Exxon Mobil Chemical Company, Inc. 4

class certification on March 16, 2007. The district court certified the

class to include “all persons who indirectly purchased Defendants’ EPDM

in the State of Iowa, other than for resale, from January 1994 through

December 2002.”

The EPDM manufacturers appealed, contending the district court

abused its discretion in certifying the class.

II. Scope of Review.

We review a district court’s decision to grant a request to certify a

class action for an abuse of discretion. Luttenegger v. Conseco Fin.

Servicing Corp., 671 N.W.2d 425, 436 (Iowa 2003). “Our class-action

rules are remedial in nature and should be liberally construed to favor

the maintenance of class actions.” Comes v. Microsoft Corp., 696 N.W.2d

318, 320 (Iowa 2005) (Comes II). When a district court’s grounds for

certification are clearly unreasonable, an abuse of discretion can be

found. Varner v. Schwan’s Sales Enters., Inc., 433 N.W.2d 304, 305

(Iowa 1988). However, if the district court “weigh[ed] and consider[ed]

the factors and [came] to a reasoned conclusion as to whether a class

action should be permitted for a fair adjudication of the controversy,” we

will affirm. Luttenegger, 671 N.W.2d at 437; accord Comes II, 696 N.W.2d

at 321.

III. Discussion.

When determining whether to certify a class action, a district court

is guided by Iowa Rules of Civil Procedure 1.261–1.263. “[A]s soon as

practicable after the commencement of a class action the court shall hold

a hearing” and determine whether the action should proceed as a class

action. Iowa R. Civ. P. 1.262(1). The court may certify a class if it finds

three requirements are established: (1) the requirements of rule 1.261

are met, (2) a class action would provide for the fair and efficient 5

adjudication of the case, and (3) the representative parties will protect

the interests of the class. Iowa R. Civ. P. 1.262(2). The requirements of

rule 1.261 are established if the class is either so numerous or

constituted in such a way that joinder is impracticable and there is a

question of law or fact common to the class. Iowa R. Civ. P. 1.261. To

determine whether a class action will provide a fair and efficient

adjudication of the case, rule 1.263 provides “the court shall consider

and give appropriate weight to [thirteen listed factors] and other relevant

factors.” Iowa R. Civ. P. 1.263(1).

We have recognized that the language of rule 1.263 indicates the

district court has “considerable discretion” in weighing the factors.

Vignaroli v. Blue Cross of Iowa, 360 N.W.2d 741, 744 (Iowa 1985). The

court will determine what weight, if any, to give to each of the listed

factors. Vos v. Farm Bureau Life Ins. Co., 667 N.W.2d 36, 45 (Iowa 2003);

Martin v. Amana Refrigeration, Inc., 435 N.W.2d 364, 369 (Iowa 1989).

“Whether or not we agree with the decision arrived at by the trial court is

not the issue. The issue is one of abuse of discretion.” Martin, 435

N.W.2d at 369.

The district court issued a twenty-two page ruling examining each

requirement for class certification, as well as each of the thirteen factors

relevant to the determination of whether a class action is a fair and

efficient method of litigation in this case. The district court described its

decision to certify the class as a “close call” and acknowledged several

concerns. When considering rule 1.263(1)(e), the court noted the

potential difficulties confronting indirect purchasers when proving injury

and damages, but ultimately concluded common questions predominate

over individual ones and weigh in favor of certifying the class. When

considering rule 1.263(1)(k), the court acknowledged the broad definition 6

of the class coupled with the potential difficulty of identifying specific

products containing the defendants’ EPDM posed significant

manageability problems which could prove insurmountable. Although

the court did determine this factor weighed against certification, the

court concluded that the requirements of rule 1.262 were met and

certified the class. The district court noted it has the authority to amend

the certification order at a later time or even to decertify the class if the

circumstances later render such action appropriate. See Iowa R. Civ. P.

1.265; Vos, 667 N.W.2d at 46.

Several of the EPDM manufacturers appealed, 2 contending the

district court abused its discretion in (1) certifying the class action

despite its recognition of the potential manageability problems and (2)

concluding common issues predominate over individual issues.

A. Manageability. The EPDM manufacturers allege the district

court correctly determined the manageability factor weighs against class

certification as a fair and efficient means to litigate the case, but argue

the court abused its discretion by certifying the class. The

manufacturers argue that because EPDM has a similar appearance to

natural rubber, plastic, and vinyl, and because of the wide range of products that use EPDM and these other substances, it will be difficult, if

not impossible, for potential class members to establish they are

members of the class. Further, the manufacturers contend, even if

potential class members can determine they purchased a product

containing EPDM during the relevant time period, it will be even more

difficult to determine if the EPDM was manufactured by one of the

defendants. They also contend the definition of the class is ambiguous

2Allof the appellants have since withdrawn their appeal with the exception of DSM Copolymers, Inc. 7

and problematic because the limitation to people who have purchased

EPDM “other than for resale” is confusing and unclear. The parties agree

that because of the prevalence of EPDM products, the class could

potentially include every resident of Iowa during the established time

frame. Thus, the EPDM manufacturers contend, the identification of

class members will require hundreds of thousands of “mini-trials” for

each putative class member to establish his or her membership in the

class.

The distribution channels of EPDM are complicated and extensive

due to the nature of the substance. EPDM is extremely versatile.

Because it is manufactured in various grades with different qualities, its

uses vary widely, and it tends to be combined with other components to

create other products. Often, these products are sold and, in turn,

combined with or implemented into other products and again resold.

(For example, consider the case of EPDM that is sold to a purchaser who

combines it with other products to create a rubber hose which is then

sold to a car manufacturer to be used in the assembly of motor vehicles.)

The end product will not have “EPDM” or the original manufacturer’s

identity stamped on it, and indeed the EPDM-containing component part

itself may be well-concealed within the final product (an appliance or

automobile). Thus, the EPDM manufacturers contend identifying the

members of the class will require a mini-trial for each potential class

member to establish that he has indeed purchased not only a product

containing EPDM and not another substance, but has purchased a

product containing the defendants’ EPDM.

Anderson does not dispute that the distribution channels are

complicated and widely varied. It contends, however, that because it will

prove class-wide injury and damages in the aggregate during trial, there 8

will be no need for mini-trials establishing that each individual class

member purchased an EPDM product for an inflated price. Assuming,

only for argument’s sake, that Anderson is successful at trial and proves

one or more of the defendants violated the Iowa Competition Law, Iowa

Code chapter 553, any potential mini-trials establishing class

membership and entitlement to damages will occur during the claims

administration process.

Anderson seeks to utilize a “top down” approach in proving class-

wide injury and asserts damages should be assessed in the aggregate as

established through expert testimony. See Comes II, 696 N.W.2d at 323–

25. Under this approach advocated by Anderson, there would be no

need during the trial to address the potential manageability problems

described by the EPDM manufacturers and noted by the district court in

its ruling. Instead, such potential problems would be confronted, if

necessary, after the trial of the liability and class-wide injury issues is

completed. Further, Anderson contends the manageability problems

asserted by the manufacturers could be avoided altogether if any

judgment for class-wide injury is distributed cy pres. See 2 William B.

Rubenstein, Alba Conte & Herbert B. Newberg, Newberg on Class Actions

§ 4:33 (4th ed. 2002) (explaining cy pres distribution of settlement

proceeds in antitrust class action litigation).

The EPDM manufacturers cite In re Phenylpropanolamine Products

Liability Litigation (PPA) to support their argument that the district court

should not have certified the class because the class members must

establish individual injury and damages. 214 F.R.D. 614, 619–20 (W.D.

Wash. 2003). In PPA, the federal district court’s order declined a request

for certification of a class because the court concluded the class

members would be virtually impossible to identify and that even allowing 9

a “fluid recovery” or cy pres procedure would not alleviate the

identification problems. 214 F.R.D. at 618–20. We do not find the

court’s decision in PPA persuasive in this case. First, although the court

in PPA determined a “fluid recovery” procedure was unsuitable, this

court has already recognized an aggregate approach to injury and

damages as appropriate in an antitrust case. See Comes II, 696 N.W.2d

at 323–24. Further, PPA is a case in which a federal district court

concluded, in the exercise of its discretion, that a class should not be

certified. PPA, 214 F.R.D. at 614, 623. In contrast, this court is

reviewing for abuse of discretion the district court’s determination that a

class should be certified.

The EPDM manufacturers next assert the exclusion from the class

of indirect purchasers who bought EPDM-containing products “other

than for resale” is vague and confusing. Specifically, they argue it is

unclear whether purchasers who intended to resell the product when

they purchased it, those who did not intend to resell but did ultimately

resell, and those who intended to resell but were unable to resell are all

excluded from the certified class. We conclude the district court’s

delineation of the class clearly is intended to exclude persons who resold

the EPDM or product containing the substance, no matter what their

intent was at the time of purchase. The definition of the class makes no

mention of the purchasers’ intent, and we see no reason the purchasers’

intent informs a determination of whether a purchaser has been harmed

by the alleged conduct of the defendant manufacturers.

In conclusion, we note the district court did conclude the potential

manageability issues weighed against certification of the class. However,

manageability is but one of thirteen factors the court considered when it

determined a class action is a fair and efficient method of litigating the 10

case. As we have already noted, rule 1.263 does not require any

particular factor be weighed more heavily than another. In fact, the rule

gives ample discretion to the district court to weigh the factors as it sees

fit. We also observe that a number of courts have concluded

manageability issues alone are rarely sufficient to refuse certification.

See In re Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124, 140–41

(2d Cir. 2001); In re Workers’ Compensation, 130 F.R.D. 99, 110 (D.

Minn. 1990). Further, the district court emphasized that if necessary, it

could modify the certification order or decertify the class altogether at a

later time. Considering these possible remedies should the class become

unmanageable, and given our belief that any need for individualized

determinations will arise, if at all, during the claims administration

process after a trial of the liability and class-wide injury issues, we

conclude the district court did not abuse its discretion.

B. Common Issues vs. Individual Issues. The EPDM

manufacturers also take issue with the district court’s application of rule

1.263(1)(e)—“[w]hether common questions of law or fact predominate

over any questions affecting only individual members.” Iowa R. Civ. P.

1.263(1)(e). The EPDM manufacturers and Anderson agree that the

claim against the manufacturers involves three elements: (1) proof of a

conspiracy to fix the price of EPDM, (2) injury to the plaintiffs, and (3)

damages. The parties further agree that the first of these elements can

be established with common proof and the third element will require

some individualized proof. However, they dispute whether the second

element may be established with common proof. Both parties offered

expert opinions supporting their positions. The district court considered

both expert opinions and noted the fighting issue between them was

whether a method of establishing class-wide injury could be devised. 11

The court concluded this issue went to the merits of the case and was “a

factual issue for the jury to determine based on expert testimony.” It

concluded that if Anderson’s expert is ultimately unable to provide a

method of calculating the alleged conspiracy’s effect on pricing, a motion

for summary judgment or directed verdict would appropriately address

the issue. Concluding the difficulties of proving injury and damages in a

class action brought by indirect purchasers are very challenging but not

insurmountable, the court determined that common issues predominate

over individual issues in this case and rule 1.263(1)(e) therefore weighs

in favor of certification of the class.

The EPDM manufacturers contend the district court’s assessment

of this factor was flawed in two respects. First, the manufacturers argue

the court should not have applied the low standard articulated in Comes

II for the evaluation of expert testimony at the class certification stage.

Second, they contend the district court abused its discretion by

concluding common issues predominated and should have refused to

certify the class had it determined otherwise.

The district court acknowledged that the EPDM manufacturers had

offered an expert opinion contradicting Anderson’s expert’s claimed

ability to assess injury on a class-wide basis. The court nonetheless

concluded it is inappropriate, during class certification proceedings, to

resolve “battles between the experts.” Citing Comes II, the district court

said “[a]t this point the Court is only concerned with ensuring that the

basis of the expert opinion is not so flawed that it would be inadmissible

as a matter of law.” The EPDM manufacturers contend the district

court’s application of the “not inadmissible” standard was erroneous

because the case Comes II relied on for the standard, Visa Check, 280 12

F.3d at 135, has since been disavowed. See In re Initial Pub. Offerings

Secs. Litig., 471 F.3d 24, 42 (2d Cir. 2006) (IPO).

After reviewing Supreme Court authority, as well as decisions from

other federal circuits, the Second Circuit Court of Appeals did disavow

the “not inadmissible” standard and joined a clear majority of

jurisdictions applying a somewhat more searching standard in the

determination of whether a class should be certified. The court

concluded [a] district judge is to assess all of the relevant evidence admitted at the class certification stage and determine whether each [class certification] requirement has been met, just as the judge would resolve a dispute about any other threshold prerequisite for continuing a lawsuit. Id.; see also Blades v. Monsanto Co., 400 F.3d 562, 575 (8th Cir. 2005)

(stating that “in ruling on class certification, a court may be required to

resolve disputes concerning the factual setting of the case” and resolve

“expert disputes concerning the import of evidence”); Unger v. Amedisys,

Inc., 401 F.3d 316, 319 (5th Cir. 2005) (requiring a careful certification

inquiry including findings); Szabo v. Bridgeport Machs., Inc., 249 F.3d

672, 676 (7th Cir. 2001) (requiring a judge to make whatever legal and

factual inquiries are necessary to determine if class certification is appropriate); Love v. Turlington, 733 F.2d 1562, 1564 (11th Cir. 1984)

(stating that while the court may not reach the merits of a claim, it also

should not artificially limit its review of the class certification

requirements in deference to that principle).

The EPDM manufacturers contend we should adopt the more

searching standard now applied by the Second Circuit and a majority of

jurisdictions. The manufacturers assert that if such a standard is

applied in this case, Anderson’s expert’s claimed ability to devise a 13

workable formula to assess damages on a class-wide basis will not

survive scrutiny.

Although the Second Circuit Court of Appeals’ decision in IPO was

partially based on amendments to the Federal Rules of Civil Procedure

which have not been adopted in Iowa, 3 we find the reasoning of IPO’s

rejection of the “not inadmissible” standard persuasive and adopt it. We,

however, do not find our disavowal of the “not inadmissible” standard

requires us to reverse the decision of the district court in this case.

First, we do not find the standard articulated in IPO to be radically

different from the standards of evaluating a motion for class certification

this court has articulated in the past. Although we have cautioned that

a certification hearing should not involve a determination of whether the

plaintiffs will prevail on the merits, “that is not to say that the court may

not require sufficient information to form a reasonable judgment in

deciding whether to certify a class.” Martin, 435 N.W.2d at 367–68.

“[T]he question of predominance necessitates a ‘close look’ at ‘the

difficulties likely to be encountered in the management of a class

action.’ ” Vos, 667 N.W.2d at 46 (quoting Rothwell v. Chubb Life Ins. Co.

of Am., 191 F.R.D. 25, 28–29 (D.N.H. 1998)). While the decision in IPO

requires a more searching analysis than earlier Second Circuit

precedent, the Second Circuit Court of Appeals also warned courts

3The Second Circuit Court of Appeals concluded that [t]wo changes arguably combine to permit a more extensive inquiry into whether [class certification] requirements are met than was previously appropriate. First, the amended rule removes . . . the provision that class certification “may be conditional.” Second, the amended rule replaces the provision . . . that a class certification decision be made “as soon as practicable” with a provision requiring the decision “at an early practicable time.” IPO, 471 F.3d at 39. 14

against engaging in such an extensive analysis of an expert’s credibility

that it must make a decision on the merits of the case. 471 F.3d at 41.

Additionally, we conclude the opinion of Anderson’s expert survives

this more searching scrutiny. Anderson’s expert, Dr. Conner, concluded

that based on his extensive experience and studies, 4 “all class members

were similarly affected by paying a higher price for the defendants’ EPDM

during the class period than they otherwise would have paid in the

absence of the defendants’ anticompetitive conduct.” He additionally

asserted that “there is a reasonable method sanctioned by orthodox

economic principles that will permit the computation of class-wide

damages using a common formula.” In his affidavit, Dr. Conner

described four different methods of calculating the class-wide damages

commonly accepted in federal and state courts, which he has used in

other class action litigation. In his deposition, Dr. Conner conceded he

could not tell at this early juncture which of the four methods would

prove to be the most effective and reliable “[b]ecause that would require

actual immersion into the data.” He explained that while he had not

attempted to include calculations of the manufacturers’ overcharge in his

4Dr. Conner has been a professor in the Department of Agricultural Economics at Purdue University since 1989 where he teaches price analysis, industrial- organization economics, and quantitative research methods primarily to graduate students. He earned his Ph.D. and M.S. in Agricultural Economics from the University of Wisconsin at Madison. His research has specialized in industrial-organization economics, and in the last ten years, more specifically in cartel studies and antitrust enforcement. He has published or is preparing to publish more than fifty academic publications analyzing various facets of the economics of price-fixing or antitrust enforcement. This research has been stimulated by his involvement as an expert in various class action lawsuits. He has most recently submitted expert reports in cases alleging price-fixing conspiracies in the marketing of lysine, methionine, smokeless tobacco, fed cattle, district heating pipes, and grocery wholesaling. Dr. Conner has served as a consultant to the U.S. Subcommittee on Multinational Corporations, the U.S. Congress’ Office of Technology Assessment, the Antitrust Division of the U.S. Department of Justice, the National Association of Attorneys General, the Organization of Economic Cooperation and Development, and the United Nations. 15

report because of his limited access to industry pricing records, he had

prepared a “back-of-the-envelope” estimate of the percentage price

change during the collusive period and estimated that the direct

overcharge amount extracted by the defendants from the market was

about nine percent.

The EPDM manufacturers contend Dr. Conner wavered in his

assertion that he could calculate class-wide injury.

Q: Have you concluded that it would be possible to determine the effect of injury in this case on a class-wide basis? A: I’ve not yet determined that. I see no impediments to doing such an analysis with further discovery in the future and information from end users, from retailers. I have no reason to suspect that it’s not feasible to form such an analysis. But I don’t have—I have not yet been provided with—with prices, for example, at lower levels of the EPDM channels that would allow me to make a preliminary conclusion in the matter. Q: Have you formed a conclusion about whether it would be possible to prove damages to the [putative] class in this case on a class-wide basis? A: I am confident that one or more of the methods that I outlined in this affidavit will permit me or some other well-trained analyst to do so. We do not find Dr. Conner’s testimony as faltering as the

manufacturers would characterize it. Rather we read any hesitation of

Dr. Conner to be a reluctance to identify the most appropriate method of

calculating the indirect overcharge until he has access to more complete

records following thorough discovery. As we noted before, when pressed,

Dr. Conner offered a preliminary rough estimate of the direct overcharge,

but consistently declined to estimate the indirect overcharge because he

did not yet have enough information to calculate it.

The defendants also assert the district court should have

considered the opinion of their expert who contradicted Dr. Conner’s

assertion that it would be possible to calculate damages on a class-wide

basis. The EPDM manufacturers’ expert, Dr. Snyder, in a well-written 16

and persuasive report, criticized Dr. Conner’s conclusions on two key

issues. First, Dr. Snyder challenged Dr. Conner’s lack of knowledge of

the EPDM industry and channels of distribution. Second, Dr. Snyder

claimed Dr. Conner’s methods are simplistic and insufficient to calculate

class-wide damages in an industry as complex and wide-ranging as the

EPDM industry. We conclude the first of these issues would be

appropriately considered by the court when making a class certification

decision, but the second goes to the heart of the merits of the case, and

as such, should be deferred by the trial court, even under the standard

articulated in IPO. While a court should consider all of the relevant

evidence admitted at the class certification stage and resolve any factual

disputes necessary to determine if the class certification requirements

are met, the court “should not assess any aspect of the merits unrelated

to a [class certification] requirement” and has the discretion to limit

discovery and the extent of a hearing “to assure that a class certification

motion does not become a pretext for a partial trial of the merits.” Id.

Considering all the evidence admitted in the class certification

proceedings, including Dr. Snyder’s report, we conclude the trial court

did not abuse its discretion in concluding Anderson had submitted

sufficient evidence tending to demonstrate that class-wide injury can be

quantified in this case. See Comes II, 696 N.W.2d at 322–23. Given that

the class certification decision must be made “as soon as practicable

after the commencement of a class action,” rule 1.262(1), we would

expect a district court to consider a plaintiff’s expert’s limited access to

discovery (as well as the defendants’ expert’s superior access to the

defendant’s records) when assessing the experts’ opinions in the early

stages of complex litigation. Dr. Snyder’s contention that Dr. Conner’s

methods are flawed and incapable of calculating injury and damages to 17

the class as a whole constitutes a challenge going directly to the merits of

the case and should not be resolved at this preliminary stage. We note,

as did the trial court, that “a safety net is provided for cases in which

certification is improvidently granted: the court may decertify the class at

a later time.” Comes II, 696 N.W.2d at 324. The trial court may also

modify the certification order by narrowing the class or establishing

subclasses. 5 Iowa R. Civ. P. 1.265.

IV. Conclusion.

We find no abuse of discretion in the district court’s decision to

certify the class action lawsuit.

AFFIRMED.

All justices concur except Cady, J., who takes no part.

5Dr. Conner himself suggested, indirectly, that subclasses may be appropriate

for this class, [b]ecause the pass-on rates may vary according to which channel one is studying. The task of an analyst faced with this problem of determining damages may—it may turn out that the subclasses make more sense from the point of view of economic analysis than developing a model for the entire class. He, however, asserted that he would need “to gather the appropriate data and do the appropriate analysis in order to determine the pass-on rate in the channels as a whole or in individual channels involving EPDM.”