Kent Eubank v. Pella Corporation ( 2014 )

In the
United States Court of Appeals
For the Seventh Circuit
____________________
Nos. 13-2091, -2133, -2136, -2162, -2202
KENT EUBANK, et al.,
Plaintiffs-Appellants,
and
LEONARD E. SALTZMAN, et al.,
Plaintiffs-Appellees,
v.
PELLA CORPORATION and PELLA WINDOWS
AND DOORS, INC.,
Defendants-Appellees.
APPEALS OF: RON PICKERING and MICHAEL J. SCHULZ,
Objecting class members.
____________________
Appeals from the United States District Court for the
Northern District of Illinois, Eastern Division.
No. 06 C 4481 — James B. Zagel, Judge.
____________________
ARGUED APRIL 22, 2014 — DECIDED JUNE 2, 2014
____________________
2 Nos. 13-2091, -2133, -2136, -2162, -2202
Before POSNER, WILLIAMS, and TINDER, Circuit Judges.
POSNER, Circuit Judge. The class action is an ingenious
procedural innovation that enables persons who have suf-
fered a wrongful injury, but are too numerous for joinder of
their claims alleging the same wrong committed by the same
defendant or defendants to be feasible, to obtain relief as a
group, a class as it is called. The device is especially im-
portant when each claim is too small to justify the expense of
a separate suit, so that without a class action there would be
no relief, however meritorious the claims. Normally only a
few of the claimants are named as plaintiffs (sometimes only
one, though there are several in this case). The named plain-
tiffs are the representatives of the class—fiduciaries of its
members—and therefore charged with monitoring the law-
yers who prosecute the case on behalf of the class (class
counsel). They receive modest compensation, in addition to
their damages as class members, for their normally quite
limited services—often little more than sitting for a deposi-
tion—as class representatives. Invariably they are selected
by class counsel, who as a practical matter control the litiga-
tion by the class. The selection of the class representatives by
class counsel inevitably dilutes their fiduciary commitment.
The class action is a worthwhile supplement to conven-
tional litigation procedure, David L. Shapiro, “Class Actions:
The Class As Party and Client,” 73 Notre Dame L. Rev. 913,
923–24 (1998); Arthur R. Miller, “Of Frankenstein Monsters
and Shining Knights: Myth, Reality, and the ‘Class Action
Problem’,” 92 Harv. L. Rev. 664, 666–68 (1979), but it is con-
troversial and embattled, see Robert H. Klonoff, “The De-
cline of Class Actions,” 90 Wash. U. L. Rev. 729, 731–33
(2013), in part because it is frequently abused. Martin H. Re-
Nos. 13-2091, -2133, -2136, -2162, -2202 3
dish, Wholesale Justice: Constitutional Democracy and the Prob-
lem of the Class Action Lawsuit 1–2 (2009); Jonathan R. Macey
& Geoffrey P. Miller, “The Plaintiffs’ Attorney’s Role in
Class Action and Derivative Litigation: Economic Analysis
and Recommendations for Reform,” 58 U. Chi. L. Rev. 1, 3–4
(1991); John C. Coffee, Jr., “Rethinking the Class Action: A
Policy Primer on Reform,” 62 Ind. L.J. 625, 627 (1987). The
control of the class over its lawyers usually is attenuated, of-
ten to the point of nonexistence. Except for the named plain-
tiffs, the members of the class are more like beneficiaries
than like parties; for although they are authorized to appeal
from an adverse judgment, Smith v. Bayer Corp., 131 S. Ct.
2368, 2379 (2011); Devlin v. Scardelletti, 536 U.S. 1, 9–10 (2002),
they have no control over class counsel. In principle the
named plaintiffs do have that control, but as we’ve already
hinted this is rarely true in practice. Class actions are the
brainchildren of the lawyers who specialize in prosecuting
such actions, and in picking class representatives they have
no incentive to select persons capable or desirous of moni-
toring the lawyers’ conduct of the litigation.
A high percentage of lawsuits is settled—but a study of
certified class actions in federal court in a two-year period
(2005 to 2007) found that all 30 such actions had been settled.
Emery G. Lee III et al., “Impact of the Class Action Fairness
Act on the Federal Courts” 2, 11 (Federal Judicial Center
2008). The reasons that class actions invariably are settled are
twofold. Aggregating a great many claims (sometimes tens
or even hundreds of thousands—occasionally millions) often
creates a potential liability so great that the defendant is un-
willing to bear the risk, even if it is only a small probability,
of an adverse judgment. At the same time, class counsel, un-
governed as a practical matter by either the named plaintiffs
4 Nos. 13-2091, -2133, -2136, -2162, -2202
or the other members of the class, have an opportunity to
maximize their attorneys’ fees—which (besides other ex-
penses) are all they can get from the class action—at the ex-
pense of the class. The defendant cares only about the size of
the settlement, not how it is divided between attorneys’ fees
and compensation for the class. From the selfish standpoint
of class counsel and the defendant, therefore, the optimal
settlement is one modest in overall amount but heavily tilted
toward attorneys’ fees. As we said in Creative Montessori
Learning Centers v. Ashford Gear LLC, 662 F.3d 913, 918 (7th
Cir. 2011), “we and other courts have often remarked the in-
centive of class counsel, in complicity with the defendant’s
counsel, to sell out the class by agreeing with the defendant
to recommend that the judge approve a settlement involving
a meager recovery for the class but generous compensation
for the lawyers—the deal that promotes the self-interest of
both class counsel and the defendant and is therefore opti-
mal from the standpoint of their private interests. Reynolds v.
Beneficial National Bank, [288 F.3d 277, 279 (7th Cir. 2002)];
Culver v. City of Milwaukee, [277 F.3d 908, 910 (7th Cir. 2002)];
Greisz v. Household Bank (Illinois), N.A., 176 F.3d 1012, 1013
(7th Cir. 1999); Duhaime v. John Hancock Mutual Life Ins. Co.,
183 F.3d 1, 7 (1st Cir. 1999); In re General Motors Corp. Pick-Up
Truck Fuel Tank Products Liability Litigation, 55 F.3d 768, 805
(3d Cir. 1995); Plummer v. Chemical Bank, 668 F.2d 654, 658
(2d Cir. 1982).”
Fortunately the settlement, including the amount of at-
torneys’ fees to award to class counsel, must be approved by
the district judge presiding over the case; unfortunately
American judges are accustomed to presiding over adver-
sary proceedings. They expect the clash of the adversaries to
generate the information that the judge needs to decide the
Nos. 13-2091, -2133, -2136, -2162, -2202 5
case. And so when a judge is being urged by both adver-
saries to approve the class-action settlement that they’ve ne-
gotiated, he’s at a disadvantage in evaluating the fairness of
the settlement to the class. In re General Motors Corp. Pick-Up
Truck Fuel Tank Products Liability Litigation, supra, 55 F.3d at
789–90; Redish, supra, at 188.
Enter the objectors. Members of the class who smell a rat
can object to approval of the settlement. See, e.g., Reynolds v.
Beneficial National Bank, supra, 288 F.3d at 287–88; Edward
Brunet, “Class Action Objectors: Extortionist Free Riders or
Fairness Guarantors,” 2003 U. Chi. Legal F. 403, 411–12. If
their objections persuade the judge to disapprove it, and as a
consequence a settlement more favorable to the class is nego-
tiated and approved, the objectors will receive a cash award
that can be substantial, as in In re Trans Union Corp. Privacy
Litigation, 629 F.3d 741 (7th Cir. 2011).
In this case, despite the presence of objectors, the district
court approved a class action settlement that is inequitable—
even scandalous. The case underscores the importance both
of objectors (for they are the appellants in this case—without
them there would have been no appellate challenge to the
settlement) and of intense judicial scrutiny of proposed class
action settlements.
The suit was filed in the summer of 2006, almost eight
years ago. Federal jurisdiction was based on the Class Action
Fairness Act’s grant of federal jurisdiction over class actions
in which there is at least minimal (as distinct from complete)
diversity of citizenship. 28 U.S.C. § 1332(d)(2)(A). The de-
fendants are Pella Corporation and an affiliate that we can
ignore. Pella is a leading manufacturer of windows. The suit
alleges that its “ProLine Series” casement windows (a case-
6 Nos. 13-2091, -2133, -2136, -2162, -2202
ment window is a window attached to its frame by hinges at
the side) manufactured and sold between 1991 and 2006 had
a design defect that allowed water to enter behind the win-
dow’s exterior aluminum cladding and cause damage to the
window’s wooden frame and to the house itself. Pella’s sale
of the defective windows is alleged to have violated the
product-liability and consumer-protection laws of a number
of states in which the windows were sold.
The district judge certified two separate classes: one for
customers who had already replaced or repaired their defec-
tive windows, the other for those who hadn’t. The latter
class sought only declaratory relief and so was nationwide,
but the former sought damages and was limited to custom-
ers in six states, with a separate subclass for each state. We
upheld the certifications over Pella’s objections in Pella Corp.
v. Saltzman, 606 F.3d 391 (7th Cir. 2010) (per curiam).
Class counsel negotiated a settlement of the class action
with Pella in the fall of 2011. The district judge gave final
approval to the settlement in 2013, precipitating the objec-
tors’ appeals. The settlement agreement ignores the certifica-
tion of the two classes and purports to bind a single nation-
wide class consisting of all owners of Pella ProLine windows
containing the defect, whether or not the owners have al-
ready replaced or repaired the windows. This provision is
the first of many red flags that the judge failed to see: “the
adversity among subgroups requires that the members of
each subgroup cannot be bound to a settlement except by
consents given by those who understand that their role is to
represent solely the members of their respective subgroups.”
In re Joint Eastern & Southern District Asbestos Litigation, 982
F.2d 721, 743 (2d Cir. 1992); see also Amchem Products, Inc. v.
Nos. 13-2091, -2133, -2136, -2162, -2202 7
Windsor, 521 U.S. 591, 627–28 (1997); Smith v. Sprint Commu-
nications Co., 387 F.3d 612, 614–15 (7th Cir. 2004).
Initially there was only one named plaintiff, a dentist
named Leonard E. Saltzman. His son-in-law, Paul M. Weiss,
was lead counsel for the class, continuing in that role
throughout the district court proceedings that culminated in
the approval of the settlement. Technically the law firm of
which he is the founder and senior partner (Complex Litiga-
tion Group LLC) is a lead class counsel too, along with two
of his partners in the firm. The settlement agreement desig-
nates still another firm as a lead class counsel as well; but the
fee petition describes that firm as merely a class counsel. The
agreement gave lead class counsel “sole discretion” to allo-
cate the award of attorneys’ fees to which the parties had
agreed among the class counsel, and Weiss proposed to allo-
cate 73 percent of the fees to his own firm. Realistically he
was the lead class counsel.
Weiss’s wife—Saltzman’s daughter—is a lawyer too, and
a partner in her husband’s firm. Both spouses are defendants
in a lawsuit charging them with misappropriation of the as-
sets of their former law firm, Freed & Weiss LLC, and other
misconduct relating to that firm. Freed v. Weiss, No. 2011-
CH-41529 (Ill. Cook County Ch. Div.). Weiss is also a de-
fendant in a second, similar suit, Lang v. Weiss, No. 2012-CH-
05863 (Ill. Cook County Ch. Div.). (The two suits are dis-
cussed in Sarah Zavala, “Cook County Suits Involve Alleged
Takeover at Freed and Weiss,” Madison-St. Clair Record,
March 7, 2012, pp. 1, 8.) The Freed & Weiss firm was still an-
other class counsel in the present case; and one of the objec-
tors points out that “the dissolution and descent into open
warfare that consumed Freed & Weiss in 2011 and 2012
8 Nos. 13-2091, -2133, -2136, -2162, -2202
clearly rendered that firm inadequate class counsel, especial-
ly in light of the articulated financial needs of the partners
that drove the settlement of this case.” And six weeks ago
the Hearing Board of the Illinois Attorney Registration and
Disciplinary Commission recommended in a 94-page report
that the Supreme Court of Illinois suspend Weiss from prac-
ticing law for 30 months because of repeated misconduct. In
re Paul M. Weiss, No. 08 CH 116 (Ill. Att’y Registration &
Disciplinary Commission Hearing Board, Apr. 17, 2014). The
recommended penalty is severe by Illinois standards; the
state allows lawyers sanctioned with “disbarment” to apply
for reinstatement to the bar after 60 (in some cases just 36)
months. Ill. S. Ct. R. 767(a); Illinois Attorney Registration &
Disciplinary Commission, Annual Report of 2013, at 21, 25.
The impropriety of allowing Saltzman to serve as class
representative as long as his son-in-law was lead class coun-
sel was palpable. See Greisz v. Household Bank (Illinois), 176
F.3d 1012, 1014 (7th Cir. 1999); Petrovic v. Amoco Oil Co., 200
F.3d 1140, 1155 (8th Cir. 1999); Zylstra v. Safeway Stores, Inc.,
578 F.2d 102, 104 (5th Cir. 1978); Turoff v. May Co., 531 F.2d
1357, 1360 (6th Cir. 1976) (per curiam); “Developments in the
Law—Class Actions,” 89 Harv. L. Rev. 1318, 1585–86 n. 29
(1976). Weiss may have been desperate to obtain a large at-
torney’s fee in this case before his financial roof fell in on
him.
Early in the case four other class members had been add-
ed as plaintiffs, making a total of five including Saltzman.
When the settlement was presented to the district court for
preliminary approval, the four class members who had been
added as named plaintiffs opposed it, leaving only Saltzman
among the original class members to support it. But pursu-
Nos. 13-2091, -2133, -2136, -2162, -2202 9
ant to a motion filed by George Lang, who at the time was a
partner of Weiss, four other class members were added as
named plaintiffs. (Lang says that Weiss rather than he
picked them.)
Weiss removed the original four class members who had
opposed the settlement; naturally their replacements joined
Saltzman in supporting it.
Lang now represents the defrocked named plaintiffs,
who are four of the six objectors. A lawyer’s switching sides
in the same lawsuit would normally be considered a fatal
conflict of interest, but the courts are lenient when it is a
class action lawyer. E.g., Bash v. Firstmark Standard Life Ins.
Co., 861 F.2d 159, 161 (7th Cir. 1988). For often “only the at-
torneys who have represented the class, rather than any of
the class members themselves, have substantial familiarity
with the prior proceedings, the fruits of discovery, the actual
potential of the litigation. And when an action has continued
over the course of many years, the prospect of having those
most familiar with its course and status be automatically
disqualified whenever class members have conflicting inter-
ests would substantially diminish the efficacy of class ac-
tions as a method of dispute resolution.” In re “Agent Or-
ange” Product Liability Litigation, 800 F.2d 14, 18–19 (2d Cir.
1986).
As finally approved by the district judge, the settlement
directed Pella to pay $11 million in attorneys’ fees to class
counsel. The basis of this figure was the plaintiffs’ claim that
the settlement was worth $90 million to the class. Were that
so, then considering the multistate scope of the suit and per-
haps the length of time that elapsed between its filing and
the approval of the settlement by the district court in May
10 Nos. 13-2091, -2133, -2136, -2162, -2202
2013 (our “perhaps” reflecting doubt that the time was well
spent), the fee award, equal to 12 percent of the amount of
the settlement earmarked for the class members, would have
been defensible. But the settlement did not specify an
amount of money to be received by the class members as
distinct from class counsel. Rather it specified a procedure
by which class members could claim damages. So there was
an asymmetry: class counsel was to receive its entire award
of attorneys’ fees up front; class members were to obtain
merely contingent claims, albeit with a (loosely) estimated
value of $90 million (actually far less, as we’ll see).
The named plaintiffs were each awarded compensation
(an “incentive award,” as it is called) for their services to the
class of either $5,000 or $10,000, depending on their role in
the case. Saltzman, being the lead class representative, was
slated to be a $10,000 recipient.
Although the judge rightly made incentive awards to the
class representatives who had opposed the settlement as
well as to those who had approved it, the settlement agree-
ment itself had provided for incentive awards only to the
representatives who supported the settlement. This created a
conflict of interest: any class representative who opposed the
settlement would expect to find himself without any com-
pensation for his services as representative. Still another
questionable provision of the settlement, which the judge
refused to delete, made any reduction in the $11 million at-
torneys’ fee award revert to Pella, rather than being added to
the compensation of the class members.
Not only did the settlement agreement not quantify the
benefits to the class members, but the judge approved it be-
fore the deadline for filing claims. He made no attempt to
Nos. 13-2091, -2133, -2136, -2162, -2202 11
estimate how many claims were likely to be filed, though
without such an estimate no responsible prediction of the
value of the settlement to the members of the class could be
made. Furthermore, the judge’s approval of the settlement
(over the objection of the former class representatives and
other class members) is squeezed into two two-page orders
(the second addressed to the attorneys’ fee award) that ig-
nore virtually all the objections to the settlement. Unheeded
was our warning that “because class actions are rife with po-
tential conflicts of interest between class counsel and class
members, district judges presiding over such actions are ex-
pected to give careful scrutiny to the terms of proposed set-
tlements in order to make sure that class counsel are behav-
ing as honest fiduciaries for the class as a whole.” Mirfasihi v.
Fleet Mortgage Corp., 356 F.3d 781, 785 (7th Cir. 2004) (cita-
tions omitted).
The settlement should have been disapproved on multi-
ple grounds. To begin with, it was improper for the lead
class counsel to be the son-in-law of the lead class repre-
sentative. Class representatives are, as we noted earlier, fi-
duciaries of the class members, and fiduciaries are not al-
lowed to have conflicts of interest without the informed con-
sent of their beneficiaries, which was not sought in this case.
Only a tiny number of class members would have known
about the family relationship between the lead class repre-
sentative and the lead class counsel—a relationship that cre-
ated a grave conflict of interest; for the larger the fee award
to class counsel, the better off Saltzman’s daughter and son-
in-law would be financially—and (which sharpened the con-
flict of interest) by a lot. They may well have had an acute
need for an infusion of money, in light not only of Weiss’s
ethical embroilment, which cannot help his practice, but also
12 Nos. 13-2091, -2133, -2136, -2162, -2202
of the litigation against him by his former law partners and
his need for money to finance his new firm. The appellees
(primarily Saltzman, who is still a named plaintiff, and Pella)
point out that Saltzman was one of five class representatives,
and the other four didn’t have a conflict of interest. But the
four other original class representatives had opposed the set-
tlement, whereupon they had been replaced by new named
plaintiffs—selected by the conflicted lead class counsel.
Weiss’s ethical embroilment was another compelling rea-
son for kicking him and Saltzman off the case. The discipli-
nary proceeding against Weiss was already under way when
the settlement agreement was negotiated. It was very much
in his personal interest, as opposed to the interest of the class
members, to get the settlement signed and approved before
the disciplinary proceeding culminated in a sanction that
might abrogate his right to share in the attorneys’ fee award
in this case. He could negotiate a quick settlement only by
giving ground to Pella, which upon discovering the box that
Weiss was in would have stiffened its terms (it plays hard-
ball, as its conduct throughout this litigation has demon-
strated).
So Weiss’s ethical troubles should have disqualified him
from serving as class counsel even if his father-in-law hadn’t
been in the picture. Another suspicious feature of the settle-
ment, doubtless also related to Weiss’s woes, was Pella’s
agreeing to a $2 million advance of attorneys’ fees to lead
class counsel before notice of the settlement was sent to the
members of the class.
Counsel for a certified class is appointed by the district
judge presiding over the class action, and in deciding to ap-
point a lawyer to be class counsel the court “may consider,”
Nos. 13-2091, -2133, -2136, -2162, -2202 13
besides the lawyer’s competence, experience, and related
professional qualifications, “any other matter pertinent to
counsel’s ability to fairly and adequately represent the inter-
ests of the class.” Fed. R. Civ. P. 23(g)(1)(B), (g)(4). “When
class counsel have demonstrated a lack of integrity, a court
can have no confidence that they will act as conscientious
fiduciaries of the class.” Creative Montessori Learning Centers
v. Ashford Gear LLC, supra, 662 F.3d at 918. Weiss was unfit to
represent the class.
Rule 23(a)(4) of the Federal Rules of Civil Procedure re-
quires that “the representative parties will fairly and ade-
quately protect the interests of the class.” This both Saltzman
and the other class representatives who approved the set-
tlement failed to do. The settlement that the district judge
approved is stacked against the class. Pella itself estimates
the value of the settlement to the class at only $22.5 mil-
lion—and that is an overestimate. The settlement strews ob-
stacles in the path of any owner of a defective ProLine Series
casement window. A member of the class may either file a
claim with Pella, period, or file a claim that he must submit
to arbitration with Pella. If he chooses the first option, he is
limited to a maximum damages award of $750 per “Struc-
ture,” confusingly defined not as a window but as the entire
building containing the window. There’s also a per-window
damages cap that ranges from $60 to $100 (with an addition-
al $0 to $250 for the cost of installation), depending on when
the class member purchased his window and when he re-
placed it. And the cap falls to zero unless he gave “notice” to
Pella before replacing the defective window.
A class member who chooses arbitration can receive up
to $6000 per “Structure” (defined the same way), and doesn’t
14 Nos. 13-2091, -2133, -2136, -2162, -2202
have to prove that his window or windows were in fact de-
fective, only that they were in the category of Pella windows
that contained the design defect. But if Pella convinces the
arbitrator that the damage the claimant is seeking compensa-
tion for was not caused by the defect or by “any other defect
in the structure” (whatever that means), or that the claimant
was compensated for the damage from some other source,
the claimant gets nothing; and likewise if Pella successfully
interposes a complete defense, such as that the statute of lim-
itations had run. The settlement allows Pella to assert ten
categories of defenses, including “natural weathering.” And
the limitations periods applicable to the class members’
claims vary from three to five years and involve different
accrual and tolling rules. Statutes of repose are also in the
picture.
Pella also reserved in the settlement agreement the right
to plead and prove partial defenses such as comparative
fault and failure to mitigate damages. And some claimants
are entitled only to “coupons” (discounts on future purchas-
es of Pella windows, discounts that may be worth very little
to current owners of Pella’s defective windows)—a warning
sign of a questionable settlement. Synfuel Technologies, Inc. v.
DHL Express (USA), Inc., 463 F.3d 646, 654 (7th Cir. 2006); In
re HP Inkjet Printer Litigation, 716 F.3d 1173, 1179–80 (9th Cir.
2013); Christopher R. Leslie, “The Need to Study Coupon
Settlements in Class Action Litigation,” 18 Geo. J. Legal Ethics
1395, 1396–98 (2005); Geoffrey P. Miller & Lori S. Singer,
“Nonpecuniary Class Action Settlements,” Law & Contemp.
Probs., vol. 60, Autumn 1997, pp. 97, 108; cf. 28 U.S.C. § 1712
(Coupon Settlements).
Nos. 13-2091, -2133, -2136, -2162, -2202 15
Some class members may be entitled only to an extension
of warranty, under a program (the “ProLine Service En-
hancement Program”) that Pella had adopted before the set-
tlement and that requires class members to deduct $100 per
window from the cost of installation or other labor services
required to replace it. The $90 million estimate of the value
of the settlement to the class includes the value of these war-
ranty extensions even though they were a contractual enti-
tlement that preceded the settlement rather than being con-
ferred by it and thus were not part of the value created by
the settlement, although the settlement does forbid Pella to
revoke the extensions, which confers a bit of extra value.
The claim forms are long—12 pages for the “simple”
claim with its $750 ceiling, 13 pages for the claim that has the
higher ceiling ($6000) but requires the claimant to run the
gauntlet of arbitration, doubtless without assistance of coun-
sel or expert witnesses, because the legal fees and experts’
fees would quickly mount to or above $6000, leaving the
claimant with nothing or even less than nothing: additional
bills to pay. There is no provision for shifting the legal or ex-
pert-witness costs of a victorious claimant in the arbitration
proceeding to Pella.
Both forms require a claimant to submit a slew of arcane
data, including the “Purchase Order Number,” “Glass Etch
Information,” “Product Identity Stamp,” and “Unit ID La-
bel” of each affected window. The claim forms are so com-
plicated that Pella could reject many of them on the ground
that the claimant had not filled out the form completely and
correctly.
And that’s assuming that class members even attempt to
file claims. The notice of settlement that was sent to them is
16 Nos. 13-2091, -2133, -2136, -2162, -2202
divided into 27 sections, some with a number of subsections.
For example, the section on eligibility for benefits under the
settlement lists nine criteria that must be satisfied while the
section on “How Do I Get Out of the Settlement?” specifies
six requirements that must be met for a class member to be
allowed to opt out of the settlement. And to object to the set-
tlement the class member must satisfy seven other criteria,
one of which is again multiple, as it requires listing “each
specific reason for your objection.”
Considering the modesty of the settlement, the length
and complexity of the forms, and the unfamiliarity of the av-
erage homeowner with arbitration, we’re not surprised that
only 1276 claims (of which only 97 sought arbitration) had
been filed as of February 2013, out of the more than 225,000
notices that had been sent to class members. The claims
sought in the aggregate less than $1.5 million and were like-
ly to be worth even less because Pella would be almost cer-
tain to prevail in some, maybe most, of the arbitration pro-
ceedings. It’s been found that on average consumers prevail
in arbitration roughly half the time, and those who win are
awarded roughly half of what they seek. Christopher R.
Drahozal & Samantha Zyontz, “An Empirical Study of AAA
Consumer Arbitrations,” 25 Ohio St. J. Dispute Resolution 843,
898–900 (2010). The implication is that Pella would be able to
knock 75 percent off the damages sought by class members
who filed claims that were submitted to arbitration.
A class recovery of little more than $1 million is a long
way from the $90 million that the district judge thought the
class members likely to receive were the suit to be litigated.
It’s true that another 9500 or so simple claims were filed after
the district court entered its final judgment, plus another
Nos. 13-2091, -2133, -2136, -2162, -2202 17
1387 claims that would require arbitration. But Pella’s esti-
mate that the class will recover $22.5 million assumes against
all reason that every one of the claims will reap the maxi-
mum authorized benefits—$750 for the simple claims and
$6000 for the claims that go to arbitration. And that recovery
would be only $17 million, not $22.5 million (Pella contends,
however, that the extension of its warranty is worth another
$5.5 million to the class). There is no evidence that Pella
would pay the maximum benefits on all, or indeed on any,
of the claims.
If the average payment were half the amount of the
claim—a very generous assumption given the estimate of a
75 percent success rate for Pella—the aggregate value of the
settlement to the class ($8.5 million) would be less than the
attorneys’ fees ($11 million). Even the $8.5 million figure is
an exaggeration, because the settlement subtracts from the
award compensation received from any other source—and
one of the other sources is the warranty program.
We don’t understand the judge’s valuing the settlement
at $90 million or thinking the feeble efforts of class counsel
led by Weiss to obtain benefits for the class (as distinct from
benefits for themselves in the form of generous attorneys’
fees) worth $11 million. The restrictions that Pella was al-
lowed to place on the settlement would, if upheld, enor-
mously reduce the class members’ recovery of their losses,
and the residue is to be returned to Pella. Class counsel sold
out the class.
The class as we said could not expect to receive more
than $8.5 million from the settlement, given all the obstacles
that the terms of the settlement strewed in the path of the
class members. And even that figure seems too high. For if
18 Nos. 13-2091, -2133, -2136, -2162, -2202
the class received that amount, this would indicate that Pella
had agreed to pay attorneys’ fees equal to 56 percent of the
total settlement ($11 million = .56 × ($8.5 million + $11 mil-
lion)) in order to induce class counsel to settle the case on
terms that would minimize Pella’s overall liability.
We note the remarkable statement in Saltzman’s brief de-
fending the settlement that “in comparison to this $90 mil-
lion independent valuation of the Settlement, a trial of the
certified claims here, even with a complete victory, would
result in an award of $0.” Zero? But if Pella has no liability,
why would it agree to a $33.5 million settlement ($22.5 mil-
lion in estimated damages plus the $11 million in attorneys’
fees)? Saltzman appears to believe that the alternative of liti-
gating the class action to judgment would be infeasible be-
cause the court would go crazy trying to determine the
damages of each of several, maybe many, thousand class
members. He neglects to mention that we rejected this ar-
gument when we approved class certification. Pella Corp. v.
Saltzman, supra, 606 F.3d at 395–96; see also 1966 Advisory
Committee Notes to Fed. R. Civ. P. 23; Butler v. Sears, Roebuck
& Co., 727 F.3d 796, 801 (7th Cir. 2013); In re Whirlpool Corp.
Front-Loading Washer Products Liability Litigation, 722 F.3d
838, 860–61 (6th Cir. 2013); Tardiff v. Knox County, 365 F.3d 1,
6–7 (1st Cir. 2004); 2 Newberg on Class Actions § 4:54, pp. 205–
10 (5th ed. 2012). Pella argues that it would fight the indi-
vidual damages claims if the case were litigated. But the set-
tlement agreement allows it to fight the damages claims
submitted to it pursuant to the agreement.
In the district court Saltzman valued the case if it went to
trial at $50 million. If he was lying and actually thinks the
Nos. 13-2091, -2133, -2136, -2162, -2202 19
case worthless, how could he have been an effective class
representative even if he had had no conflict of interest?
The mystery deepens: Pella thinks the case if tried would
be worth only $14.5 million to the class members. If that is
so, why has it agreed to a settlement that it claims will cost it
$33.5 million? Because it would incur legal fees and other
expenses of more than $19 million ($14.5 million + $19 mil-
lion = $33.5 million)? But if the case were tried, class counsel
would incur heavy expenses as well, which would induce it
to settle for less than $14.5 million. The truth must be that,
protected by the bristling technicalities of the settlement
agreement, Pella does not believe that the settlement will
cost it anywhere near $14.5 million.
If Saltzman is right and damages if the case were tried
would be zero, a settlement of $90 million would be a re-
markable achievement. (Also an inexplicable one.) But the
district judge did not find that the trial would yield zero
damages. He didn’t estimate the likely outcome of a trial, as
he should have done in order to evaluate the adequacy of
the settlement. Reynolds v. Beneficial National Bank, supra, 288
F.3d at 285.
Saltzman as we said defends the $90 million figure as an
“independent valuation” of the settlement. But the only evi-
dence we can find supporting that valuation is the affidavit
of an accountant—hired and paid by Weiss’s law firm, so
hardly independent. Maybe by “independent” Saltzman is
referring (though he doesn’t say so) to the fact that the set-
tlement was mediated by two retired judges. One, however,
stopped mediating (we don’t know why) before the negotia-
tions were completed and the other limited his mediation to
issues of attorneys’ fees.
20 Nos. 13-2091, -2133, -2136, -2162, -2202
Saltzman and Pella argue that the objectors did not pre-
sent an expert witness to support their estimate of the value
of the litigation, and Saltzman did: the brother of one of
Saltzman’s lawyers! Anyway Saltzman has implicitly repu-
diated his expert, who did not testify that the value of the
suit if litigated was $0.
Saltzman and Pella point out that the notice of the set-
tlement sent to the class members provoked few objections.
Of course not; it was not intended to; it was incomplete and
misleading. It failed to mention that four of the five original
class representatives had opposed the settlement and been
promptly replaced by other persons, selected by class coun-
sel; that the only original representative who had supported
the settlement was the father-in-law of the lead class counsel
who was both in financial trouble and ethically challenged;
that up to half the recipients of the notice would if they filed
a claim and it was accepted receive only a coupon discount
on a future purchase of a Pella window; and that four of the
original class representatives believed the notice of the set-
tlement misleading because it implied that class members
would be guaranteed at least $750 or $6000 in response to
their claim, whereas these were ceilings and were not even
potential payments to those class members entitled only to
coupons. The judge was informed of these objections to the
notice but declined to order it modified. He said that the no-
tice was “fair,” that it was “a neutral communication from
the court.” It was not neutral and it did not provide a truth-
ful basis for deciding whether to opt out. The judge said the
objectors could send their own notice to the class members.
But what would the recipient of two conflicting notices do?
And it wouldn’t be just two. For if the objectors sent their
own notice class counsel would send out a rebuttal notice.
Nos. 13-2091, -2133, -2136, -2162, -2202 21
Better for the court to make sure that the single notice it sent
would be a responsible communication rather than an un-
candid communication from class counsel than to subject the
class members to a blizzard of conflicting notices.
All this is academic, however, because opting out of a
class action is very rare. Virtually no one who receives notice
that he is a member of a class in a class action suit opts out.
He doesn’t know what he could do as an opt-out. He’s un-
likely to hire a lawyer to litigate over a window. In fact the
opt-outs in this case were only one twentieth of one percent
of the recipients of the notice of approved settlement. A
study of other product-liability class actions found that the
average opt-out percentage was less than one tenth of one
percent. Theodore Eisenberg & Geoffrey Miller, “The Role of
Opt-Outs and Objectors in Class Action Litigation: Theoreti-
cal and Empirical Issues,” 57 Vand. L. Rev. 1529, 1549 (2004);
see also Mars Steel Corp. v. Continental Illinois National Bank &
Trust Co. of Chicago, 834 F.2d 677, 680–81 (7th Cir. 1987). Con-
trary to the statement in Pella’s brief, a low opt-out rate is no
evidence that a class action settlement was “fair” to the
members of the class.
In sum, almost every danger sign in a class action settle-
ment that our court and other courts have warned district
judges to be on the lookout for was present in this case. See,
e.g., Synfuel Technologies, Inc. v. DHL Express (USA), Inc., su-
pra, 463 F.3d at 654; Smith v. Sprint Communications Co., 387
F.3d 612, 614 (7th Cir. 2004); Mirfasihi v. Fleet Mortgage Corp.,
supra, 356 F.3d at 785–86; Reynolds v. Beneficial National Bank,
supra, 288 F.3d at 282–83; Crawford v. Equifax Payment Ser-
vices, Inc., 201 F.3d 877, 880 (7th Cir. 2000); In re Bluetooth
Headset Products Liability Litigation, 654 F.3d 935, 946–47 (9th
22 Nos. 13-2091, -2133, -2136, -2162, -2202
Cir. 2011); Weinberger v. Great Northern Nekoosa Corp., 925
F.2d 518, 525 (1st Cir. 1991). Most were not even mentioned
by the district judge, and those that were received a brush-
off. The settlement flunked the “fairness” standard by the
one-sidedness of its terms and the fatal conflicts of interest
on the part of Saltzman and Weiss. This is a case in which
“the lawyers support the settlement to get fees; the defend-
ants support it to evade liability; the court can’t vindicate the
class’s rights because the friendly presentation means that it
lacks essential information.” Kamilewicz v. Bank of Boston
Corp., 100 F.3d 1348, 1352 (7th Cir. 1996) (dissent from denial
of rehearing en banc).
A couple of loose ends remain to be tied up:
1. Saltzman has moved to dismiss the appeals on the
ground that the appellants—objectors to the settlement ap-
proved by the district judge—lack standing to litigate their
objections. Since absent objectors have standing to appeal
from an adverse judgment, Devlin v. Scardelletti, supra, 536
U.S. at 14, named objectors must as well. Even named plain-
tiffs who settle nevertheless have standing to appeal a denial
of class certification. Espenscheid v. DirectSat USA, LLC, 688
F.3d 872, 876 (7th Cir. 2012).
2. Objector Schulz asks us to sanction Saltzman’s lawyers
for filing the motion on standing. Saltzman’s removal as lead
plaintiff and his lawyers’ removal as class counsel are sanc-
tion enough; because the motion on standing was indeed
frivolous, little time was spent on it either by us judges or by
the objectors’ lawyers. Both motions (standing and sanc-
tions) are therefore denied.
To conclude:
Nos. 13-2091, -2133, -2136, -2162, -2202 23
After eight largely wasted years, much remains to be
done in this case. For starters, Saltzman, Paul Weiss, and
Weiss’s firm, Complex Litigation Group, must be replaced as
class representative (Saltzman), and as class counsel (Weiss
and his firm), respectively. And since we are rejecting the
settlement agreement, the plaintiffs named in the third
amended complaint, whom that agreement caused to be
substituted for the original named plaintiffs (other than
Saltzman), must be discharged and the four original named
plaintiffs (whom we’ve called the “defrocked” plaintiffs) re-
instated.
The judgment is reversed and the case remanded for fur-
ther proceedings in conformity with this opinion.
REVERSED AND REMANDED.