12‐4671‐cv(L) In re Payment Card Interchange Fee and Merchant Discount Antitrust Litigation

UNITED STATES COURT OF APPEALS

FOR THE SECOND CIRCUIT

August Term, 2015

(Argued: September 28, 2015 Decided: June 30, 2016)

Docket Nos. 12‐4671‐cv(L); 12‐4708(CON); 12‐4765(CON); 13‐4719(CON); 13‐4750(CON); 13‐4751(CON); 13‐4752(CON); 14‐32(CON); 14‐117(CON); 14‐119(CON); 14‐133(CON); 14‐157(CON); 14‐159(CON); 14‐192(CON); 14‐197(CON); 14‐219(CON); 14‐241(CON); 14‐250(CON); 14‐266(CON); 14‐303(CON); 14‐331(CON); 14‐349(CON); 14‐404(CON); 14‐422(CON); 14‐443(CON); 14‐480(CON); 14‐497(CON); 14‐530(CON); 14‐567(CON); 14‐584(CON); 14‐606(CON); 14‐663(CON); 14‐837(CON)

‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐x

IN RE PAYMENT CARD INTERCHANGE FEE AND MERCHANT DISCOUNT ANTITRUST LITIGATION

‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐x

Before: WINTER, JACOBS, and LEVAL, Circuit Judges.

This antitrust class action was brought on behalf of approximately 12

million merchants against Visa and MasterCard, which are the two largest credit

card issuing networks in the United States, as well as against various issuing and

acquiring banks, alleging a conspiracy in violation of Section 1 of the Sherman Act. After nearly ten years of litigation, the parties agreed to a settlement that

released all claims in exchange for disparate relief to each of two classes: up to

$7.25 billion would go to an opt‐out class, and a non‐opt‐out class would get

injunctive relief. The district court certified these two settlement‐only classes,

and approved the settlement as fair and reasonable. On this appeal, numerous

objectors and opt‐out plaintiffs argue that this class action was improperly

certified and that the settlement was unreasonable and inadequate. We conclude

that the class plaintiffs were inadequately represented in violation of Rule 23(a)(4)

and the Due Process Clause. Accordingly, we vacate the district court’s

certification of this class action and reverse the approval of the settlement.

Vacated, reversed, and remanded.

Judge Leval concurs in a separate opinion.

THOMAS C. GOLDSTEIN (Eric F. Citron, on the brief), Goldstein & Russell P.C., Washington, DC; Stephen R. Neuwirth, Sanford I. Weisburst, Steig D. Olson, and Cleland B. Welton II, Quinn Emanuel Urquhart & Sullivan, LLP, New York, NY; Jeffrey I. Shinder, Gary J. Malone, and A. Owen Glist, Constantine Cannon LLP, New York, NY; Michael J. Canter, Robert N. Webner, and Kenneth J. Rubin, Vorys, Sater, Seymour and Pease LLP, Columbus

2 OH; Gregory A. Clarick, Clarick Gueron Reisbaum LLP, New York, NY, for Objectors‐Appellants and Plaintiffs‐Appellants (Merchant Appellants).

PHILIP C. KOROLOGOS, Boies, Schiller & Flexner LLP, New York, NY, for Objectors‐Appellants American Express Company, et al.

JENNIFER M. SELENDY (William H. Pratt, on the brief), Kirkland & Ellis LLP, New York, NY, for Objectors‐Appellants Discover Bank, et al.

JASON A. YURASEK (Anahit Samarjian, on the brief), Perkins Coie LLP, San Francisco, CA, for Objectors‐Appellants First Data Corporation, et al.

Andrew G. Celli, Jr. and Debra L. Greenberger, Emery Celli Brinckerhoff & Abady LLP, New York, NY, for Objectors‐Appellants (Merchant Trade Groups).

Jerrold S. Parker and Jay L.T. Breakstone, Parker Waichman, LLP, Port Washington, NY; Thomas P. Thrash and Marcus N. Bozeman, Thrash Law Firm, P.A., Little Rock, AR; Phillip Duncan and Richard Quintus, Duncan Firm, P.A., Little Rock, AR, for Appellant Retailers and Merchants Objectors.

3

Elizabeth Wolstein, Schlam Stone & Dolan LLP, New York, NY, for Objectors‐Appellants U.S. PIRG and Consumer Reports.

Anthony F. Shelley, Adam P. Feinberg, Laura G. Ferguson, Michael N. Khalil, and Katherine E. Pappas, Miller & Chevalier Chartered, Washington, DC, for Appellants Blue Cross and Blue Shield Entities and Wellpoint Entities.

Steve A. Miller, Denver, CO, for Appellant The Iron Barley Restaurant LLC.

John J. Pentz, Sudbury, MA, for Appellants Unlimited Vacations and Cruises, Inc., et al.

N. Albert Bacharach, Jr., Gainesville, FL, for Appellant Optical Etc. LLC.

Christopher A. Bandas, Corpus Christi, TX, for Objectors‐Appellants 1001 Property Solutions, LLC, et al.

PAUL D. CLEMENT (Jeffrey M. Harris and Candice C. Wong, on the brief), Bancroft PLLC, Washington, DC; K. Craig, Wildfang, Thomas J. Undlin, Ryan W. Marth, and Bernard Persky, Robins Kaplan, Miller & Ciresi L.L.P., Minneapolis, MN; H. Laddie Montague, Merrill G. Davidoff, and Michael J. Kane, Berger & Montague, P.C., Philadelphia, PA; Bonny E. Sweeney,

4 Joseph D. Daley, and Alexandra S. Bernay, Robbins Geller Rudman & Dowd LLP, San Diego, CA; Joseph Goldberg, Freedman Boyd Goldberg Urias & Ward, P.A., Albuquerque, MN, for Plaintiffs‐Appellees.

CARTER G. PHILIPS (David F. Graham, Robert N. Hochman, Benjamin R. Nagin, Eamon P. Joyce, and Mark D. Taticchi, on the brief), Sidley Austin LLP, Washington, DC; Robert C. Mason, Arnold & Porter LLP, New York, NY; Robert J. Vizas, Arnold & Porter LLP, San Francisco, CA; Mark R. Merley and Matthew A. Eisenstein, Arnold & Porter LLP, Washington, DC; Richard J. Holwell, Michael S. Shuster, and Demian Ordway, Holwell Shuster & Goldberg LLP, New York, NY; Matthew Freimuth and Wesley R. Powell, Willkie Farr & Gallagher LLP, New York, NY; Kenneth A. Gallo, Paul, Weiss, Rifkind, Wharton & Garrison LLP, Washington, DC; Mark P. Ladner and Michael B. Miller, Morrison & Foerster LLP, New York, NY; Andrew J. Frackman and Abby F. Rudzin, O=Melveny & Myers LLP; James P. Tallon, Shearman & Sterling LLP, New York, NY; Richard L. Creighton and Drew M. Hicks, Keating Muething & Klekamp PLL, Cincinnati, OH; John P. Passarelli and James M. Sulentic, Kutak Rock LLP, Omaha, NE; Peter E. Greene, Boris Bershteyn, and Peter S. Julian, Skadden, Arps, Slate, Meagher & Flom LLP, New York, NY; Jonathan S. Massey and Leonard A. Gail, Massey & Gail,

5 Washington, DC; Ali M. Stoeppelweth, Wilmer Cutler Pickering Hale and Dorr LLP, Washington, DC; John M. Majoras and Joseph W. Clark, Jones Day, Washington, DC; Teresa T. Bonder, Valarie C. Williams, and Kara F. Kennedy, Alston & Bird LLP, Atlanta, GA; Jonathan B. Orleans and Adam S. Mocciolo, Pullman & Comley, LLC, Bridgeport, CT; Robert P. LoBue and William F. Cavanaugh, Patterson Belknap Webb & Tyler LLP, New York, NY, for Defendants‐Appellees.

DENNIS JACOBS, Circuit Judge:

This antitrust class action was brought on behalf of approximately 12

million merchants against Visa U.S.A. Inc. (“Visa”) and MasterCard International

Incorporated (“MasterCard”), which are the two largest credit card issuing

networks in the United States, as well as against various issuing and acquiring

banks (collectively with Visa and MasterCard, the “defendants”), alleging a

conspiracy in violation of Section 1 of the Sherman Act. After nearly ten years of

litigation, the parties agreed to a settlement that released all claims in exchange

for disparate relief for each of two classes: up to $7.25 billion would go to an

opt‐out class, and a non‐opt‐out class would get injunctive relief. The district

court certified these two settlement‐only classes, and approved the settlement as

6 fair and reasonable. On this appeal, numerous objectors and opt‐out plaintiffs

argue that this class action was improperly certified and that the settlement was

unreasonable and inadequate. We conclude that the class plaintiffs were

inadequately represented in violation of Rule 23(a)(4) and the Due Process

Clause. Accordingly, we vacate the district court’s certification of this class

action and reverse the approval of the settlement.

BACKGROUND

Detailed information about how the credit card industry operates is set out

in the district court opinion approving the settlement in this case, In re Payment

Card Interchange Fee and Merchant Discount Antitrust Litig. (“Payment Card I”),

, 214‐15 (E.D.N.Y. 2013), and in our previous opinions dealing

with past antitrust lawsuits against Visa and MasterCard, Wal‐Mart Stores, Inc. v.

Visa U.S.A., Inc.,

, 101‐02 (2d Cir. 2005); United States v. Visa U.S.A.,

Inc.,

, 234‐37 (2d Cir. 2003); In re Visa Check/MasterMoney Antitrust

Litig. (“Visa Check”),

, 129‐31 (2d Cir. 2001). This section of the

opinion lays out only the facts and procedural history needed to explain our

analysis and result.

7 In general terms, a Visa or MasterCard credit card transaction is processed

as follows: the customer presents a credit card to pay for goods or services to the

merchant; the merchant relays the transaction information to the acquiring bank;

the acquiring bank processes the information and relays it to the network (here,

Visa or MasterCard); the network relays the information to the issuing bank; if the

issuing bank approves the transaction, that approval is relayed to the acquiring

bank, which then relays it to the merchant. If the transaction is approved, the

merchant receives the purchase price minus two fees: the “interchange fee” that

the issuing bank charged the acquiring bank and the “merchant discount fee” that

the acquiring bank charged the merchant.

In a given transaction, the interchange fee that the acquiring bank pays

(and is in turn paid by the merchant) varies depending on the credit card network

and the type of credit card. Thus, the American Express credit‐card network

generally charges a higher interchange fee than the Visa or MasterCard networks.

And Visa and MasterCard have different product levels within their credit card

portfolios, such as cards that give consumers generous rewards, and typically

charge a higher interchange fee than cards that offer few rewards or none. The

difference in interchange fee between American Express and Visa or MasterCard

8 is one at the brand level, while the difference between, e.g., a rewards card from

Visa and a no‐rewards card from Visa is one at the product level.

Plaintiffs are all merchants who accept Visa‐ and MasterCard‐branded

credit cards and are therefore bound by the issuers’ network rules. Plaintiffs

challenge as anti‐competitive several of the following network rules (which are

effectively identical as between Visa and MasterCard). The “default

interchange” fee applies to every transaction on the network (unless the merchant

and issuing bank have entered into a separate agreement). The

“honor‐all‐cards” rule requires merchants to accept all Visa or MasterCard credit

cards if they accept any of them, regardless of the differences in interchange fees.

Multiple rules prohibit merchants from influencing customers to use one type of

payment over another, such as cash rather than credit, or a credit card with a

lower interchange fee. These “anti‐steering” rules include the “no‐surcharge”

and “no‐discount” rules, which prohibit merchants from charging different prices

at the point of sale depending on the means of payment.

Plaintiffs allege that these Visa and MasterCard network rules, working in

tandem, allow the issuing banks to impose an artificially inflated interchange fee

that merchants have little choice but to accept. The argument is that the

9 honor‐all‐cards rule forces merchants to accept all Visa and MasterCard credit

cards (few merchants can afford to accept none of them); the anti‐steering rules

prohibit them from nudging consumers toward cheaper forms of payment; the

issuing banks are thus free to set interchange fees at a supra‐competitive rate; and

that rate is effectively locked in via the default interchange fee because the issuing

banks have little incentive to deviate from it unless a given merchant is huge

enough to have substantial bargaining power.

The first consolidated complaint in this action was filed in 2006.

Developments since then have altered the credit card industry in important ways.

Both Visa and MasterCard conducted initial public offerings that converted each

from a consortium of competitor banks into an independent, publicly traded

company. The “Durbin Amendment” to the Dodd‐Frank Wall Street Reform

and Consumer Protection Act of 2010 limited the interchange fee that issuing

banks could charge for debit card purchases, and allowed merchants to discount

debit card purchases relative to credit card purchases. Finally, pursuant to a

consent decree with the Department of Justice in 2011, Visa and MasterCard

agreed to permit merchants to discount transactions to steer consumers away

10 from credit cards use. None of these developments affected the honor‐all‐cards

or no‐surcharging rules, or the existence of a default interchange fee.

Notwithstanding these pro‐merchant industry developments, the plaintiffs

pressed on. Discovery included more than 400 depositions, 17 expert reports, 32

days of expert deposition testimony, and the production of over 80 million pages

of documents. The parties fully briefed a motion for class certification, a motion

to dismiss supplemental complaints, and cross‐motions for summary judgment.

Beginning in 2008, the parties participated in concurrent settlement negotiations

assisted by well‐respected mediators. At the end of 2011, the district judge and

the magistrate judge participated in the parties’ discussions with the mediators.

In October 2012, after several more marathon negotiations with the mediators

(including one more with the district court and magistrate judges), the parties

executed the Settlement Agreement. The district court granted preliminary

approval of the proposed settlement on November 27, 2012, and final approval on

December 13, 2013. Payment Card I,

.

The Settlement Agreement divides the plaintiffs into two classes: one – the

Rule 23(b)(3) class – covers merchants that accepted Visa and/or MasterCard from

January 1, 2004 to November 28, 2012; the other – the Rule 23(b)(2) class – covers

11 merchants that accepted (or will accept) Visa and/or MasterCard from November

28, 2012 onwards forever. The former class would be eligible to receive up to

$7.25 billion in monetary relief; the latter would get injunctive relief in the form of

changes to Visa=s and MasterCard’s network rules. Because of the difference

between Rule 23(b)(3) and Rule 23(b)(2), members of the first class (which

receives money damages in the settlement) could opt out, but members of the

second, forward‐looking class (which receives only injunctive relief) could not.

The most consequential relief afforded the (b)(2) class was the ability to

surcharge Visa‐ and MasterCard‐branded credit cards at both the brand and

product levels. That is, a merchant could increase the price of a good at the point

of sale if a consumer presents (for example) a Visa card instead of cash, or a Visa

rewards card instead of a Visa card that yields no rewards. The incremental

value and utility of this relief is limited, however, because many states, including

New York, California, and Texas, prohibit surcharging as a matter of state law.

See, e.g., Expressions Hair Design v. Schneiderman,

(2d Cir.

2015) (upholding the New York ban on credit‐card surcharges); Rowell v.

Pettijohn,

(5th Cir. 2016) (upholding the Texas ban on credit‐card

surcharges). But see Dana’s R.R. Supply v. Attorney Gen., Florida,

,

12 1249 (11th Cir. 2015) (striking down Florida ban on credit‐card surcharges).

Moreover, under the most‐favored‐nation clause included in the Settlement

Agreement, merchants that accept American Express cannot avail themselves of

the surcharging relief because American Express effectively prohibits

surcharging, and the Settlement Agreement permits surcharging for Visa or

MasterCard only if the merchant also surcharges for use of cards issued by

competitors such as American Express.

Visa and MasterCard also agreed to modify their network rules to reflect

that they will: negotiate interchange fees with groups of merchants in good faith,

lock‐in the benefits of the Durbin Amendment and Department of Justice consent

decree, and permit a merchant that operates multiple businesses under different

names or banners to accept Visa or MasterCard at fewer than all of its businesses.

The Settlement Agreement provides that all of the injunctive relief will

terminate on July 20, 2021.

In return, the plaintiffs are bound by a release that waives any claims they

would have against the defendants for: all of the conduct challenged in the

operative complaint, all other policies and practices (concerning credit card

transactions) that were in place as of November 27, 2012, and any substantially

13 similar practices they adopt in the future. While the injunctive relief for the

(b)(2) class will expire on July 20, 2021, this release has no end date. It operates

in perpetuity, provided only that Visa and MasterCard keep in place the several

rules that were modified by the injunctive relief provided to the (b)(2) class

(including, inter alia, permitting merchants to surcharge), or impose rules that are

substantially similar to the modified rules. That is, after July 20, 2021, for as long

as Visa and MasterCard elect to leave in place their network rules as modified by

the Settlement Agreement or adopt rules substantially similar thereto, the

defendants continue to enjoy the benefit of the release as to all claims the plaintiffs

potentially had against the defendants for any of the network rules existing as of

November 27, 2012.

If, after July 20, 2021, the Visa or MasterCard networks rules are changed

such that they are no longer substantially similar to their form as modified by the

Settlement Agreement, then merchants are freed from the release as to claims

arising out of that new network rule – but only as to such claims. For example, if

Visa or MasterCard revert to their pre‐Settlement Agreement rules by forbidding

merchants from surcharging, then the release will not bar future merchants

included in the (b)(2) class from bringing antitrust claims arising out of the

14 prohibition on surcharging; but the rest of release would remain in effect, so that a

suit by the future plaintiff could not challenge any of the unchanged network

rules, such as the honor‐all‐cards rule or imposition of default interchange fees.

In sum, regardless what Visa or MasterCard do with their network rules after July

20, 2021, no merchant will ever be permitted to bring claims arising out of the

network rules that are unaffected by this Settlement Agreement, including most

importantly, the honor‐all‐cards rule or existence of default interchange fees.

Appellants, including those that opted out from the (b)(3) class and

objected to the (b)(2) class, argue that the (b)(2) class was improperly certified and

that the settlement was inadequate and unreasonable.

DISCUSSION

Certification of a class is reviewed for abuse of discretion, i.e., whether the

decision (i) rests on a legal error or clearly erroneous factual finding, or (ii) falls

outside the range of permissible decisions. In re Literary Works in Elec.

Databases Copyright Litig. (“Literary Works”),

(2d Cir. 2011).

The district court’s factual findings are reviewed for clear error; its conclusions of

law are reviewed de novo. Charron v. Wiener,

(2d Cir. 2013).

15 Class actions are an exception to the rule that only the named parties

conduct and are bound by litigation. See Hansberry v. Lee,

, 40‐41

(1940). “In order to justify a departure from that rule, a class representative must

be part of the class and possess the same interest and suffer the same injury as the

class members.” Wal‐Mart v. Dukes,

(2011) (internal

quotation marks and citations omitted). That principle is secured by Rule

23(a)(4) and the Due Process Clause. Rule 23(a)(4), which requires that “the

representative parties . . . fairly and adequately protect the interests of the class,”

“serves to uncover conflicts of interest between named parties and the class they

seek to represent,” as well as the “competency and conflicts of class counsel.”

Amchem Prods., Inc. v. Windsor,

, 626 n.20 (1997). “[T]he Due

Process Clause of course requires that the named plaintiff at all times adequately

represent the interests of the absent class members.” Phillips Petroleum Co. v.

Shutts,

(1985). Class actions and settlements that do not comply

with Rule 23(a)(4) and the Due Process Clause cannot be sustained.

We conclude that class members of the (b)(2) class were inadequately

represented in violation of both Rule 23(a)(4) and the Due Process Clause.

Procedural deficiencies produced substantive shortcomings in this class action

16 and the settlement. As a result, this class action was improperly certified and the

settlement was unreasonable and inadequate.

I

Under Rule 23(a)(4), “[a]dequacy is twofold: the proposed class

representative must have an interest in vigorously pursuing the claims of the

class, and must have no interests antagonistic to the interests of other class

members.” Denney v. Deutsche Bank AG,

(2d Cir. 2006); see

also Robinson v. Metro‐North Commuter R.R. Co.,

(2d Cir. 2001)

(“Two factors generally inform whether class representatives satisfy the Rule

23(a)(4) requirement: ‘(1) absence of conflict and (2) assurance of vigorous

prosecution.’”) To assure vigorous prosecution, courts consider whether the class

representative has adequate incentive to pursue the class’s claim, and whether

some difference between the class representative and some class members might

undermine that incentive.

. To avoid antagonistic interests, any

“fundamental” conflict that goes “to the very heart of the litigation,” Charron, 731

F.3d at 249‐50 (internal citations omitted), must be addressed with a “structural

assurance of fair and adequate representation for the diverse groups and

individuals” among the plaintiffs. Amchem,

. One common

17 structural protection is division of the class into “homogenous subclasses under

Rule 23(c)(4)(B), with separate representation to eliminate conflicting interests of

counsel.” Ortiz v. Fibreboard Corp.,

(1999).

“Adequacy must be determined independently of the general fairness

review of the settlement; the fact that the settlement may have overall benefits for

all class members is not the ‘focus’ in ‘the determination whether proposed

classes are sufficiently cohesive to warrant adjudication.’” Denney,

(quoting Ortiz,

). The focus of the Rule 23(a) inquiry remains

on “inequity and potential inequity at the precertification stage.” Ortiz,

. So when (as here) the district court certifies the class at the same time it

approves a settlement, the requirements of Rule 23(a) “demand undiluted, even

heightened, attention.” Amchem,

.

A

The Supreme Court wrote the ground rules for adequate representation in

the settlement‐only class context in Amchem and Ortiz, two asbestos cases. Our

recent decision in Literary Works contributed a gloss on the subject.

The single‐class proposed settlement in Amchem potentially encompassed

millions of plaintiffs who had been exposed to asbestos, without distinction

18 between those who had already manifested asbestos‐related injuries and sought

“generous immediate payments,” and those who had not manifested injury and

sought “an ample, inflation‐protected fund for the future.” Amchem,

. A single class representative could not adequately represent both interests.

The two subgroups had “competing interests in the distribution of a settlement

whose terms reflected ‘essential allocation decisions designed to confine

compensation and to limit defendants’ liability.’” Literary Works,

(quoting Amchem,

). The antagonistic interests were so

pronounced, on an issue so crucial, that the settlement required a “structural

assurance of fair and adequate representation for the diverse groups and

individuals.” Amchem,

.

Two years later, the Supreme Court again considered a settlement‐only

class action that joined present and future claimants in a single class, and

emphasized: “it is obvious after Amchem that a class divided between holders of

present and future claims . . . requires division into homogenous subclasses

under Rule 23(c)(4)(B), with separate representation to eliminate conflicting

interests of counsel.” Ortiz,

. A second fatal deficiency in the

Ortiz settlement was that all present claimants were treated equally,

19 notwithstanding that some had claims that were more valuable. “It is no answer

to say . . . that these conflicts may be ignored because the settlement makes no

disparate allocation of resources as between the conflicting classes” for the “very

decision to treat them all the same is itself an allocation decision with results

almost certainly different from the results that [the disparate claimants] would

have chosen.”

. These fault lines between present and future

plaintiffs, and among plaintiffs with differently valued claims, were so

fundamental that they required “structural protection” in the form of subclasses

with separate counsel.

Literary Works contained the same “ingredients of conflict identified in

Amchem and Ortiz.” Literary Works,

. The settlement divided

class claims into three categories, capped defendants’ overall liability at $18

million, and used a formula for splitting this amount. The settlement was less

generous to the third category, and required the holders of those claims to

exclusively bear the risk of over‐subscription, i.e., their recovery alone would be

reduced to bring the total payout down to $18 million. The class representatives

of the single class included individuals with claims in each category; nevertheless,

we held that (at a minimum) class members with claims only in the third category

20 required separate representation because their interests were antagonistic to the

others on a matter of critical importance – how the money would be distributed.

.

Since some named representatives held claims across all three categories,

the class did not encompass mutually exclusive groups as in Amchem; still, each

impermissibly “served generally as representative for the whole, not for a

separate constituency.”

at 251 (quoting Amchem,

). Class

representatives with claims in all three categories naturally would want to

maximize their overall recovery regardless of allotment across categories,

whereas class members with claims only in the third category would want to

maximize the compensation for that category in particular. A great risk thus

arose that class representatives would sell out the third category of claims for

terms that would tilt toward the others. As it transpired, the resulting settlement

awarded the third category less, and taxed that lesser recovery with all the risk

that claim would exceed the liability cap.

We did not conclude that the third category’s “inferior recovery [w]as

determinative evidence of inadequate representation.” Id. at 253. The claims in

third category were objectively the weakest. “The problem, of course, [wa]s that

21 we ha[d] no basis for assessing whether the discount applied to Category C’s

recovery appropriately reflect[ed] that weakness.” Id. We could not know the

right value of the category C claims “without independent counsel pressing its

most compelling case.” Id. While the settlement “was the product of an

intense, protected, adversarial mediation, involving multiple parties,” including

“highly respected and capable” mediators and associational plaintiffs, these

features of the negotiation could not “compensate for the absence of independent

representation” because there could be no assurance that anyone “advanced the

strongest arguments in favor” of the disfavored claims. Id. at 252‐53. The

eventual settlement proved that “[o]nly the creation of subclasses, and the

advocacy of an attorney representing each subclass, can ensure that the interests

of that particular subgroup are in fact adequately represented.” Id. at 252.

Divided loyalties are rarely divided down the middle.

B

Like the settlement‐only classes in Amchem, Ortiz, and Literary Works, the

unitary representation of these plaintiffs was inadequate. Class representatives

had interests antagonistic to those of some of the class members they were

representing. The fault lines were glaring as to matters of fundamental

22 importance. Such conflicts and absence of incentive required a sufficient

“structural assurance of fair and adequate representation,” Amchem,

, but none was provided.

The conflict is clear between merchants of the (b)(3) class, which are

pursuing solely monetary relief, and merchants in the (b)(2) class, defined as

those seeking only injunctive relief. The former would want to maximize cash

compensation for past harm, and the latter would want to maximize restraints on

network rules to prevent harm in the future. Amchem tells us that such

divergent interests require separate counsel when it impacts the “essential

allocation decisions” of plaintiffs’ compensation and defendants’ liability.

Amchem,

. The Settlement Agreement does manifest tension on

an “essential allocation decision”: merchants in the (b)(3) class would share in up

to $7.25 billion of damages, while merchants in the (b)(2) class would enjoy the

benefit of some temporary changes to the defendants’ network rules. The same

counsel represented both the (b)(3) and the (b)(2) classes. The class counsel and

class representatives who negotiated and entered into the Settlement Agreement

were in the position to trade diminution of (b)(2) relief for increase of (b)(3) relief.

However, “it is obvious after Amchem that a class divided between holders of

23 present and future claims . . . requires division into homogenous

subclasses . . . with separate representation.” Ortiz,

.

Moreover, many members of the (b)(3) class have little to no interest in the

efficacy of the injunctive relief because they no longer operate, or no longer accept

Visa or MasterCard, or have declining credit card sales. By the same token,

many members of the (b)(2) class have little to no interest in the size of the

damages award because they did not operate or accept Visa or MasterCard before

November 28, 2012, or have growing credit card sales. Unitary representation of

separate classes that claim distinct, competing, and conflicting relief create

unacceptable incentives for counsel to trade benefits to one class for benefits to

the other in order somehow to reach a settlement.

Class counsel stood to gain enormously if they got the deal done. The (up

to) $7.25 billion in relief for the (b)(3) class was the “largest‐ever cash settlement in

an antitrust class action.” Payment Card I,

. For their

services, the district court granted class counsel $544.8 million in fees. In re

Payment Card Interchange Fee and Merchant Discount Antitrust Litig.

(“Payment Card II”),

(E.D.N.Y. 2014). The district court

calculated these fees based on a graduated percentage cut of the (b)(3) class’s

24 recovery; thus counsel got more money for each additional dollar they secured for

the (b)(3) class. But the district court’s calculation of fees explicitly did not rely

on any benefit that would accrue to the (b)(2) class,

at 442 n.4, and class

counsel did not even ask to be compensated based on the size or significance of

the injunctive relief.

The resulting dynamic is the same as in Ortiz. As the

Supreme Court recognized in that case: when “the potential for gigantic fees” is

within counsel’s grasp for representation of one group of plaintiffs, but only if

counsel resolves another group of plaintiffs’ claims, a court cannot assume class

counsel adequately represented the latter group’s interests. Ortiz,

. We expressly do not impugn the motives or acts of class counsel.

Nonetheless, class counsel was charged with an inequitable task.

The trouble with unitary representation here is exacerbated because the

members of the worse‐off (b)(2) class could not opt out. The (b)(2) merchants are

stuck with this deal and this representation. We do not decide whether

providing these class members with opt out rights would be a sufficient

“structural assurance of fair and adequate representation,” Amchem,

, to overcome the lack of separate class counsel and representative. Cf. Visa

25 Check,

. It is enough to say that this feature of the Settlement

Agreement compounded the problem.

One aspect of the Settlement Agreement that emphatically cannot remedy

the inadequate representation is the assistance of judges and mediators in the

bargaining process. True, “a court‐appointed mediator’s involvement in

pre‐certification settlement negotiations helps to ensure that the proceedings

were free of collusion and undue pressure.” D’Amato v. Deutsche Bank,

, 85 (2d Cir. 2001). But even “an intense, protected, adversarial mediation,

involving multiple parties,” including “highly respected and capable” mediators

and associational plaintiffs, does not “compensate for the absence of independent

representation.” Literary Works, 654 F.3d at 252‐53. The mission of mediators

is to bring together the parties and interests that come to them. It is not their role

to advance the strongest arguments in favor of each subset of class members

entitled to separate representation, or to voice the interests of a group for which

no one else is speaking.

Nor is the problem cured by the partial overlap of merchants who get cash

as members of the (b)(3) class and become members of the (b)(2) class as they

continue to accept Visa or MasterCard. The force of Amchem and Ortiz does not

26 depend on the mutually exclusivity of the classes; it was enough that the classes

did not perfectly overlap. We held as much in Literary Works, reasoning that

named plaintiffs with claims in multiple subgroups cannot adequately represent

the interests of any one subgroup because their incentive is to maximize their own

total recovery, rather than the recovery for any single subgroup. Amchem

observed that “where differences among members of a class are such that

subclasses must be established, we know of no authority that permits a court to

approve a settlement . . . on the basis of consents by members of a unitary class,

some of whom happen to be members of the distinct subgroups.” Amchem,

(quoting In re Joint E. and S. Dist. Asbestos Litig.,

, 742‐43

(2d Cir. 1992), modified on reh’g,

(2d Cir. 1993)).

Moreover, whatever overlap presently exists is partial and shrinking with

time. As of the September 12, 2013 fairness hearing, class counsel reported that

the class was composed of about 12 million merchants. That figure of course

does not include merchants that have come into being since then, or those that

will come into being in the future, all of whom will be members of only the (b)(2)

class. The membership of the (b)(3) class, on the other hand, is fixed and finite.

27 Over time, the initial overlap will be reduced, and the gap between the interests of

the (b)(3) and (b)(2) classes will continue to widen.

None of this is to say that (b)(3) and (b)(2) classes cannot be combined in a

single case, or that (b)(3) and (b)(2) classes necessarily and always require

separate representation. Problems arise when the (b)(2) and (b)(3) classes do not

have independent counsel, seek distinct relief, have non‐overlapping

membership, and (importantly) are certified as settlement‐only. The

requirements of Rule 23(a) are applied with added solicitude in the

settlement‐only class context because “the certification of a mandatory settlement

class ‘effectively concludes the proceeding save for the final fairness hearing,’ and

there is thus a heightened risk of conflating the fairness requirements of Rule

23(e) with the independent requirement of ‘rigorous adherence to those

provisions of the Rule designed to protect absentees,’ such as Rules 23(a) and (b).”

Charron,

(quoting Ortiz,

). As in Amchem, Ortiz,

and Literary Works, settlements that are approved simultaneously with class

certification are especially vulnerable to conflicts of interest because the

imperatives of the settlement process, which come to bear on the defendants, the

class counsel, and even the mediators and the court itself, can influence the

28 definition of the classes and the allocation of relief. For this reason, we scrutinize

such settlements more closely.

Of course we have blessed multi‐class settlements that were the product of

unitary representation, but those were entered into after class certification. For

example, we approved a settlement negotiated by unitary counsel in Charron; but

before doing so, we “note[d] that unlike the situation in Amchem, Ortiz, and

Literary Works, the settlement here was not being approved at the same time that

the class was being certified.” Charron,

. Accordingly, we were

more skeptical of allegations that subclass conflicts required separate

representation.

True, Charron observed “[a]ll class settlements value some

claims more highly than others, based on their perceived merits, and strike

compromises based on probabilistic assessments,”

but that observation has

less force in the settlement‐only context. Charron also spoke of counsel trading

one claim for another (which may be permissible); in the settlement‐only class

action, we are concerned that counsel will trade the interests of one class for

another (which is not).

We have reason to think that that occurred here. Structural defects in this

class action created a fundamental conflict between the (b)(3) and (b)(2) classes

29 and sapped class counsel of the incentive to zealously represent the latter.

Apparently, the only unified interests served by herding these competing claims

into one class are the interests served by settlement: (i) the interest of class counsel

in fees, and (ii) the interest of defendants in a bundled group of all possible

claimants who can be precluded by a single payment. This latter interest

highlights the next problem with the Settlement Agreement.

II

This opinion already concludes that class plaintiffs were inadequately

represented. Accordingly, the settlement and release that resulted from this

representation are nullities. See Stephenson v. Dow Chem. Co.,

, 260

(2d Cir. 2001), aff’d in part by an equally divided court and vacated in part,

(2003) (“Res judicata generally applies to bind absent class members

except where to do so would violate due process” and “[d]ue process requires

adequate representation at all times throughout the litigation.”). This outcome is

confirmed by the substance of the deal that was struck. Like the Supreme Court

in Amchem, we “examine a settlement’s substance for evidence of prejudice to the

interests of a subset of plaintiffs” when “assessing the adequacy of

representation.” Literary Works,

. Here, the bargain that was

30 struck between relief and release on behalf of absent class members is so

unreasonable that it evidences inadequate representation.

“It is familiar doctrine of the federal courts that members of a class not

present as parties to the litigation may be bound by the judgment where they are

in fact adequately represented by parties who are present” consistent with “the

requirements of due process and full faith and credit.” Hansberry, 311 U.S. at

42‐43 (emphasis added); see also Stephenson, 273 F.3d at 261 (“Part of the due

process inquiry (and part of the Rule 23(a) class certification requirements)

involves assessing adequacy of representation and intra‐class conflicts.”).

Similarly, “[p]laintiffs in a class action may release claims that were or could have

been pled in exchange for settlement relief”; but this authority “is limited by the

‘identical factual predicate’ and ‘adequacy of representation’ doctrines.”

Wal‐Mart Stores,

. “[W]here class plaintiffs have not adequately

represented the interests of class members,” any “[c]laims arising from a shared

set of facts will not be precluded.”

.

A

As discussed above, Literary Works concluded that inadequate

representation was demonstrated by the relief afforded to a subset of the class.

31 Similarly, the release in Stephenson was itself proof of inadequate representation,

wheras the release in Wal‐Mart Stores did not impugn the class’s representation.

Considered together, these cases illustrate when the tradeoff between relief and

release as applied to a class member can violate due process.

Literary Works held that class members with claims in one of the categories

were inadequately represented not only because they did not receive separate

representation, but also because they solely bore the risk that the total amount

claimed would exceed a preset liability cap. We observed that this feature of the

settlement could not be justified by the relative weakness of those claims because

that fact was already accounted for. Literary Works,

. We could

discern no reason for subjecting the single category of claims to the whole risk of

over‐subscription; nor could the settlement’s proponents.

. When

“one category [of class members are] targeted for [worse treatment] without

credible justification” it “strongly suggests a lack of adequate representation for

those class members who hold only claims in this category.”

In Stephenson, we considered a collateral attack on a class action that had

established a settlement fund for individuals injured by exposure to Agent

Orange. The underlying litigation provided compensation only for those who

32 discovered their injury before 1994, yet released all future claims. Two

individuals who fell within the class definition of individuals injured by Agent

Orange, but who learned of their injury after 1994, challenged the release as

applied to them. Analogizing the case to Amchem and Ortiz, we concluded that

the two individuals were inadequately represented in the prior litigation because

the settlement purported to resolve all future claims but “the settlement fund was

permitted to terminate in 1994” and “[n]o provision was made for post‐1994

claimants.” Stephenson, 273 F.3d at 260‐61. The two challengers could not have

been adequately represented if their class representative negotiated a settlement

and release that extinguished their claims without affording them any recovery.

The result violated due process; the plaintiffs could not be bound by the

settlement release.

.

A similar challenge was raised to the settlement release in Wal‐Mart Stores,

which foreclosed all claims arising from the same factual predicate as that alleged

in the complaint. Objectors argued that they were inadequately represented

because class representatives did not pursue certain claims as vigorously as

others. We rejected this basis for objection because “adequate representation of a

particular claim is determined by the alignment of interests of class members, not

33 proof of vigorous pursuit of that claim.” Wal‐Mart Stores,

.

Stephenson was “not directly on point” because in the Agent Orange settlement

(as in the Amchem and Ortiz settlements) “future claims had not been considered

separately from claims involving current injury” despite these two groups having

clearly divergent interests.

. The objectors in Wal‐Mart Stores did not

allege divergent interests; they had disagreements about which claims were most

valuable and what relief was adequate. Moreover, the settlement in Wal‐Mart

Stores covered only a past, finite period and did not preclude future suits over

conduct post‐dating the settlement.

No future claimants or claims were

covered by the Wal‐Mart Stores settlement or release. Finally, every claimant

from the objecting groups benefitted from the settlement.

.

B

Merchants in the (b)(2) class that accept American Express or operate in

states that prohibit surcharging gain no appreciable benefit from the settlement,

and merchants that begin business after July 20, 2021 gain no benefit at all. In

exchange, class counsel forced these merchants to release virtually any claims

they would ever have against the defendants. Those class members that

34 effectively cannot surcharge and those that begin operation after July 20, 2021

were thus denied due process.

No one disputes that the most valuable relief the Settlement Agreement

secures for the (b)(2) class is the ability to surcharge at the point of sale. To the

extent that the injunctive relief has any meaningful value, it comes from

surcharging, not from the buying‐group provision, or the all‐outlets provision, or

the locking‐in of the Durbin Amendment and DOJ consent decree. For this

reason, it is imperative that the (b)(2) class in fact benefit from the right to

surcharge. But that relief is less valuable for any merchant that operates in New

York, California, or Texas (among other states that ban surcharging), or accepts

American Express (whose network rules prohibit surcharging and include a

most‐favored nation clause). Merchants in New York and merchants that accept

American Express can get no advantage from the principal relief their counsel

bargained for them.

It may be argued that the claims of the (b)(2) class are weak and can

command no benefit in settlement. However, that argument would seem to be

foreclosed because other members of the same class with the same claims – those

that do not take American Express and operate in states that permit surcharging –

35 derive a potentially substantial benefit. There is no basis for this unequal

intra‐class treatment: the more valuable the right to surcharge (a point the parties

vigorously dispute), the more unfair the treatment of merchants that cannot avail

themselves of surcharging.

This is not a case of some plaintiffs forgoing settlement relief. A

significant proportion of merchants in the (b)(2) class are either legally or

commercially unable to obtain incremental benefit from the primary relief

negotiated for them by their counsel, and class counsel knew at the time the

Settlement Agreement was entered into that this relief was virtually worthless to

vast numbers of class members. Alternative forms of relief might have conferred

a real and palpable benefit, such as remedies that affected the default interchange

fee or honor‐all‐cards rule. This is not a matter of certain merchants (e.g., those

based in New York and those that accept American Express) arguing that class

counsel did not bargain for their preferred form of relief, did not press certain

claims more forcefully, or did not seek certain changes to the network rule books

more zealously. This is a matter of class counsel trading the claims of many

merchants for relief they cannot use: they actually received nothing.

36 Another fault line within the (b)(2) class runs between merchants that will

have accepted Visa or MasterCard before July 20, 2021, and those that will come

into being thereafter. The former are at least guaranteed some form of relief,

while the latter are at the mercy of the defendants to receive relief because the

Settlement Agreement explicitly states that the defendants’ obligation to provide

any injunctive relief terminates on July 20, 2021. Like the servicemen with latent

injury in Stephenson, the post‐July 20, 2021 merchants are future claimants who

had their claims settled for nothing. There is no evidence to suggest that

merchants operating after July 20, 2021 would have weaker claims than those

operating before July 20, 2021; yet, the Settlement Agreement consigns the former

to an unambiguously inferior position. As in Literary Works, we conclude that

such arbitrary harsher treatment of class members is indicative of inadequate

representation.

Merchants that cannot surcharge, and those that open their doors after July

20, 2021, are also bound to an exceptionally broad release. The Settlement

Agreement releases virtually any claim that (b)(2) class members would have had

against the defendants for any of the defendants’ thousands of network rules.

And unlike the relief, which expires on July 20, 2021, the release operates

37 indefinitely. Therefore, after July 20, 2021, the (b)(2) class remains bound to the

release but is guaranteed nothing. This release permanently immunizes the

defendants from any claims that any plaintiff may have now, or will have in the

future, that arise out of, e.g., the honor‐all‐cards and default interchange rules.

Even if the defendants revert back to all their pre‐Settlement Agreement practices,

the release continues to preclude any claim based on any rule that was not altered

by the Settlement Agreement. The defendants never have to worry about future

antitrust litigation based on their honor‐all‐cards rules and their default

interchange rules.

That is because the only claims that merchants post‐July 20, 2021 may have

are ones relating to those network rules that are explicitly changed by the

injunctive relief in the Settlement Agreement. Those claims will become

actionable only if the defendants elect to revert to their pre‐Settlement Agreement

rules. Of course, it remains to be seen how much the mandated rules will cost

the defendants or benefit the merchants, but either way, the defendants win. If

the defendants see that permitting surcharging had little effect on their business,

they can decide to maintain the rules changes provided for in the injunctive relief

so that only merchants that do not accept American Express and do not operate in

38 states like New York, California, and Texas will be able to avail themselves of that

limited relief. On the other hand, if the defendants observe that surcharging

took a significant toll on their business, they can revert to prohibiting surcharging

and expose themselves to lawsuits that are limited to challenging the surcharging

ban. In all events, merchants that cannot surcharge receive valueless relief while

releasing a host of claims of unknown value.

This bargain is particularly unreasonable for merchants that begin

accepting Visa or MasterCard after July 20, 2021. They will be deemed to have

released all of their claims pertaining to a whole book of rules, including (perhaps

most importantly) the honor‐all‐cards and default interchange rules, and in

return have the chance that the defendants will permit surcharging. In substance

and effect, merchants operating after July 20, 2021 give up claims of potential

value and receive nothing that they would not otherwise have gotten. Since

there was no independent representation vigorously asserting these merchants’

interests, we have no way to ascertain the value of the claims forgone. See

Literary Works,

.

In sum, this release has much in common with the releases in Stephenson,

Amchem, and Ortiz. Like those, this release applies to future claims and

39 claimants, and disadvantaged class members are bound to it. The Settlement

Agreement waives any claim any (b)(2) merchant would have against any

defendant arising out of any of the current network rules, or those imposed in the

future that are substantially similar thereto. The (b)(2) class had no notice and no

opportunity to opt out of this deal. (At least the authors in Literary Works could

opt out from their inadequate representation.) This Settlement Agreement is

also distinguishable from releases that have passed muster. For example, the

settlement release in Wal‐Mart Stores (another merchant class action against Visa

and MasterCard) did not bind future claimants and did not preclude new suits for

similar conduct in the future. Wal‐Mart Stores,

. And our

approval of the Charron settlement release explicitly distinguished it from those

in Amchem, Ortiz, and Literary Works on the ground that it did not extinguish

claims other than those that were the subject of relief in the settlement. Charron,

.

Merchants that cannot surcharge (by reason of state law or rules of

American Express) and those that begin operating after July 20, 2021 suffer an

unreasonable tradeoff between relief and release that demonstrates their

representation did not comply with due process. We of course acknowledge that

40 “[b]road class action settlements are common, since defendants and their cohorts

would otherwise face nearly limitless liability from related lawsuits in

jurisdictions throughout the country.” Wal‐Mart Stores,

. And it

is true that “[p]arties often reach broad settlement agreements encompassing

claims not presented in the complaint in order to achieve comprehensive

settlement of class actions, particularly when a defendant’s ability to limit his

future liability is an important factor in his willingness to settle.” Literary

Works, 654 F.3d at 247‐48. But the benefits of litigation peace do not outweigh

class members’ due process right to adequate representation.

CONCLUSION

For the foregoing reasons, we vacate the district court’s certification of the

class, reverse approval of the settlement, and remand for further proceedings not

inconsistent with this opinion.

41 LEVAL, Circuit Judge, concurring:

I concur in Judge Jacobs’s thoughtful opinion. I write separately, however,

to note another, perhaps deeper, problem with the settlement. Under its terms,

one class of Plaintiffs accepts substantial payments from the Defendants, in return

for which they compel Plaintiffs in another class, who receive no part of the

Defendants’ payments, to give up forever their potentially valid claims, without

ever having an opportunity to reject the settlement by opting out of the class.

Opinions of the Supreme Court directly hold that this arrangement violates the

due process rights of those compelled to surrender their claims for money

damages.

Representatives brought this class action on behalf of approximately 12

million merchants against Visa and MasterCard, alleging that a number of the

Defendants’ practices violate the antitrust laws, and seeking both damages for

past injury and an injunction barring future violations. Eventually, the

Defendants reached a proposed settlement with the Representatives. The

settlement provides that the Defendants would pay approximately $ 7.25 billion

to compensate merchants for damages suffered up to November 28, 2012 (when

the district court granted preliminary approval of the settlement). The settlement

1 also entails a commitment by the Defendants, enforced by injunction, to abandon

some (not all) of their challenged practices for nine years—until July 20, 2021. The

Defendants would be free after that date to resume the practices they temporarily

abandoned and would also be free from the outset to continue forever the

challenged practices they did not agree to abandon. In return for what the

Defendants gave up, a class consisting of all merchants that would ever in the

future accept Visa and MasterCard is compelled to release forever the Defendants

from any and all claims for past or future conduct (other than the conduct

enjoined) that relate in any way to any of Defendants’ practices that are alleged or

could have been alleged in the suit. While I do not speculate on the merits of the

Plaintiffs’ claims, the fact that the Defendants were willing to pay $7.25 billion,

apparently the largest antitrust cash settlement in history, suggests that the claims

were not entirely devoid of merit.

What is particularly troublesome is that the broad release of the Defendants

binds not only members of the Plaintiff class who receive compensation as part of

the deal, but also binds in perpetuity, without opportunity to reject the

settlement, all merchants who in the future will accept Visa and MasterCard,

including those not yet in existence, who will never receive any part of the

2 money. This is not a settlement; it is a confiscation. No merchants operating from

November 28, 2012, until the end of time will ever be allowed to sue the

Defendants, either for damages or for an injunction, complaining of any conduct

(other than that enjoined) that could have been alleged in the present suit. The

future merchants are barred by the court’s adoption of the terms of the settlement

from suing for relief from allegedly illegal conduct, although they have no ability

to elect not to be bound by it. One class of Plaintiffs receives money as

compensation for the Defendants’ arguable past violations, and in return gives up

the future rights of others. The Supreme Court has addressed such circumstances

and ruled that an adjudication coming to this result is impermissible.

In Phillips Petroleum Co. v. Shutts,

(1985), the Supreme Court

reasoned that a claim for money damages—a “chose in action”— is “a

constitutionally recognized property interest possessed by each of the plaintiffs”

whose claims are represented in a class action.

. In order for a court “to

bind an absent plaintiff concerning a claim for money damages or similar relief at

law, it must provide minimal procedural due process protection. . . . [D]ue

process requires at a minimum that an absent plaintiff be provided with an

3 opportunity to remove himself from the class . . . .”

at 811–12. That opportunity

was lacking here.

Following Shutts, the Court unanimously held in Wal‐Mart Stores, Inc. v.

Dukes,

(2011), that claims for monetary relief cannot be

certified under Rule 23(b)(2), as here, because of the possibility that “individual

class members’ compensatory‐damages claims would be precluded by litigation

they had no power to hold themselves apart from.”

(emphasis added).

Dukes did not involve a settlement agreement, but that does not make its

precedent any less applicable to this case. If a class may not even be certified

because of the risk that adjudication of its rights might violate the due process

rights of its members by forcibly depriving them of claims, then necessarily an

adjudication of a class’s rights that in fact forcibly deprives the members of their

claims is also unacceptable. Because the terms of this settlement preclude all

future merchants that will accept the Defendants’ cards (the (b)(2) class) from

bringing claims without their having had an opportunity to opt out (or even

object), the Supreme Court’s rulings in Shutts and Dukes make clear that a court

cannot accept it.

4 The practical effects of this settlement underscore why this is so. Although

no court will ever have ruled that the Defendants’ practices are lawful, no person

or entity will ever have the legal right to sue to challenge those practices, and no

person or entity, past, present, or future has had or will have the opportunity to

refuse to be a part of the class so bound. For this reason, as well as those noted in

Judge Jacobs’s opinion, we must reject the settlement.

5