In the
United States Court of Appeals
For the Seventh Circuit
____________________
Nos. 11‐8029, 12‐8030
LARRY BUTLER, et al., individually and on behalf of all others
similarly situated,
Plaintiffs‐Appellants, Cross‐Appellees,
v.
SEARS, ROEBUCK AND CO.,
Defendant‐Appellee, Cross‐Appellant.
____________________
Appeals from the United States District Court for the
Northern District of Illinois, Eastern Division.
Nos. 06 C 07023, 07 C 00412, 08 C 01832 —
Sharon Johnson Coleman, Judge.
____________________
On Remand from the Supreme Court of the
United States, No. 12‐1067
__________________________
SUBMITTED JULY 8, 2013 — DECIDED AUGUST 22, 2013
____________________
Before POSNER, RIPPLE, AND HAMILTON, Circuit Judges.
POSNER, Circuit Judge. The Supreme Court has vacated
our judgment in this class action suit (reported at 702 F.3d
2 Nos. 11‐8029, 12‐8030
359 (7th Cir. 2012)) and remanded the case to us for recon‐
sideration in light of Comcast Corp. v. Behrend, 133 S. Ct. 1426
(2013). Sears, Roebuck & Co. v. Butler, 133 S. Ct. 2768 (2013)
(mem.).
This suit, a diversity suit based on the breach‐of‐
warranty laws of six states, is really two class actions be‐
cause the classes have different members and different
claims, though both arise from alleged defects in Kenmore‐
brand Sears washing machines sold in overlapping periods
beginning in 2001 and 2004. One class action complains of a
defect that causes mold (the “mold claim”), the other of a de‐
fect that stops the machine inopportunely (the “control‐unit
claim”). The district court denied certification of the class
complaining about the defect that causes mold and granted
certification of the class complaining about the defect that
causes the sudden stoppage. The plaintiffs asked us to re‐
verse the denial, and we did so; Sears asked us to reverse the
grant, and we refused.
Sears asks us to remand the case to the district court for a
fresh ruling on certification in light of Comcast, or alterna‐
tively to deny certification in both class actions. The plain‐
tiffs ask us to reinstate our judgment, granting certification
in both.
Sears’ request for a remand to the district court is based
to a significant degree on new evidence that has come to
light since the district court ruled on certification in Septem‐
ber 2011. But the case remains pending in the district court,
and, as Sears itself emphasizes, rulings on certification in
class action suits are tentative and can be revisited by the
district court as changed circumstances require. Fed. R. Civ.
P. 23(c)(1)(C); Advisory Committee Notes to 1966 Amend‐
Nos. 11‐8029, 12‐8030 3
ment of Rule 23(c)(1); Amgen Inc. v. Connecticut Retirement
Plans & Trust Funds, 133 S. Ct. 1184, 1202 n. 9 (2013); Johnson
v. Meriter Health Services Employee Retirement Plan, 702 F.3d
364, 370 (7th Cir. 2012). What could it mean to remand a case
to a court before which the case is pending?
The question presented by the Supreme Court’s remand
is one of law—whether the Comcast decision cut the ground
out from under our decision ordering that the two classes be
certified. There is no point in delaying our decision on re‐
mand to await consideration by the district court of factual
issues that may be moot on the basis of the Comcast decision.
The claim in the mold class action is that because of the
low volume and temperature of the water in the front‐
loading machines compared to its volume and temperature
in the traditional top‐loading machines, they don’t clean
themselves adequately and as a result mold accumulates
that emits bad odors. Traditional household cleaners do not
eliminate the molds or the odors. Roughly 200,000 of these
Kenmore‐brand machines are sold each year and there have
been many thousands of complaints of bad odors by the ma‐
chines’ owners.
Sears contends that Whirlpool (the manufacturer of the
washing machines) made a number of design modifications,
and as a result different models are differently defective;
Sears does not contend that any of the design changes elimi‐
nated the odor problem, only that they diminished it. The
basic question presented by the mold claim—are the ma‐
chines defective in permitting mold to accumulate and gen‐
erate noxious odors?—is common to the entire mold class,
although damages are likely to vary across class members
(the owners of the washing machines). A class action is the
4 Nos. 11‐8029, 12‐8030
efficient procedure for litigation of a case such as this, a case
involving a defect that may have imposed costs on tens of
thousands of consumers, yet not a cost to any one of them
large enough to justify the expense of an individual suit. A
determination of liability could be followed by individual
hearings to determine the damages sustained by each class
member. The parties probably would agree on a schedule of
damages based on the cost of fixing or replacing class mem‐
bers’ mold‐contaminated washing machines. In that event
the hearings would be brief; indeed the case would probably
be quickly settled.
We added that if it turned out as the litigation unfolded
that there were large differences in the mold problem among
the differently designed washing machines, the district
judge might decide to create subclasses (and for the further
reason that Sears’ liability might vary across the states em‐
braced by the class action because of differences among
those states’ laws), but that this possibility was not an obsta‐
cle to certification of a single mold class at the outset.
Sears argued that most members of the plaintiff class had
not experienced any mold problem. But if so, we pointed
out, that was an argument not for refusing to certify the class
but for certifying it and then entering a judgment that would
largely exonerate Sears—a course it should welcome, as all
class members who did not opt out of the class action would
be bound by the judgment.
The second class action involves a computer device that
gives instructions to a washing machine’s various moving
parts. In 2004 the company that supplied these control units
in Kenmore washing machines altered its manufacturing
process in a way that caused some control units mistakenly
Nos. 11‐8029, 12‐8030 5
to “believe” that a serious error had occurred and therefore
to order the machine to shut down, though actually there
had been no error. The plaintiffs allege that Sears knew
about the problem yet charged each owner of a defective
machine hundreds of dollars to repair the central control
unit, and that after the defect was corrected in 2005, Sears
continued to ship machines containing the earlier‐
manufactured, defective units.
The principal issue in the control‐unit class action is
whether the control unit is indeed defective. The only indi‐
vidual issues concern the amount of harm to particular class
members, and we pointed out that it was more efficient for
the principal issue—common to all class members—to be
resolved in a single proceeding than for it to be litigated sep‐
arately in hundreds of different trials. But we added that, as
with the mold class action, the district court would want to
consider whether to create different subclasses of the control
unit class for the different states because of different state
laws.
So how does the Supreme Court’s Comcast decision bear
on the rulings, just summarized, in our first decision?
Comcast holds that a damages suit cannot be certified to
proceed as a class action unless the damages sought are the
result of the class‐wide injury that the suit alleges. Comcast
was an antitrust suit, and the Court said that “if [the plain‐
tiffs] prevail on their claims, they would be entitled only to
damages resulting from reduced overbuilder competition,
since that is the only theory of antitrust impact accepted for
class‐action treatment by the District Court. It follows that a
model purporting to serve as evidence of damages in this
class action must measure only those damages attributable
6 Nos. 11‐8029, 12‐8030
to that theory. If the model does not even attempt to do that,
it cannot possibly establish that damages are susceptible of
measurement across the entire class for purposes of Rule
23(b)(3).” 133 S. Ct. at 1433. “[A] methodology that identifies
damages that are not the result of the wrong” is an impermissi‐
ble basis for calculating class‐wide damages. Id. at 1434 (em‐
phasis added). “For all we know, cable subscribers in
Gloucester County may have been overcharged because of
petitioners’ alleged elimination of satellite competition (a
theory of liability that is not capable of classwide proof).” Id. (em‐
phasis added). And on the next page of its opinion the Court
quotes approvingly from Federal Judicial Center, Reference
Manual on Scientific Evidence 432 (3d ed. 2011), that “the first
step in a damages study is the translation of the legal theory of
the harmful event into an analysis of the economic impact of
that event.” (emphasis the Court’s). None of the parties had
even challenged the district court’s ruling that class certifica‐
tion required “that the damages resulting from…[the anti‐
trust violation] were measurable ‘on a class‐wide basis’
through use of a ‘common methodology.’” 133 S. Ct. at 1430.
Unlike the situation in Comcast, there is no possibility in
this case that damages could be attributed to acts of the de‐
fendants that are not challenged on a class‐wide basis; all
members of the mold class attribute their damages to mold
and all members of the control‐unit class to a defect in the
control unit.
Sears argues that Comcast rejects the notion that efficiency
is a proper basis for class certification, and thus rejects our
statement that “predominance” of issues common to the en‐
tire class, a requirement of a damages class action under
Rule 23(b)(3), “is a question of efficiency.” 702 F.3d at 362.
Nos. 11‐8029, 12‐8030 7
But in support of its argument Sears cites only the statement
in the dissenting opinion in Comcast that “economies of time
and expense” favor class certification, 133 S. Ct. at 1437—a
statement that the majority opinion does not contradict.
Sears is wrong to think that anything a dissenting opinion
approves of the majority must disapprove of.
Sears compares the design changes that may have affect‐
ed the severity of the mold problem to the different antitrust
liability theories in Comcast. But it was not the existence of
multiple theories in that case that precluded class certifica‐
tion; it was the plaintiffs’ failure to base all the damages they
sought on the antitrust impact—the injury—of which the
plaintiffs were complaining. In contrast, any buyer of a
Kenmore washing machine who experienced a mold prob‐
lem was harmed by a breach of warranty alleged in the
complaint.
Furthermore and fundamentally, the district court in our
case, unlike Comcast, neither was asked to decide nor did de‐
cide whether to determine damages on a class‐wide basis.
As we explained in McReynolds v. Merrill Lynch, Pierce, Fen‐
ner & Smith, Inc., 672 F.3d 482, 491–92 (7th Cir. 2012), distin‐
guishing Wal‐Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (2011),
a class action limited to determining liability on a class‐wide
basis, with separate hearings to determine—if liability is es‐
tablished—the damages of individual class members, or
homogeneous groups of class members, is permitted by Rule
23(c)(4) and will often be the sensible way to proceed. See
Advisory Committee Notes to 1966 Amendment of Rule
23(b)(3); Pella Corp. v. Saltzman, 606 F.3d 391, 393–94 (7th Cir.
2010) (per curiam).
8 Nos. 11‐8029, 12‐8030
But if we are right that this is a very different case from
Comcast, why did the Supreme Court remand the case to us
for reconsideration in light of that decision? The answer
must lie in the emphasis that the majority opinion places on
the requirement of predominance and on its having to be
satisfied by proof presented at the class certification stage
rather than deferred to later stages in the litigation. See 133
S. Ct. at 1432–33. The Court doesn’t want a class action suit
to drag on for years with the parties and the district judge
trying to figure out whether it should have been certified.
Because the class in Comcast was (in the view of the majority)
seeking damages beyond those flowing from the theory of
antitrust injury alleged by the plaintiffs, the possibility
loomed that “questions affecting only individual members”
of the class would predominate over questions “common to
class members,” rather than, as Rule 23(b)(3) requires, the
reverse.
Sears argued passionately in its petition for certiorari that
we had failed to make a sufficiently rigorous inquiry into
predominance in allowing the two classes (the mold class
and the control‐unit class) to be certified. The petition was
filed before the Supreme Court issued its decision in Com‐
cast, and the Court may have felt that Sears should be al‐
lowed to amend its submission in light of Comcast and sub‐
mit its amended argument to us in the first instance.
Sears thinks that predominance is determined simply by
counting noses: that is, determining whether there are more
common issues or more individual issues, regardless of rela‐
tive importance. That’s incorrect. An issue “central to the va‐
lidity of each one of the claims” in a class action, if it can be
resolved “in one stroke,” can justify class treatment. Wal‐
Nos. 11‐8029, 12‐8030 9
Mart Stores, Inc. v. Dukes, supra, 131 S. Ct. at 2551. That was
said in the context of Rule 23(a)(2), the rule that provides
that class actions are permissible only when there are issues
common to the members of the class (as of course there are
in this case). But predominance requires a qualitative as‐
sessment too; it is not bean counting. In Amgen Inc. v. Con‐
necticut Retirement Plans & Trust Funds, supra, 133 S. Ct. at
1196, the Court said that the requirement of predominance is
not satisfied if “individual questions…overwhelm questions
common to the class,” and in Amchem Products, Inc. v. Wind‐
sor, 521 U.S. 591, 623 (1997), it said that the “predominance
inquiry tests whether proposed classes are sufficiently cohe‐
sive to warrant adjudication by representation.” And in In re
Inter‐Op Hip Prosthesis Liability Litigation, 204 F.R.D. 330, 345
(N.D. Ohio 2001), we read that “common issues need only
predominate, not outnumber individual issues.” Or as we
put it in Messner v. Northshore University HealthSystem, 669
F.3d 802, 819 (7th Cir. 2012), “Under the district court’s ap‐
proach [which our decision in Messner rejected], Rule
23(b)(3) would require not only common evidence and
methodology, but also common results for members of the
class. That approach would come very close to requiring
common proof of damages for class members, which is not
required. To put it another way, the district court asked not
for a showing of common questions, but for a showing of
common answers to those questions. Rule 23(b)(3) does not
impose such a heavy burden.”
It would drive a stake through the heart of the class ac‐
tion device, in cases in which damages were sought rather
than an injunction or a declaratory judgment, to require that
every member of the class have identical damages. If the is‐
sues of liability are genuinely common issues, and the dam‐
10 Nos. 11‐8029, 12‐8030
ages of individual class members can be readily determined
in individual hearings, in settlement negotiations, or by crea‐
tion of subclasses, the fact that damages are not identical
across all class members should not preclude class certifica‐
tion. Otherwise defendants would be able to escape liability
for tortious harms of enormous aggregate magnitude but so
widely distributed as not to be remediable in individual
suits. As we noted in Carnegie v. Household Int’l, Inc., 376 F.3d
656, 661 (7th Cir. 2004), “the more claimants there are, the
more likely a class action is to yield substantial economies in
litigation. It would hardly be an improvement to have in lieu
of this single class 17 million suits each seeking damages of
$15 to $30…. The realistic alternative to a class action is not
17 million individual suits, but zero individual suits, as only
a lunatic or a fanatic sues for $30” (emphasis in original).
The present case is less extreme: tens of thousands of class
members, each seeking damages of a few hundred dollars.
But few members of such a class, considering the costs and
distraction of litigation, would think so meager a prospect
made suing worthwhile.
There is a single, central, common issue of liability:
whether the Sears washing machine was defective. Two sep‐
arate defects are alleged, but remember that this class action
is really two class actions. In one the defect alleged involves
mold, in the other the control unit. Each defect is central to
liability. Complications arise from the design changes and
from separate state warranty laws, but can be handled by the
creation of subclasses. See, e.g., Johnson v. Meriter Health Ser‐
vices Employee Retirement Plan, supra, 702 F.3d at 365 (10 sub‐
classes). These are matters for the district judge to consider
in the first instance, and Sears will be able to present to her
Nos. 11‐8029, 12‐8030 11
the evidence it’s obtained since the district judge ruled on
certification almost two years ago.
One last point. Shortly before our original decision, the
Sixth Circuit had upheld the certification of a single mold
class in a case identical to this one (the defendant, Whirlpool,
was the manufacturer of the defective Kenmore‐brand wash‐
ing machines), except that it did not involve the other claim
in our case, the control unit claim. In re Whirlpool Corp. Front‐
Loading Washer Products Liability Litigation, 678 F.3d 409 (6th
Cir. 2012). Whirlpool sought certiorari, and the Supreme
Court granted it, vacated the court of appeals’ judgment,
and remanded the case, just as in our case. 133 S. Ct. 1722
(2013) (mem.). On remand the Sixth Circuit, denying as we
have done the defendant’s motion to remand to the district
court, and interpreting Comcast as we do, concluded that the
requirement of predominance had been satisfied. In re
Whirlpool Corp. Front‐Loading Washer Products Liability Litiga‐
tion, No. 10‐4188, 2013 WL 3746205, at *2, *16–19 (6th Cir. Ju‐
ly 18, 2013). The concordance in reasoning and result of our
decision and the Sixth Circuit’s decision averts an intercir‐
cuit conflict.
Our judgment of November 13, 2012, is hereby reinstat‐
ed.