Shiyang Huang v. Equifax Inc.

999 F.3d 1247

Opinion

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT

________________________

No. 20-10249

________________________

D.C. Docket No. 1:17-md-02800-TWT In re Equifax Inc. Customer Data Security Breach Litigation

SHIYANG HUANG, et al., Movants-Appellants,

BRIAN F. SPECTOR, et al., Plaintiffs-Appellees,

versus

EQUIFAX INC., et al.,

Defendants-Appellees.

________________________

Appeals from the United States District Court for the Northern District of Georgia

________________________

(June 3, 2021)

Before MARTIN, GRANT, and BRASHER, Circuit Judges. MARTIN, Circuit Judge:

This appeal arises from the 2017 data privacy breach of Equifax Inc. and its affiliates (collectively “Equifax”). After the breach came to light, scores of class actions against Equifax flooded the courts. The cases were consolidated in the Northern District of Georgia, where Plaintiffs and Equifax eventually settled their dispute, resulting in “the largest and most comprehensive recovery in a data breach case in U.S. history by several orders of magnitude.” But try as they might, the parties could not please everyone. Of the approximately 147 million class members, 388 people objected to the settlement. Even so, the District Court approved the settlement, certified the settlement class, awarded attorney’s fees and expenses, and approved incentive awards for the class representatives. Several of the objectors appealed, challenging the District Court’s approval order as well as some related rulings.

This case highlights the role objectors play in the settlement of class actions.

We begin with the knowledge that settlements are “highly favored in the law” because “they are a means of amicably resolving doubts and uncertainties and preventing lawsuits.” In re Nissan Motor Corp. Antitrust Litig., 552 F.2d 1088, 1105 (5th Cir. 1977) (quotation marks omitted).1 The settlement here is a prime example. Absent the settlement, the class action could have faced serious hurdles

1 In Bonner v. City of Prichard, 661 F.2d 1206 (11th Cir. 1981) (en banc), we adopted as binding precedent all decisions of the former Fifth Circuit handed down before October 1, 1981. Id. at 1209.

3

to recovery, and now the class is entitled to significant settlement benefits that may not have even been achieved at trial. And you need not take our word for this. The Federal Trade Commission, the Consumer Financial Protection Bureau, and the Attorneys General for 48 states, the District of Columbia, and Puerto Rico all support the settlement.

Yet as we mentioned, not everyone bound by this class action settlement agrees with it, and class members who oppose the settlement have the right to object. See Fed. R. Civ. P. 23(e)(5)(A). Often times objectors play a “beneficial role in opening a proposed settlement to scrutiny and identifying areas that need improvement.” David F. Herr, Annotated Manual for Complex Litigation § 21.643 (4th ed. 2021) [hereinafter “Manual for Complex Litigation”]. And because objectors have the right to object, it is our obligation to closely review the issues they present. Consistent with our obligation, we have studied the hundreds of pages of briefing, sifted through the flurry of Rule 28(j) letters, and familiarized ourselves with the enormous record in this case. After this careful consideration, and with the benefit of oral argument, we affirm the District Court’s rulings in full, subject to one small asterisk. Specifically, after the District Court approved incentive awards for the class representatives, a panel of this Court held that such awards are prohibited. See Johnson v. NPAS Sols., LLC, 975 F.3d 1244, 1260 (11th Cir. 2020). As in NPAS Solutions, we must reverse the District Court’s

4

ruling on the incentive awards alone and remand this case to the District Court solely for the limited purpose of vacating those awards. See id.

I. BACKGROUND

In 2017, Equifax, a consumer reporting agency, announced it had been subject to a data privacy breach affecting the personal information of almost 150 million Americans. The breach involved some of the most sensitive personal information possible: all nine digits of Americans’ Social Security numbers, coupled with their names, dates of birth, and addresses, among other things. Over 300 class actions against Equifax were filed across the nation, all of which came to be consolidated and transferred by the Judicial Panel on Multidistrict Litigation to then-Chief Judge Thomas W. Thrash in the Northern District of Georgia.2 The District Court established separate tracks for the consumer claims and the financial institution claims. This appeal relates to the consumer claims.

In 2018, Plaintiffs filed a 559-page consolidated class action complaint against Equifax. The complaint included 96 named plaintiffs who brought a host of statutory and common law claims under federal and state law. These claims included violations of the Fair Credit Reporting Act, the Georgia Fair Business Practices Act, and various state consumer protection and data breach statutes.

2 Chief Judge Thrash ended his service as Chief Judge for the Northern District of Georgia earlier this year. For consistency, we refer to him by his former title.

5

Plaintiffs also brought claims for negligence, negligence per se, unjust enrichment, and breach of contract. Plaintiffs alleged that, due to the data breach, they are “subject to a pervasive, substantial and imminent risk of identity theft and fraud.” They also alleged that they have spent time, money, and effort attempting to mitigate the risk of identity theft and that many have already been victims of identity theft.

Equifax filed a motion to dismiss the complaint in its entirety, which the District Court granted in part and denied in part. The District Court dismissed the Fair Credit Reporting Act claims, the Georgia Fair Business Practices Act claims, as well as some state statutory claims. However, it allowed the negligence and negligence per se claims under Georgia law, as well as other state statutory claims, to go forward. All the while, the parties engaged in robust settlement negotiations. Layn Phillips, a retired federal district court judge with experience in data breach cases, served as the mediator. The parties’ efforts paid off. After 18 months of negotiations, they reached a settlement agreement. The parties then consulted and negotiated with various federal and state regulators and revised their agreement as a result of those consultations. Ultimately, the Federal Trade Commission, the Consumer Financial Protection Bureau, and the Attorneys General for 48 states, the District of Columbia, and Puerto Rico settled with Equifax, agreeing that the

6

settlement fund in this case provides redress to consumers. In July 2019, the parties presented their final settlement agreement to the District Court.

The District Court described the parties’ settlement as “the largest and most comprehensive recovery in a data breach case in U.S. history by several orders of magnitude.” Under the terms of the settlement, Equifax agreed to pay an initial $380.5 million into a fund to benefit the class members and to pay attorney’s fees and expenses, incentive awards, as well as notice and administration costs. The settlement includes the following benefits for each class member: 3

• Reimbursement for up to $20,000 of documented, out-of-pocket losses fairly traceable to the data breach (e.g., the cost of freezing a credit file, professional fees due to identity theft);

• Compensation of $25 per hour for up to 20 hours (subject to a $38 million cap) for time spent taking preventative measures or dealing with identity theft, with no documentation needed for the first 10 hours;

• Four years of three-bureau credit monitoring and identity protection services through Experian;

• An additional six years of one-bureau credit monitoring and identity protection services through Equifax, which will be provided separately by Equifax and not paid for from the settlement fund;

3 The settlement class includes the “approximately 147 million U.S. consumers identified by Equifax whose personal information was compromised as a result of the cyberattack and data breach announced by Equifax Inc. on September 7, 2017.”

7

• Alternative cash compensation (subject to a $31 million cap) for class members who already have credit monitoring and who do not wish to enroll in the settlement’s programs; 4 and

• Seven years of identity restoration services through Experian to help class members who believe they may have been victims of identity theft.

Beyond these class benefits, Equifax will pay an additional $125 million if needed to satisfy claims for out-of-pocket losses and potentially $2 billion more if all 147 million class members sign up for credit monitoring. In no circumstance does money in the settlement fund revert back to Equifax. Instead, if money remains in the settlement fund after the claim periods, the settlement provides ways in which the above class benefits are increased. Equifax is also required to spend a minimum of $1 billion on data security over five years and to comply with certain data security requirements. Its compliance will be audited by an independent assessor and subject to the District Court’s enforcement powers if it fails to comply.

The District Court ordered that notice of the settlement agreement be provided to the class, such that class members had the opportunity to opt-out of the

4 When the settlement was first announced to the public, media reports said consumers could get $125 in alternative cash compensation under the settlement. The original short-form notice was ambiguous—it simply stated class members “can request” and “may be eligible” for $125 if they already had credit monitoring. However, the long-form notice, which was posted the same day that class members could start making claims, stated in no uncertain terms that consumers who already had credit monitoring could get up to $125, which would be reduced on a proportional basis if the $31 million cap was exceeded. After the media reports, class counsel cleared up this confusion, and those who had already submitted a claim for the alternative cash compensation were given the opportunity to instead choose credit monitoring.

8

class or object to the settlement. The District Court required those who wished to object to provide certain information about their objections in order to prevent a “chaotic” objection process. To provide notice of the settlement to the class, class counsel adopted “an innovative and comprehensive program,” including multiple emails, a social media campaign, newspaper and radio advertising, a settlement website, and a call center to answer questions. The response from the class was “unprecedented,” as the claims rate exceeded 10 percent of the class. By contrast, in another recent data breach case, the claims rate was only about 1.7 percent. As we’ve mentioned, out of the approximately 147 million class members, 388 people objected.

In December 2019, the District Court held a hearing to consider the motions for final approval of the proposed class settlement, attorney’s fees and expenses, and incentive awards for the class representatives. After hearing arguments from Plaintiffs, Equifax, and the objectors who wished to speak, the District Court issued its rulings from the bench. The District Court approved the settlement as fair, reasonable, and adequate under the factors set forth in Federal Rule of Civil Procedure 23(e) and Bennett v. Behring Corp., 737 F.2d 982 (11th Cir. 1984). The District Court then approved the requested attorney’s fees and expenses as well as incentive awards for the class representatives.

9

After issuing its oral rulings, the District Court directed Plaintiffs’ counsel to prepare a written order “summariz[ing] [its] rulings on the motions and [its] adoption basically of the arguments that have been made by the Plaintiffs and by Equifax in the hearing today.” The District Court instructed Plaintiffs to obtain Equifax’s approval before submitting the proposed order to the court, which it would then “consider signing.” The District Court later issued a written order memorializing its rulings. The order approved the settlement; certified the settlement class, finding that the class satisfied the requirements of Rule 23(a) and (b)(3); and approved the requested attorney’s fees and expenses and incentive awards for the class representatives. Finally, the order overruled the objections to the settlement and made findings that some of the objectors were serial objectors.5 Several objectors appealed, and the District Court granted Plaintiffs’ motion to require the objectors to post appeal bonds in order to ensure payment of costs on appeal.

5 Serial objectors are those who bring objections that are merely “boilerplate and immaterial, while their true goal is to get paid some fee to go away.” 4 William B. Rubenstein, Newberg on Class Actions § 13:20 (5th ed. 2021) [hereinafter “Newberg”]. The District Court’s findings on this topic are largely unrelated to the merits of this appeal and may be dicta in any event. We do not review those findings here. See Keating v. City of Miami, 598 F.3d 753, 761 (11th Cir. 2010) (“[A]n appellate court ‘reviews judgments, not statements in opinions.’”).

10

With the dust now settled,6 this consolidated case presents five appeals filed by six objectors: George Cochran, John Davis, Theodore Frank and David Watkins (who filed a single appeal and are collectively referred to as “Mr. Frank”), Shiyang Huang, and Mikell West. Collectively, we refer to the six objectors as the “Objectors.” This is their appeal.

II. DISCUSSION

The Objectors raise a wide array of issues for our consideration. We start by addressing the jurisdictional questions. From there, and in hopes of maintaining some semblance of organization, we proceed in as close to chronological order as this record permits. We begin our discussion of the merits by addressing the Objectors’ challenge to the requirements the District Court imposed on them in its order directing notice of the settlement to the class. We next consider the Objectors’ various challenges to the District Court’s approval order: the process used in adopting the order and the court’s decisions approving the class action settlement, certifying the settlement class, awarding attorney’s fees and expenses, and approving incentive awards for the class representatives. Finally, we address the Objectors’ challenge to the appeal bonds imposed by the District Court.

6 A total of nine objectors appealed the District Court’s orders. Two of those nine objectors filed a single appeal, so eight appeals were filed in this Court. This Court sua sponte dismissed two of the eight appeals for lack of jurisdiction. And in an order issued together with this opinion, we now dismiss the appeal filed by Christopher Andrews, leaving us with five appeals filed by six objectors.

11

A. Jurisdiction We start now with two jurisdictional questions, which we consider de novo.

See Jacobson v. Fla. Sec’y of State, 974 F.3d 1236, 1245 (11th Cir. 2020).7 First, we address whether Plaintiffs had Article III standing to bring their claims. Second, we consider whether Article III’s case-or-controversy requirement ceased to be met once the parties agreed to settle their dispute.

1. Article III Standing 8 In order for a federal court to have jurisdiction under Article III of the Constitution, a plaintiff must have standing to bring the lawsuit. See Lujan v. Defs. of Wildlife, 504 U.S. 555, 559–60, 112 S. Ct. 2130, 2135–36 (1992). And for the plaintiff to have standing, he must “show that the defendant harmed him, and that a court decision can either eliminate the harm or compensate for it.”

7 The parties do not dispute that we have jurisdiction over the Objectors’ appeals. This is for good reason. In Devlin v. Scardelletti, 536 U.S. 1, 122 S. Ct. 2005 (2002), the Supreme Court held that nonnamed class members “who have objected in a timely manner to approval of the settlement at the fairness hearing have the power to bring an appeal without first intervening.” Id. at 14, 122 S. Ct. at 2013. Otherwise, class members would be deprived of “the power to preserve their own interests in a settlement that will ultimately bind them, despite their expressed objections before the trial court.” Id. at 10, 122 S. Ct. at 2011. Although Devlin involved objectors to a Rule 23(b)(1) settlement, which did not permit objectors to opt out of the settlement, its logic also applies to objectors to a Rule 23(b)(3) settlement who did not opt out (like those here) because they are bound by the settlement. See, e.g., Fidel v. Farley, 534 F.3d 508, 512–13 (6th Cir. 2008).

8 Mr. Huang says Plaintiffs were required to prove they had Article III standing with evidentiary support at the final approval stage, yet he says Plaintiffs failed to do so. However, Mr. Huang’s cited cases do not actually support his proposition. In any event, he does not raise any factual doubt about Plaintiffs’ standing, so we need not decide this issue here.

12

Muransky v. Godiva Chocolatier, Inc., 979 F.3d 917, 924 (11th Cir. 2020) (en banc). More to the point, the “irreducible constitutional minimum” of standing contains three requirements. Lujan, 504 U.S. at 560, 112 S. Ct. at 2136. First, the plaintiff must have suffered an “injury in fact,” which means the injury is “concrete and particularized” and “actual or imminent,” as opposed to “conjectural or hypothetical.” Id. (quotation marks omitted). Second, the plaintiff’s injury must be “fairly traceable” to the challenged conduct of the defendant and not the result of some action by a third party not before the court. Id. (quotation marks omitted and alterations adopted). Finally, it must be likely that the plaintiff’s injury will be redressed by a favorable court decision. Id. at 561, 112 S. Ct. at 2136. These requirements apply with full force in a class action, Muransky, 979 F.3d at 924, and even at the settlement approval stage, as a “court is powerless to approve a proposed class settlement if it lacks jurisdiction over the dispute, and federal courts lack jurisdiction if no named plaintiff has standing,” Frank v. Gaos, 586 U.S. __, 139 S. Ct. 1041, 1046 (2019) (per curiam). On the other hand, only one named plaintiff must have standing as to any particular claim in order for it to advance. Wilding v. DNC Servs. Corp., 941 F.3d 1116, 1124–25 (11th Cir. 2019).

Mr. Huang argues Plaintiffs lacked Article III standing to bring their claims for two reasons. First, he says those Plaintiffs who have not had their identities stolen have not suffered an injury in fact. Second, he says those Plaintiffs who

13

have not had their identities stolen cannot have their injuries redressed by the settlement, as the settlement does not stop third parties from committing identity theft. We address each issue in turn.

i. Injury in Fact

We now turn to the question of whether Plaintiffs who have not had their identities stolen suffered an injury in fact. We hold that they have. Again, to establish standing, a plaintiff’s injury must be (1) concrete, (2) particularized, and (3) either actual or imminent. Lujan, 504 U.S. at 560, 112 S. Ct. at 2136. Only the first and third elements are at issue here, so we focus on them in more detail.

An injury is concrete if the harm is “real.” Muransky, 979 F.3d at 926 (quotation marks omitted). Economic injuries are “[c]ertainly” concrete. Debernardis v. IQ Formulations, LLC, 942 F.3d 1076, 1084 (11th Cir. 2019). So are identity theft and damages resulting from such theft, see Resnick v. AvMed, Inc., 693 F.3d 1317, 1323 (11th Cir. 2012), as well as wasted time, Salcedo v. Hanna, 936 F.3d 1162, 1173 (11th Cir. 2019). A plaintiff can also satisfy the concreteness element by showing a “material” risk of harm. Muransky, 979 F.3d at 927 (quotation marks omitted). This Court has said this is a “high standard” that requires courts to consider the “magnitude of the risk.” Id. This Court has also addressed injuries incurred while mitigating a risk of harm, such as purchasing a credit freeze or spending time or effort to minimize a risk of identity theft. “[A]ny

14

assertion of wasted time and effort necessarily rises or falls along with this Court’s determination of whether” a risk of injury is a concrete harm. Id. at 931. For that reason, when a plaintiff faces a sufficient risk of harm, the time, money, and effort spent mitigating that risk are also concrete injuries.

We now turn to the actual-or-imminent element. When there is no actual injury, an imminent injury must be “certainly impending,” as allegations of “possible future injury are not sufficient.” Clapper v. Amnesty Int’l USA, 568 U.S. 398, 409, 133 S. Ct. 1138, 1147 (2013) (emphases and quotation marks omitted). It need not be “literally certain” that the injury will come about, but there must be a “substantial” risk. Id. at 414 n.5, 133 S. Ct. at 1150 n.5 (quotation marks omitted).

Applying these principles to the case before us, Plaintiffs have plausibly alleged an injury in fact.9 Plaintiffs alleged that “hackers obtained at least 146.6 million names, 146.6 million dates of birth, 145.5 million Social Security numbers, 99 million addresses, 17.6 million driver’s license numbers, 209,000 credit card numbers, and 97,500 tax identification numbers.” With this information, Plaintiffs alleged that “identity thieves can create fake identities, fraudulently obtain loans and tax refunds, and destroy a consumer’s credit-worthiness.” Plaintiffs also

9 Mr. Huang says Plaintiffs forfeited any arguments in support of standing by not raising them in the District Court. But Plaintiffs pled countless allegations of injury in their complaint. We therefore reject Mr. Huang’s argument.

15

alleged they “remain subject to a pervasive, substantial and imminent risk of identity theft and fraud” due to the “highly-sensitive nature of the information stolen,” and that they spent time, money, or effort dealing with the breach. Given the colossal amount of sensitive data stolen, including Social Security numbers, names, and dates of birth, and the unequivocal damage that can be done with this type of data, we have no hesitation in holding that Plaintiffs adequately alleged that they face a “material” and “substantial” risk of identity theft that satisfies the concreteness and actual-or-imminent elements. See Muransky, 979 F.3d at 927; Clapper, 568 U.S. at 414 n.5, 133 S. Ct. at 1150 n.5.

The actual identity theft already suffered by some Plaintiffs further demonstrates the risk of identity theft all Plaintiffs face—though actual identity theft is by no means required when there is a sufficient risk of identity theft. Here, dozens of Plaintiffs allege they have already had their identities stolen and thus suffered injuries in many different ways. Specifically, those who suffered identity theft had numerous unauthorized charges and accounts made in their name; incurred specific numerical drops in their credit scores; had their ability to obtain loans affected; purchased credit monitoring; and spent time, money, and effort trying to mitigate their injuries, including disputing fraudulent activity, filing police reports, and otherwise dealing with identity theft. There is no dispute that these Plaintiffs’ allegations of identity theft and resulting damages “constitute[] an

16

injury in fact under the law.” 10 Resnick, 693 F.3d at 1323. As such, the allegations of some Plaintiffs that they have suffered injuries resulting from actual identity theft support the sufficiency of all Plaintiffs’ allegations that they face a risk of identity theft. Indeed, in Tsao v. Captiva MVP Restaurant Partners, LLC, 986 F.3d 1332 (11th Cir. 2021), our Court recently recognized that “some allegations of actual misuse or actual access to personal data” support Article III standing for “a data breach based on an increased risk of theft or misuse.” Id. at 1340 (collecting cases); see also, e.g., McMorris v. Carlos Lopez & Assocs., LLC, 995 F.3d 295, 301–02 (2d Cir. 2021) (“[C]ourts have been more likely to conclude that plaintiffs have established a substantial risk of future injury where they can show that at least some part of the compromised dataset has been misused.”) (collecting cases).

Beyond the sufficient risk of identity theft and resulting injuries, a vast number of Plaintiffs who have not yet suffered identity theft also allege they have spent time, money, and effort mitigating the risk of identity theft. Their efforts include purchasing credit freezes, monitoring their financial accounts, and purchasing credit monitoring, among other things. As explained above, because

10 These Plaintiffs’ allegations of this sort of “injury in fact” provide them with Article III standing. And as noted, only one named plaintiff must have standing for any particular claim to advance. Wilding, 941 F.3d at 1124–25. This means we could also undertake a claim-by-claim analysis of the many claims in this case to determine if there is at least one named plaintiff with the sort of injury required to bring each claim. But because we conclude that all Plaintiffs have adequately alleged a sufficient risk of identity theft, we need not undertake this additional task.

17

the risk of harm here is a sufficient injury, the allegations of mitigation injuries made by these Plaintiffs are also sufficient. See Muransky, 979 F.3d at 931 (“[A]ny assertion of wasted time and effort necessarily rises or falls along with this Court’s determination of whether the risk posed . . . is itself a concrete harm.”).

Plaintiffs have easily shown an injury in fact.

ii. Redressability

With the issue of injury now resolved, we move on to address whether Plaintiffs’ injuries are redressed by the settlement. Mr. Huang says those Plaintiffs who have not had their identities stolen cannot have their injuries redressed by the settlement because the settlement does not stop third parties from committing identity theft. We need not linger on this issue, as Mr. Huang’s argument misunderstands the allegations of the complaint as well as the nature of the settlement. The Plaintiffs who have not suffered identity theft did not sue Equifax in order to stop third parties from committing identity theft. Instead, they sued Equifax because of their injuries associated with the risk of identity theft. As discussed, these injuries include the time, money, and effort spent mitigating the risk of identity theft, including purchasing credit freezes, monitoring their financial accounts, and purchasing credit monitoring, among other things.

The settlement redresses the injuries resulting from these mitigation efforts.

Specifically, for each class member, the settlement includes reimbursement for up

18

to $20,000 of documented, out-of-pocket losses fairly traceable to the data breach (e.g., the cost of purchasing credit freezes and credit monitoring), and compensation of $25 per hour for up to 20 hours for time spent taking preventative measures against identity theft. And while the additional settlement benefits of 10 years of credit monitoring and seven years of identity restoration services might not stop a third party from committing identity theft, these benefits will help limit Plaintiffs’ injuries. Credit monitoring can quickly alert Plaintiffs to an identity theft, and identity restoration services will help minimize the time and money spent by Plaintiffs to combat an identity theft. The settlement thus redresses Plaintiffs’ injuries. See Lujan, 504 U.S. at 561, 112 S. Ct. at 2136.

2. Case-or-Controversy Requirement With the issue of standing resolved, we now consider Mr. Huang’s other argument concerning Article III jurisdiction. In his view, the District Court lacked jurisdiction to approve the settlement because once the parties agreed to settle their dispute, there was not a case or controversy between the parties. Of course, Article III permits federal courts to address only “cases and controversies,” which limits their jurisdiction to “questions presented in an adversarial context.” Graham v. Butterworth, 5 F.3d 496, 498–99 (11th Cir. 1993) (citing Flast v. Cohen, 392 U.S. 83, 94–95, 88 S. Ct. 1942, 1949–50 (1968)). The controversy must exist at all

19

stages of the litigation. Preiser v. Newkirk, 422 U.S. 395, 401, 95 S. Ct. 2330, 2334 (1975).

We are aware of no court that has adopted Mr. Huang’s idea that a district court is somehow divested of jurisdiction (and thus lacks authority to approve the settlement) once parties agree to settle a class action. As we understand Mr. Huang’s position, no class action could ever be approved, because as soon as the parties decide to settle, the case or controversy would vanish, and the court would therefore lack jurisdiction to approve the settlement.

To the contrary, we hold that Article III’s case-or-controversy requirement is satisfied throughout the settlement process because the litigation remains in an adversarial posture during that process. First, the parties themselves remain in adversarial positions until the district court approves the settlement. Rule 23(e) states a class action “may be settled . . . only with the court’s approval.” Fed. R. Civ. P. 23(e). This means the parties’ decision to settle a class action is not consummated until the district court actually approves it. Cf. Haven Realty Corp. v. Coleman, 455 U.S. 363, 371 n.10, 102 S. Ct. 1114, 1120 n.10 (1982) (holding that a settlement agreement did not moot certain claims because the agreement was “still subject to the approval of the District Court”). Indeed, the parties remain adversaries all throughout the settlement approval process because until approval, the settlement is not final, and if the district court rejects the settlement, the parties

20

would continue their litigation. See In re Asbestos Litig., 90 F.3d 963, 988 (5th Cir. 1996) (holding Article III’s case-or-controversy requirement was satisfied, notwithstanding a settlement, in light of the “the adversarial positions which the parties occupied before settlement negotiations and the positions to which they will return if the settlement is not approved”), vacated on other grounds, Ortiz v. Fibreboard Corp., 521 U.S. 1114, 117 S. Ct. 2503 (1997) (mem.).

Second, because the district court acts as a fiduciary for the class, there remains adversity between the class and the defendant. Rule 23(e) requires the district court to ensure the settlement is “fair, reasonable, and adequate.” Fed. R. Civ. P. 23(e)(2). The district court thus takes on a type of fiduciary role for the class, NPAS Sols., 975 F.3d at 1253, and works to ensure the settlement is “noncollusive in nature,” 4 Newberg § 13:40; see also Manual for Complex Litigation § 21.61 (“[T]he judge must adopt the role of a skeptical client and critically examine . . . the proposed settlement terms[.]”). Our Court directs district judges to exercise “careful scrutiny” in order to “guard against settlements that may benefit the class representatives or their attorneys at the expense of absent class members.” Holmes v. Cont’l Can Co., 706 F.2d 1144, 1147 (11th Cir. 1983) (quotation marks omitted). Third and finally (and as this case demonstrates), objectors cause the settlement process to be more adversarial. While the settling parties may agree about the prospect of settlement, class action settlements are

21

routinely subjected to objections that “provide the court an adversarial presentation of the issues under review, bringing the decision-making process closer to a familiar judicial decision.” 4 Newberg § 13:40. B. Requirements Imposed on the Objectors Having established jurisdiction, we now turn to the Objectors’ various challenges to the District Court’s decisions in this case. After Plaintiffs and Equifax presented their final settlement agreement to the District Court, that court ordered notice of the settlement agreement to be provided to the class, such that members of the class had the opportunity to opt-out of the class or object to the settlement. In the order directing notice to the class, the District Court imposed a number of administrative requirements on those class members who wished to object. The District Court explained that it imposed these requirements because in a class action case it previously handled, an objector came in “out of the blue” and created a “really chaotic process.” It also found that such requirements can help “expose objections that are lawyer-driven and filed with ulterior motives.”

Among other things, the District Court required that each objection include:

the objector’s name and address; the objector’s personal signature; the grounds for the objection; previous objections in recent class actions; and dates on which the objector was available to be deposed. In addition, if the objector had counsel who intended to speak at the fairness hearing, the objection needed to include the legal

22

and factual basis for the objection and the evidence to be offered at the hearing. Finally, if the objector had counsel who sought compensation from anyone other than the objector, the objection needed to include counsel’s previous objections in recent class actions, counsel’s experience in class action litigation, and information on the fees sought. Mr. Davis says these administrative requirements infringed on the objectors’ right to be heard and to be represented by counsel in their objections. More to the point, he says that by imposing these requirements on objectors, including those with counsel, the District Court limited their right to object and deterred objections.11 We review a district court’s management of a class action for abuse of discretion. See Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 342, 98 S. Ct. 2380, 2385 (1978) (applying abuse of discretion standard to a district court’s order “concern[ing] the conduct of class actions” under Rule 23). Rule 23 grants district courts broad discretion to manage class actions. See Nissan Motor Corp., 552 F.2d at 1096 (“In the management of class actions, [Rule] 23 necessarily vests the district courts with a broad discretion to enable efficacious administration of the course of proceedings before it.”); Gulf Oil Co. v. Bernard, 452 U.S. 89, 100, 101 S. Ct. 2193, 2200 (1981) (“Because of the potential for abuse, a district court has

11 Mr. Davis also says the requirements allowed the District Court to reject objections on technical grounds. However, the District Court considered and rejected all objections on their merits “whether or not the objections [were] procedurally valid.”

23

both the duty and the broad authority to exercise control over a class action and to enter appropriate orders governing the conduct of counsel and parties.”). For instance, Rule 23 authorizes district courts to “prescribe measures to prevent undue repetition or complication in presenting evidence or argument,” to “impose conditions on the representative parties or on intervenors,” and to “deal with similar procedural matters.” Fed. R. Civ. P. 23(d)(1)(A), (C), (E). At the same time, district courts’ discretion is “not unlimited.” Gulf Oil, 452 U.S. at 100, 101 S. Ct. at 2200.

Mr. Davis has failed to show the District Court abused its discretion here.

The District Court explicitly imposed the requirements outlined here, not to deter objections or for some arbitrary purpose, but for the express purpose of avoiding a “chaotic process” in evaluating the objections. The District Court said it found these requirements help “expose objections that are lawyer-driven and filed with ulterior motives.” The District Court was well within its broad discretion to impose the requirements for these stated purposes. See id.; see also, e.g., In re Deepwater Horizon, 739 F.3d 790, 809 (5th Cir. 2014) (requirement imposed on objector a “legitimate exercise” of court’s discretion to minimize abuse); Manual for Complex Litigation § 21.662 (courts may be “inclined to find [discovery from objectors] useful to assess the validity of the objections”); 4 Newberg § 13:33

24

(“[C]lass counsel may seek discovery from objectors on issues such as the objectors’ . . . relationships with the professional objector counsel.”). 12 Beyond that, the requirements the District Court imposed were not particularly burdensome. Most requirements were clerical in nature, such as simply providing information. The most potentially burdensome requirement was being deposed, yet in many instances that was no more than a possibility. And of course, depositions are a normal part of litigation, see Fed. R. Civ. P. 30, including for objectors to class settlements, see In re Cathode Ray Tube (CRT) Antitrust Litig., 281 F.R.D. 531, 532, 534 (N.D. Cal. 2012) (ordering objector to sit for deposition regarding the “bases for his objection” and his “relationship with ‘professional’ or ‘serial’ objector counsel”); see also Granillo v. FCA US LLC, 2018 WL 4676057, at *7 (D.N.J. Sept. 28, 2018) (“[C]ourts across the country have approved . . . depositions of objectors who have voluntarily inserted themselves into [an] action[.]” (quotation mark omitted)); In re Netflix Privacy

12 Mr. Davis notes that a recent amendment to Rule 23 requires district courts to approve any agreement between an objector and class counsel in which payment is “provided in connection with” a decision to forgo or withdraw an objection. See Fed. R. Civ. P. 23(e)(5)(B). In his view, this means objectors no longer bring meritless objections with the hope of being paid off, and thus the District Court’s requirements were unnecessary. But Mr. Davis’s argument assumes the amendment completely eliminated this type of extortion, which may not be a settled question. See, e.g., In re Foreign Exch. Benchmark Rates Antitrust Litig., 334 F.R.D. 62, 64 (S.D.N.Y. 2019) (“Approving agreements in these circumstances would serve only to encourage objectors or their attorneys to extract this type of payment[.]”). And even if the question is settled, this District Court did not abuse its discretion when it imposed the requirements for the other reasons discussed.

25

Litig., 2013 WL 6173772, at *5 (N.D. Cal. Nov. 25, 2013) (“[W]hile absent class members are not normally included in discovery, Objectors have voluntarily inserted themselves into this action, and as such, depositions of the Objectors are relevant and proper.”). 13 To be sure, discovery requirements may in some cases “dissuade class members from exercising their right to object.” 4 Newberg § 13:33. But here, the District Court found that any concerns about requirements deterring objections were “at odds with the number of objections received” and the fact that “few objectors had difficulty meeting these criteria.” Mr. Davis has not shown the District Court erred in making this finding, especially given that the requirements were not particularly burdensome.

Of course, district courts must remain mindful that burdensome requirements could deter objectors from exercising their right to object while also fulfilling their obligation to manage class actions. This can be a difficult task. Whether a district court abuses its discretion in striking the right balance will invariably depend on the facts of each case, and the breadth of a court’s discretion in this regard will tend to ebb and flow with the size and administrative difficulties

13 There seems to be a dispute about whether all objectors were subject to depositions or just those that were represented by counsel. But Mr. Davis, who was not represented by counsel in his objection, admits that Plaintiffs sought to depose him. Thus, it appears both represented and non-represented objectors were subject to depositions.

26

of the class action. With this class of approximately 147 million members, the District Court acted well within its discretion to impose the requirements it did. C. Order Certifying Class and Approving Settlement At the final hearing, after hearing arguments from Plaintiffs, Equifax, and various objectors, and after giving its oral rulings, the District Court directed Plaintiffs’ counsel to draft a proposed order “summariz[ing] [the District Court’s] rulings on the motions and [its] adoption basically of the arguments that have been made by the Plaintiffs and by Equifax in the hearing today.” The District Court instructed Plaintiffs’ counsel to obtain Equifax’s approval before submitting the proposed order to the court, which it would then “consider signing.” The District Court acted pursuant to its local rule, which states, “[u]nless the Court directs otherwise, all orders . . . orally announced by the district judge in Court shall be prepared in writing by the attorney for the prevailing party.” N.D. Ga. R. 7.3. There is no indication in the record that the proposed order was ever disclosed to the class or filed on the docket. In fact, Plaintiffs acknowledge they emailed the proposed order directly to the District Court.

Some Objectors challenge this procedure on various grounds. These challenges include the assertions that: the District Court erred in adopting a proposed order “ghostwritten” by Plaintiffs’ counsel; engaged in impermissible ex parte communications and violated various rules by failing to disclose the proposed

27

order to the class; and erred by not including the proposed order in the appellate record. The Objectors also request that this case be reassigned to a different judge on remand.14 We consider each of these issues in turn after addressing one preliminary matter. Specifically, it is unclear how much of the proposed order— none at all, only some, or even verbatim—the District Court adopted. The Objectors ask us to assume that the District Court adopted the proposed order in full, and Plaintiffs and Equifax don’t ask us to do otherwise. For the purposes of our review, we therefore assume the District Court adopted the proposed order verbatim.

1. Ghostwritten Order Mr. Frank and Mr. West say the District Court erred in adopting a proposed order “ghostwritten” by Plaintiffs’ counsel. This Court has “repeatedly condemned the ghostwriting of judicial orders by litigants,” and cases admonishing courts for

14 Independently, Mr. Frank also argues that the District Court improperly relied on a declaration filed by Professor Robert Klonoff, who writes on class actions. Mr. Frank says the declaration was inadmissible under Federal Rule of Evidence 702 because Professor Klonoff provided a legal opinion. This issue is ultimately unrelated to the District Court’s decision to adopt a proposed order, but Mr. Frank raises this issue in passing when discussing the proposed order issue, so we address it briefly here. Courts have held that Rule 702 is flexible at the final approval stage. See, e.g., Int’l Union, United Auto., Aerospace, and Agric. Implement Workers of Am. v. Gen. Motors Corp., 497 F.3d 615, 636–37 (6th Cir. 2007); see also 4 Newberg § 13:42 (“[T]raditional rules of evidence do not necessarily apply to the fairness hearing.”). However, we need not decide whether Rule 702 applies at the final approval stage because even if Rule 702 applies—and even if Professor Klonoff’s declaration violated Rule 702—Mr. Frank fails to show the error was anything other than harmless. See Furcron v. Mail Ctrs. Plus, LLC, 843 F.3d 1295, 1304 (11th Cir. 2016) (“[E]ven a clearly erroneous evidentiary ruling will be affirmed if harmless.”). Although the District Court said Professor Klonoff’s declaration was “particularly helpful,” it expressly stated its rulings were “not dependent upon his declaration.”

28

the verbatim adoption of such orders are “legion.” In re Colony Square Co., 819 F.2d 272, 274–75 (11th Cir. 1987). When such a practice is permitted, the drafting party has an “overwhelming” “temptation to overreach and exaggerate.” Id. at 275. Beyond that, the “quality of judicial decisionmaking suffers when a judge delegates the drafting of orders to a party,” as “the writing process requires a judge to wrestle with the difficult issues before him and thereby leads to stronger, sounder judicial rulings.” Id. 15 Even so, as the parties acknowledge, our Court has not enforced a per se rule prohibiting this practice. Even though this Court has sharply critiqued the practice of having the prevailing party author court orders, we have continued to approve courts’ adoption of proposed orders, some even verbatim. See, e.g., In re Dixie Broad., Inc., 871 F.2d 1023, 1029–30 (11th Cir. 1989) (refusing to vacate “ghostwritten” order when judge told all counsel “in open court” that he asked a party’s counsel to draft the order, the other parties did not request the opportunity to review the draft order or make objections to it, and the parties had ample opportunity to argue their case); Colony Square, 819 F.2d at 276–77 (practice “not fundamentally unfair” because the judge “reached a firm decision” before asking counsel to draft the proposed order, which the judge said must reach a particular

15 We note that Northern District of Georgia Rule 7.3 does not appear to be in keeping with the admonitions of our Court about this practice.

29

result and discuss specific points, and because the losing party “had ample opportunity to present its arguments”); Fields v. City of Tarpon Springs, 721 F.2d 318, 320–21 (11th Cir. 1983) (per curiam) (district judge “did not abdicate his adjudicative role” in the “wholesale adoption of plaintiff’s proposed order” because the judge “had command of the legal issues and the evidentiary proceedings,” “ruled on the scope and manner of the evidence presented,” and was “an active arbiter of the dispute”); see also Anderson v. City of Bessemer City, 470 U.S. 564, 572, 105 S. Ct. 1504, 1510–11 (1985) (noting criticism of “courts for their verbatim adoption of findings of fact prepared by prevailing parties” yet stating “that even when the trial judge adopts proposed findings verbatim, the findings are those of the court”).

Our guiding principle in determining whether to vacate the adoption of a proposed order is whether “the process by which the judge arrived at [the order] was fundamentally unfair.” Colony Square, 819 F.2d at 276. If a process was fundamentally fair, then the concerns ordinarily associated with a ghostwritten order are greatly tempered. Without a per se rule, we determine whether a process was fundamentally unfair by evaluating the facts of each case. Also, we glean some relevant considerations from our precedent, including: whether the losing party had “ample opportunity” to present its arguments, id. at 277; see also Dixie Broad., 871 F.2d at 1030; whether the court independently “reached a firm

30

decision” before requesting a proposed order, Colony Square, 819 F.2d at 276; see also Fields, 721 F.2d at 320–21; whether the court, in directing a party to draft the proposed order, instructed that the order “reach[] a particular result and discuss[] specific points,” Colony Square, 819 F.2d at 276; whether the court directed a party to draft the proposed order in open court or otherwise publicly, Dixie Broad., 871 F.2d at 1030; whether other parties requested the opportunity to review the proposed order or make objections to it, id.; and whether the court “had command” of the issues and proceedings and was an “active arbiter” throughout the litigation, Fields, 721 F.2d at 320–21.

Applying these considerations, we conclude the process by which the District Court adopted the proposed order was not fundamentally unfair. Mr. Frank and Mr. West both had ample opportunity to present their arguments. Both lodged detailed written objections to the settlement agreement. Both appeared through counsel at the final hearing and presented arguments. And contrary to their assertions, they did have an opportunity to respond to the order. After the District Court adopted the proposed order, Mr. West moved to amend it. And it’s not as if these opportunities to present their arguments were hollow procedures; the District Court heard from Mr. Frank and Mr. West at the fairness hearing, considered their written objections, and rejected their objections on the merits.

31

Ultimately, the District Court granted Mr. West’s motion to amend the order over Plaintiffs’ objections and issued a revised order based on West’s arguments.

The District Court reached a firm decision before ever directing Plaintiffs’

counsel to draft a proposed order. And the District Court instructed that the order reach a particular result and discuss specific points: the court told Plaintiffs’ counsel that the order should “summarize[] [its] rulings on the motions and [its] adoption basically of the arguments that have been made by the Plaintiffs and by Equifax in the hearing today.”16 It even informed Plaintiffs’ counsel that it would only “consider signing” the proposed order, meaning its instruction to prepare a proposed order was not a blank check. The District Court did all this in open court for everyone to hear, including the Objectors, and not one of them objected to the process nor requested the opportunity to review the proposed order or make objections to it. There is also no question the District Court was an active arbiter of this litigation and had great command of the proceedings. For instance, the

16 We don’t find it significant that the District Court’s written order was more detailed than its oral ruling. District courts often provide a summary ruling from the bench, which is later memorialized in a longer written order. And in any event, it is ultimately the District Court’s written order that controls in civil cases. See, e.g., Billingsley v. Jefferson County, 953 F.2d 1351, 1354 (11th Cir. 1992) (“[T]he district court’s memorandum opinion constitutes its findings of facts and conclusions of law,” as the district court was not bound by its “findings of fact, rulings, or conclusions of law made during the course of [the] trial.”); Mercantel v. Michael & Sonja Saltman Family Tr., 993 F.3d 1212, 1239 & n.23 (10th Cir. 2021) (approving ghostwritten summary judgment order, even though “the final written decision cover[ed] additional issues not explicitly addressed at the hearing,” because “at least in civil cases, a court’s written decision generally controls over . . . an earlier oral ruling”).

32

District Court issued a detailed ruling on Equifax’s motion to dismiss and engaged with the issues at the fairness hearing. Finally, the fact that the District Court granted the motion to dismiss in part and denied it in part, and that it granted Mr. West’s motion to amend the approval order over Plaintiffs’ objections, shows us the court was not beholden to any party. Because the process by which the District Court adopted its order was not fundamentally unfair, we will not vacate the order.

We caution that courts should not view this decision as condoning the District Court’s practice. Judicial ghostwriting remains most unwelcome in this Circuit. For this reason, and pursuant to our supervisory power, we strongly urge the District Court to reconsider the local rule, see N.D. Ga. R. 7.3, that brought about this problem in the first place. See Piambino v. Bailey, 757 F.2d 1112, 1145–46 (11th Cir. 1985) (stating this Court has the “power to supervise the district courts” in a “wide variety of situations,” including in formulating rules of civil litigation, in order to “ensure that the judicial process remains a fair one”).

2. Ex Parte Communications Mr. Frank and Mr. West next argue there were impermissible ex parte communications in the District Court because Plaintiffs’ counsel failed to disclose the proposed order to the class. They say the ex parte communications violated Canon 3(A)(4) of the Code of Conduct for United States Judges and various local rules.

33

Any ex parte communications were harmless error. 17 See Colony Square, 819 F.2d at 276 (citing Rushen v. Spain, 464 U.S. 114, 104 S. Ct. 453 (1983)) (noting that ex parte communications can be upheld when the error is harmless). We reach this conclusion for four reasons. First, at the Objectors’ request, we assume the District Court adopted the proposed order verbatim, so for the purposes of our review the Objectors are privy to the exact communications they claim were made ex parte. Second, as discussed above, the District Court’s process was not fundamentally unfair, and thus we can affirm its decision notwithstanding any ex parte communications. See id. at 276–77. Indeed, it appears that the District Court in Colony Square engaged in more obvious ex parte communications than what the Objectors assert here, and yet this Court held there was no fundamental unfairness and thus upheld the order at issue there. In Colony Square, the judge called the prevailing party’s lawyer after the hearing and asked him to draft the order. Id. at 274. The lawyer’s firm delivered the draft order to the judge, and the losing party was not notified of any ex parte communications. Id. Here, by contrast, the District Court requested that Plaintiffs’ counsel draft a proposed order in open court, and no one objected to the process or requested to see a copy of the proposed order.

17 Because we hold that any ex parte communications were harmless error, we need not address whether the communications violated Canon 3(A)(4) or any local rules.

34

Third, as discussed throughout this opinion, we identify no errors made by the District Court, so we cannot say any ex parte communications caused the court to err in a way that prejudiced the Objectors. 18 See United States v. Adams, 785 F.2d 917, 921 (11th Cir. 1986) (holding ex parte communications were harmless error when there was no prejudice). Finally, the record demonstrates that the Objectors had the opportunity to take up with the judge any problems they identified. After the District Court issued the approval order, about which these Objectors complain, Mr. West moved to amend it. The District Court granted Mr. West’s motion and issued a revised order based on his arguments. This too shows a lack of prejudice and thus, at most, harmless error. See id.

3. Record on Appeal In the District Court, some of the Objectors moved for the appellate record to be supplemented with the proposed order under Federal Rule of Appellate Procedure 10(e). Relevant here, Rule 10(e) says “[i]f any difference arises about whether the record truly discloses what occurred in the district court, the difference must be submitted to and settled by that court and the record conformed accordingly.” Fed. R. App. P. 10(e)(1). In other words, the appellate record need

18 As addressed later, we must remand this case to the District Court for it to vacate the incentive awards for the class representatives based on this Court’s decision in NPAS Solutions, which postdated the District Court’s order. This of course is not an error that resulted from any ex parte communications, and thus it does not amount to prejudice.

35

not be supplemented when the record “truly discloses what occurred in the district court.” Hoover v. Blue Cross & Blue Shield of Ala., 855 F.2d 1538, 1543 n.5 (11th Cir. 1988) (quotation marks omitted). The District Court ultimately denied the Objectors’ motion, finding Rule 10(e) did not apply because “the record truly discloses what occurred in the district court.”

Mr. Frank and Mr. West challenge this ruling and take issue with the fact that the proposed order is not in the record on appeal. They wish to supplement the record in order to show that the final order was a verbatim copy of the proposed order. Mr. Davis likewise says the public has a right to view the proposed order, primarily to determine whether the District Court adopted it in full. But in light of our decision to accede to the Objectors’ request that we assume a verbatim adoption of the proposed order by the District Court, there is no need to supplement the record with this material on appeal. Id. Because we assume the proposed order (which is not in the record) is identical to the approval order under review here (which is in the record), the record on appeal reflects what occurred in the District Court, at least for purposes of our review.

4. Reassignment on Remand In light of the foregoing supposed errors, as well as some other late-breaking allegations of bias, Mr. Frank and Mr. West ask us to reassign this case to a different judge on remand. As discussed below, we must remand this case to the

36

District Court solely for it to vacate the incentive awards. We therefore must briefly discuss the issue of reassignment on remand.

To begin, it’s not obvious to us that we have the authority to reassign this case. This case was assigned to Chief Judge Thrash by the United States Judicial Panel on Multidistrict Litigation (the “Panel”). Under the Panel’s rules, “[i]f for any reason the transferee judge is unable to continue [its] responsibilities, the Panel shall make the reassignment of a new transferee judge.” Rules of Procedure of the United States Judicial Panel on Multidistrict Litigation, Rule 2.1(e) (emphasis added); see In re IKO Roofing Shingle Prods. Liab. Litig., 757 F.3d 599, 600 (7th Cir. 2014) (explaining that 28 U.S.C. § 1407(b) “gives the Panel exclusive power to select the judge”).

But we need not decide that question because reassignment is not justified here. Reassignment is a “severe remedy,” which is “only appropriate where the trial judge has engaged in conduct that gives rise to the appearance of impropriety or a lack of impartiality in the mind of a reasonable member of the public.” Comparelli v. Republica Bolivariana de Venezuela, 891 F.3d 1311, 1328 (11th Cir. 2018) (quotation marks omitted). The Objectors have not shown any actual bias from Chief Judge Thrash. While they certainly disagree with his decisions, “judicial rulings alone almost never constitute a valid basis for a bias or partiality

37

motion.” Liteky v. United States, 510 U.S. 540, 555, 114 S. Ct. 1147, 1157 (1994).

Without any indication of actual bias, this Court considers three factors when deciding whether to reassign a case: “(1) whether the original judge would have difficulty putting [his] previous views and findings aside; (2) whether reassignment is appropriate to preserve the appearance of justice; [and] (3) whether reassignment would entail waste and duplication out of proportion to gains realized from reassignment.” Comparelli, 891 F.3d at 1328 (quotation marks omitted) (quoting United States v. Torkington, 874 F.2d 1441, 1447 (11th Cir. 1989) (per curiam)). No factor supports reassignment here. As to the first factor, Chief Judge Thrash would not have difficulty putting his views aside, as the record indicates he corrects his mistake and amends his orders when he thinks he reached the wrong result. For instance, he revised the approval order after Mr. West moved to amend. For the second factor, Mr. Frank and Mr. West have not shown how reassignment is appropriate to preserve the appearance of justice. Finally, the third factor clearly weighs against reassignment. This is a colossal multidistrict litigation case with over 1200 docket entries in the District Court over the course of just a few years. Reassignment would create enormous waste and duplication.

In support of reassignment, Mr. Frank and Mr. West rely heavily on Chudasama v. Mazda Motor Corp., 123 F.3d 1353 (11th Cir. 1997), in which this

38

Court reassigned a case on remand in part because of “the court’s practice of uncritically adopting counsel’s proposed orders.” Id. at 1373 n.46. Chudasama is far from on point. For one, as discussed above, the District Court’s decision to adopt the proposed order at issue here was not fundamentally unfair. Beyond that, the Objectors have not established a “practice” by the District Court of “uncritically adopting” proposed orders. Id. Finally, this case involves none of the four other considerations that contributed to this Court’s decision to reassign the case in Chudasama. See id. D. Settlement Approval We now turn to the substance of the approval order, beginning with the District Court’s approval of the settlement agreement. A class action may be settled only with court approval, which requires the court to find the settlement “fair, reasonable, and adequate” based on a number of factors. Fed. R. Civ. P. 23(e)(2). 19 This Court has also instructed district courts to consider several

19 The Rule 23(e)(2) factors include whether:

(A) the class representatives and class counsel have adequately represented the class;

(B) the proposal was negotiated at arm’s length;

(C) the relief provided for the class is adequate, taking into account:

(i) the costs, risks, and delay of trial and appeal;

(ii) the effectiveness of any proposed method of distributing relief to the class, including the method of processing classmember claims;

(iii) the terms of any proposed award of attorney’s fees, including timing of payment; and

39

additional factors called the Bennett factors. See Bennett, 737 F.2d at 986. The factors include (1) “the likelihood of success at trial”; (2) “the range of possible recovery”; (3) “the point on or below the range of possible recovery at which a settlement is fair, adequate and reasonable”; (4) “the complexity, expense and duration of litigation”; (5) “the substance and amount of opposition to the settlement”; and (6) “the stage of proceedings at which the settlement was achieved.” Id. The District Court here considered the Rule 23(e)(2) factors and the Bennett factors and found the settlement was “fair, reasonable, and adequate” and that the settlement’s relief “exceeds the relief provided in other data breach settlements and . . . is in the high range of possible recoveries if the case had successfully been prosecuted through trial.”

We review an order approving a class action settlement for abuse of discretion. Ault v. Walt Disney World Co., 692 F.3d 1212, 1216 (11th Cir. 2012). And because “[d]etermining the fairness of the settlement is left to the sound discretion of the trial court,” we will not overturn its decision “absent a

(iv) any agreement required to be identified under Rule 23(e)(3); and

(D) the proposal treats class members equitably relative to each other.

Fed. R. Civ. P. 23(e)(2). The Objectors do not challenge the District Court’s application of the Rule 23(e)(2) factors, so we do not address them in depth. Although Mr. Frank says in passing that Rule 23(e)(2)(D) was not satisfied, he does not press it with any argument or authority. We therefore treat his argument abandoned. See Access Now, Inc. v. Sw. Airlines Co., 385 F.3d 1324, 1330 (11th Cir. 2004).

40

clear showing of abuse of that discretion.” Bennett, 737 F.2d at 986 (emphasis added); see also 4 Newberg § 13:47 (“[A]ppellate courts review the approval decision under a highly deferential abuse of discretion standard.”).

This degree of deference to a decision approving a class action settlement makes sense. Settlements resolve differences and bring parties together for a common resolution. See Nissan Motor Corp., 552 F.2d at 1105 (“Settlement agreements are highly favored in the law and will be upheld whenever possible because they are a means of amicably resolving doubts and uncertainties and preventing lawsuits.” (quotation marks omitted)). Settlements also save the bench and bar time, money, and headaches. See 4 Newberg § 13:44 (“The law favors settlement, particularly in class actions and other complex cases where substantial resources can be conserved by avoiding lengthy trials and appeals.”). As such, there is a “strong judicial policy favoring settlement.” Bennett, 737 F.2d at 986.

Mr. Cochran challenges this settlement approval because he says the District Court’s approval order failed to recognize the “unique risks associated with stolen Social Security numbers,” which means the settlement includes inadequate relief to remedy those risks. From this vantage point, he thinks the District Court misapplied two of the Bennett factors: “the range of possible recovery” and “the point on or below the range of possible recovery at which a settlement is fair, adequate and reasonable.” Id.

41

Mr. Cochran has failed to show an abuse of discretion, as the record clearly demonstrates the District Court was aware of the unique risks associated with stolen Social Security numbers. The complaint alleged, among other things, that “all of the 147.9 million Americans whose information was stolen in the breach remain subject to a pervasive, substantial and imminent risk of identity theft and fraud, a risk that will continue so long as Social Security numbers have such a critical role in consumers’ financial lives.” In its order on Equifax’s motion to dismiss, the District Court acknowledged the data breach involved Social Security numbers and that “[u]sing this information, identity thieves can create fake identities, fraudulently obtain loans and tax refunds, and destroy a consumer’s credit-worthiness.” And at the fairness hearing for approval of the settlement, both Plaintiffs’ counsel and an objector discussed the risks associated with stolen Social Security numbers. Finally, the District Court’s approval order expressly highlighted that the settlement includes a “lengthy period” of credit monitoring and “identity theft insurance and identity restoration services—features designed to address identity theft.” In this way, the approval order recognized that the settlement includes measures to redress the very risks Mr. Cochran says the District Court ignored.

We also reject Mr. Cochran’s view that the District Court misapplied two of the Bennett factors: “the range of possible recovery” and “the point on or below

42

the range of possible recovery at which a settlement is fair, adequate and reasonable.” Bennett, 737 F.2d at 986. When considering these factors, the Court found the settlement was “fair, reasonable, and adequate because the settlement reflects relief the Court finds is in the high range of what could have been obtained had the parties continued to litigate.” Because the District Court was aware of the risks associated with stolen Social Security numbers and found that the settlement includes benefits to redress those risks, there is no “clear showing” that the District Court abused its discretion in applying these factors. See id. And while Mr. Cochran might wish for longer credit monitoring and identity theft restoration services, such quibbling with a settlement’s terms is not a part of an abuse of discretion review. See Cotton v. Hinton, 559 F.2d 1326, 1330 (5th Cir. 1977) (trial judge “should be hesitant to substitute its own judgment for that of counsel”); Manual for Complex Litigation § 21.61 (“The judge cannot rewrite the agreement.”); see also Bennett, 737 F.2d at 986 (“[C]ompromise is the essence of settlement.”). E. Class Certification In addition to approving the settlement, the District Court’s order also certified the class action for settlement purposes. Rule 23 sets forth a number of requirements that a class action must meet in order for a district court to certify the class. First, all four requirements in Rule 23(a) must be satisfied: (1) the class

43

must be “so numerous that joinder of all members is impracticable”; (2) there must be “questions of law or fact common to the class”; (3) the class representatives’ claims or defenses must be “typical” of the class’s claims or defenses; and (4) the class representatives must “fairly and adequately protect the interests of the class.” Fed. R. Civ. P. 23(a); see Vega v. T-Mobile USA, Inc., 564 F.3d 1256, 1265 (11th Cir. 2009). These four requirements are often referred to as the numerosity, commonality, typicality, and adequacy requirements, respectively. See Vega, 564 F.3d at 1265.

In addition to meeting these four requirements, a class action must also satisfy one of the three parts of Rule 23(b). Fed. R. Civ. P. 23(b); see Vega, 564 F.3d at 1265. The District Court here found that the class action satisfied Rule 23(b)(3). Rule 23(b)(3) requires that “questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). We review a class certification ruling for abuse of discretion. See Ault, 692 F.3d at 1216.

Only Mr. Frank challenges the District Court’s class certification ruling, and he does so only as to the adequacy requirement. 20 We thus focus solely on the

20 When discussing Article III standing, Mr. Huang briefly says Rule 23(b)(3)’s predominance requirement is not met. He does not press this assertion with any argument, so we consider it abandoned. See Access Now, 385 F.3d at 1330.

44

adequacy requirement in Rule 23(a)(4). The adequacy requirement is that a class representative “must adequately protect the interests of those he purports to represent.” Valley Drug Co. v. Geneva Pharms., LLC, 350 F.3d 1181, 1189 (11th Cir. 2003) (quotation marks omitted). To determine whether the adequacy requirement is met, we ask: “(1) whether any substantial conflicts of interest exist between the representatives and the class; and (2) whether the representatives will adequately prosecute the action.” Id. (quotation marks omitted). Mr. Frank’s challenge concerns only the first question, further narrowing our focus. According to Mr. Frank, there is a fundamental conflict of interest between the class representatives and the class because some class members had state statutory damages claims while others did not. He says there should have been subclasses and separate counsel to address these different types of claims.

Minor differences in the interests of the class representatives and the class are not enough to defeat class certification under the adequacy requirement. Id. Instead, only a “fundamental” conflict “going to the specific issues in controversy” can defeat class certification. Id. (quotation marks omitted); see 1 Newberg § 3:58 (“[N]ot every potential distinction . . . will render the representative inadequate. Only conflicts that are fundamental to the suit and that go to the heart of the litigation prevent a plaintiff from meeting the Rule 23(a)(4) adequacy requirement.” (footnote omitted)). A conflict is fundamental “where some party

45

members claim to have been harmed by the same conduct that benefitted other members of the class.” Valley Drug, 350 F.3d at 1189. This Court has also recognized that a class action “cannot be certified when its members have opposing interests” or “where the economic interests and objectives of the named representatives differ significantly from the economic interests and objectives of unnamed class members.” Id. at 1189–90 (quotation marks omitted).

Mr. Frank has failed to show the District Court abused its discretion in certifying the settlement class. There is no dispute that all these Plaintiffs’ claims arise out of the same unifying event, Equifax’s data privacy breach. Likewise, all Plaintiffs seek redress for the same injury. They all seek compensation for injuries associated with the risk of identity theft. There is also no dispute that the data breach harmed all class members and made none better off. See id. at 1189. Indeed, the class is expressly limited to the “U.S. consumers identified by Equifax whose personal information was compromised.”

It is true that some class members had state law statutory damages claims while others did not, but we don’t view that difference as a “fundamental” conflict “going to the specific issues in controversy.” Id. As Mr. Frank acknowledged at oral argument, only the District of Columbia and Utah statutory damages claims are before us on appeal. For one thing, Mr. Frank’s singular devotion to the D.C. and Utah claims ignores the fact that all class members had negligence and

46

negligence per se claims under Georgia law that united the class. What’s more, Mr. Frank fails to show that the two statutory damages claims were valuable, as he demonstrates nothing about how the claims were a sure bet. In fact, he doesn’t cite a single case in which a plaintiff recovered statutory damages under either statute in a data breach case.

Even a brief review of the D.C. and Utah claims reveals significant barriers to Plaintiffs’ success. While the D.C. Code authorizes certain damages for a data breach, this provision wasn’t enacted until 2020, well after the data breach here occurred in 2017. See Security Breach Protection Amendment Act of 2020, D.C. Laws 23-98 (2020). Perhaps Plaintiffs could have tried to frame the data breach as a violation of D.C.’s prohibition against certain unfair or deceptive trade practices, see D.C. Code § 28-3904, but that presents its own set of issues about proving that a breach by a third party was an unfair or deceptive trade practice by Equifax. The Utah claim, in turn, required Plaintiffs to show Equifax engaged in a “[c]onsumer transaction” and “knowingly or intentionally” violated an enumerated prohibition. Utah Code §§ 13-11-3(2), -4(2). Given that Equifax might not have been in privity with Plaintiffs and that this case arose out of a breach by a third party, these requirements may well have been difficult to show. Therefore, to the extent some class members had D.C. or Utah statutory damages claims while others did not, there were no opposing or economic interests that were at odds.

47

The District Court’s decision aligns with the reasoned approach adopted by courts in other data breach cases. For instance, in In re Anthem, Inc. Data Breach Litigation, 327 F.R.D. 299 (N.D. Cal. 2018), the District Court found the adequacy requirement satisfied because all class members had their personal information compromised in the same data breach and generally sought the same relief. Id. at 309–11. Beyond that, the District Court found that even though there might have been variations in state law, the named representatives included “individuals from each state” and the differences in state remedies were not “sufficiently substantial so as to warrant the creation of subclasses.” Id. at 310 (quotation marks omitted). The same reasoning applies here. The class members all had their personal information compromised in the same data breach; they seek redress for similar injuries; and, to the extent some members have statutory damages under state law and others do not, there are class representatives from every single state and, due to litigation risks, the differences in remedies are not “sufficiently substantial so as to warrant the creation of subclasses.” Id. (quotation marks omitted); see also In re Target Corp. Customer Data Sec. Breach Litig., 2017 WL 2178306, at *6 (D. Minn. May 17, 2017) (noting the “availability of potential statutory damages . . . does not, by itself, mean that the interests of these class members are antagonistic to the interests of class members from other jurisdictions,” particularly in light of

48

the “substantial barriers to any individual class member actually recovering statutory damages”).

By contrast, the decisions Mr. Frank relies on, in which courts held that class actions failed to satisfy the adequacy requirement, are inapposite. For instance, Ortiz v. Fibreboard Corp., 527 U.S. 815, 119 S. Ct. 2295 (1999) and Amchem Products, Inc. v. Windsor, 521 U.S. 591, 117 S. Ct. 2231 (1997), addressed two class actions against asbestos manufacturers, neither of which involved statutory damages claims. Ortiz, 527 U.S. at 821–22, 119 S. Ct. at 2302–03; Amchem, 521 U.S. at 597, 117 S. Ct. at 2237. The plaintiffs had diametrically different injuries within each class action. Some plaintiffs in the classes had already suffered physical injury as a result of exposure to asbestos, while others might have been at risk of injury in the future. Ortiz, 527 U.S. at 856, 119 S. Ct. at 2319; Amchem, 521 U.S. at 602–03, 117 S. Ct. at 2240. This meant some plaintiffs wanted compensation immediately while others wanted it in the future. Ortiz, 527 U.S. at 856, 119 S. Ct. at 2320; Amchem, 521 U.S. at 626, 117 S. Ct. at 2251. Here, by contrast, Plaintiffs alleged that they face the same risk of identity theft and, among other things, sought the same compensatory damages for that injury. Plaintiffs likewise all receive the same benefits to redress that shared injury. And while Ortiz was also based on the fact that some plaintiffs had more valuable claims than others, see Ortiz, 527 U.S. at 857, 119 S. Ct. at 2320, that’s not the case here. For

49

the reasons set out above, Mr. Frank has failed to show how the D.C. and Utah statutory damages claims increased the value of certain Plaintiffs’ cases.

Mr. Frank’s reliance on In re Literary Works in Electronic Databases Copyright Litigation, 654 F.3d 242 (2d Cir. 2011), even if it was binding on this Court, is even further off the mark. In Literary Works, there were three groups of claims (Categories A, B, and C, which had claims that decreased in value respectively) involving three different provisions of the Copyright Act. Id. at 246. The proposed settlement said that if all claims exceeded a set cap, Category C claims would be reduced pro rata first, such that those with just Category C claims might end up with nothing. Id. The Second Circuit reasoned the adequacy requirement of Rule 23(a)(4) was not met because Category A and Category B claims were “more lucrative” than Category C claims and because the reduction of Category C claims could “deplete the recovery of Category C-only plaintiffs in their entirety before the Category A or B recovery would be affected.” Id. at 252, 254. But here (putting aside the already addressed issue of the value of the statutory damages claims), there is no risk that any members of the class will have their ability to get settlement benefits reduced to zero because some other members got more relief from the settlement. Instead, all class members are entitled to the same class benefits.

50

At bottom, this record reflects no fundamental conflict between the class representatives and the rest of the class, and thus the adequacy requirement under Rule 23(a)(4) was satisfied. The District Court therefore did not abuse its discretion in certifying the class action. 21 F. Attorney’s Fee Award In addition to approving the settlement as fair, reasonable, and adequate and certifying the class action for settlement purposes, the District Court also approved Plaintiffs’ counsel’s request for $77.5 million in attorney’s fees. See Fed. R. Civ. P. 23(h) (“In a certified class action, the court may award reasonable attorney’s fees and nontaxable costs that are authorized by law or by the parties’ agreement.”). Relying on our precedent in Camden I Condominium Association, Inc. v. Dunkle, 946 F.2d 768 (11th Cir. 1991), the District Court applied what’s called the percentage method. In Camden I, this Court held that in common fund settlements like this one, an attorney’s fee award “shall be based upon a reasonable percentage of the fund established for the benefit of the class.” Id. at 774. The

21 Mr. Frank also says the District Court erred in finding that separate subclasses and representation would not benefit the class as a whole. Related to our conclusion that the District Court did not abuse its discretion in certifying the settlement class is our conclusion that the District Court did not err in its separate finding that subclasses would not benefit the class as a whole. See, e.g., In re Oil Spill by Oil Rig Deepwater Horizon in Gulf of Mex., on Apr. 20, 2010, 910 F. Supp. 2d 891, 918–19 (E.D. La. 2012) (“In the absence of conflicts between members of the Settlement Class, subclasses are neither necessary, useful, nor appropriate here. . . . Such rigid formalism, which would produce enormous obstacles to negotiating a class settlement with no apparent benefit, is not required and could even reduce the negotiating leverage of the class.”).

51

percentage method requires a district court to consider a number of relevant factors called the Johnson factors in order to determine if the requested percentage is reasonable. See id. at 772 & n.3, 775 (citing Johnson v. Ga. Highway Express, Inc., 488 F.2d 714 (5th Cir. 1974)).22 The District Court found the request for $77.5 million in fees was 20.36 percent of the $380.5 million common settlement fund and found this percentage was reasonable based on the Johnson factors. And while noting it was not required to do so, the District Court also used the “lodestar method” as a “cross-check on the reasonableness of a percentage-based fee” and found that “the requested fee easily passes muster if a cross-check is done.”

Two Objectors, Mr. Davis and Mr. West, challenge the District Court’s decision. We address their concerns in turn, reviewing de novo the proper standard

22 The Johnson factors include 12 factors from the opinion itself:

(1) the time and labor required; (2) the novelty and difficulty of the questions involved; (3) the skill requisite to perform the legal service properly; (4) the preclusion of other employment by the attorney due to acceptance of the case; (5) the customary fee; (6) whether the fee is fixed or contingent; (7) time limitations imposed by the client or the circumstances; (8) the amount involved and the results obtained;

(9) the experience, reputation, and ability of the attorneys; (10) the “undesirability” of the case; (11) the nature and the length of the professional relationship with the client; (12) awards in similar cases.

Camden I, 946 F.2d at 772 & n.3. The Johnson factors also include a handful of additional factors this Court added in Camden I: “the time required to reach a settlement, whether there are any substantial objections by class members or other parties to the settlement terms or the fees requested by counsel, any non-monetary benefits conferred upon the class by the settlement, and the economics involved in prosecuting a class action.” Id. at 775. The Objectors challenging the District Court’s decision to award attorney’s fees do not directly challenge its application of the factors, so we do not undertake a complete review of the factors.

52

for attorney’s fee awards and reviewing the District Court’s decision to award attorney’s fees for abuse of discretion. Loggerhead Turtle v. Cmty. Council of Volusia Cnty., 307 F.3d 1318, 1322 (11th Cir. 2002).

Mr. Davis makes two arguments. First, he says the District Court applied the wrong standard and should have applied the lodestar method,23 not the percentage method, in determining how much to award in attorney’s fees. In his view, Camden I’s percentage method is no longer good law, as he argues the Supreme Court’s decision in Perdue v. Kenny A. ex rel. Winn, 559 U.S. 542, 130 S. Ct. 1662 (2010), abrogated Camden I. Perdue is different, however, because it is a case in which the Supreme Court applied the lodestar method to claims made under a fee-shifting statute. Id. at 546, 130 S. Ct. at 1669. Second, Mr. Davis argues Camden I and the percentage method are “at odds” with a handful of other Supreme Court cases that “essentially” applied the lodestar method.

Mr. Davis’s first argument is foreclosed by our precedent. It is undisputed that this case involves a common settlement fund and in NPAS Solutions this Court expressly held that Camden I is “good law” in common fund cases and that “Perdue didn’t abrogate Camden I.” 975 F.3d at 1262 n.14; see also In re Home Depot Inc., 931 F.3d 1065, 1084–85 (11th Cir. 2019) (“There is no question that

23 Under the lodestar method, a district court determines the number of hours worked by plaintiffs’ counsel, multiplies those hours by a reasonable hourly rate, and then adjusts the final amount upward or downward based on various factors. Camden I, 946 F.2d at 772.

53

the Supreme Court precedents stretching from Hensley to Perdue are specific to fee-shifting statutes. . . . Thus, these precedents are not binding outside of the statutory context. For this reason, we have held that the Supreme Court precedent requiring the use of the lodestar method in statutory fee-shifting cases does not apply to common-fund cases.”). Indeed, although Perdue applied the lodestar method, it involved “the calculation of an attorney’s fee[] under federal fee- shifting statutes” and was based on the Supreme Court’s “prior decisions concerning the federal fee-shifting statutes.” 559 U.S. at 546, 552, 130 S. Ct. at 1669, 1672.24 Nothing in Perdue considered the appropriate method for calculating attorney’s fees in a common fund case. The percentage method therefore remains the proper method to apply when awarding attorney’s fees in common fund settlement cases.

Neither are we persuaded by Mr. Davis’s argument that Camden I and the percentage method are “at odds” with a handful of Supreme Court cases that “essentially” applied the lodestar method. The Supreme Court has never

24 That this case at one point included a claim under a fee-shifting statute (the Fair Credit Reporting Act) is of no consequence. For one thing, the District Court dismissed that claim before the parties settled this litigation. Beyond that, the parties’ settlement involved a common fund settlement, and “[w]here there has been a settlement, the basis for the statutory fee award has been discharged, and it is only the fund that remains.” Home Depot, 931 F.3d at 1082 (quotation marks omitted) (quoting Brytus v. Spang & Co., 203 F.3d 238, 246 (3d Cir. 2000)); see also, e.g., Florin v. Nationsbank of Ga., N.A., 34 F.3d 560, 563 (7th Cir. 1994) (holding a fee-shifting statute “do[es] not purport to control fee awards in cases settled with the creation of a common fund”).

54

categorically prohibited the percentage method in common fund cases. See Blum v. Stenson, 465 U.S. 886, 900 n.16, 104 S. Ct. 1541, 1550 n.16 (1984) (noting that for “the calculation of attorney’s fees under the ‘common fund doctrine’” the “reasonable fee is based on a percentage of the fund bestowed on the class”); Goldberger v. Integrated Res., Inc., 209 F.3d 43, 49 (2d Cir. 2000) (noting that Blum “provided all the impetus needed for a rejuvenation of the percentage method,” as “Blum indicates that the percentage-of-the-fund method is a viable” approach to calculating attorney’s fees in common fund cases (quotation marks omitted)). To the contrary, and as Mr. Davis acknowledges, the Supreme Court has applied the percentage method in common fund cases. Without a categorical prohibition on the percentage method in common fund settlement cases, Camden I and the percentage method remain the law in this Circuit.25 Mr. West’s arguments also fail. 26 According to Mr. West, the District Court should have considered the “economies of scale” in this case, which involves a

25 Likewise, to the extent Mr. Davis challenges the percentage awarded here based on percentages awarded in some Supreme Court cases, he fails to cite a single case in which the Supreme Court set a categorical ceiling for a reasonable percentage.

26 One argument we can reject out of hand. Mr. West says the District Court should have provided Plaintiffs’ counsel’s “lodestar material” to the class members. Mr. West faults the District Court for only providing “total hours and lodestar accumulated by the firms.” While the District Court did conduct a lodestar “cross-check on the reasonableness of a percentage-based fee,” our precedent did not require it to do so. See Camden I, 946 F.2d at 774 (“[A]ttorneys’ fees awarded from a common fund shall be based upon a reasonable percentage of the fund established for the benefit of the class. The lodestar analysis shall continue to be the applicable method used for determining statutory fee-shifting awards.”). And because the District Court

55

settlement fund of hundreds of millions of dollars (what West calls a “megafund” case). As we understand Mr. West’s position, he thinks a settlement that is ten times larger than another settlement is often not ten times harder for the lawyers to work on, such that the percentage awarded as attorney’s fees should diminish as the settlement amount gets larger.

As Mr. West admits, our precedent did not require the District Court to expressly consider the economies of scale in a megafund case in deciding how much to award in attorney’s fees. This Court required the District Court to consider the Johnson factors, and none of the factors explicitly address the economies of scale in a megafund case. See Camden I, 946 F.2d at 772 & n.3, 775. That being the case, we cannot say the District Court erred as a matter of law or abused its discretion. In any event, the District Court considered the time, labor, and amount involved and the results obtained when deciding whether the attorney’s fees were reasonable. We observe that these factors fairly capture many considerations related to the economies of scale in a megafund case.

We decline to add an additional factor requiring the District Court to expressly consider the economies of scale in a megafund case. See id. at 775 (observing that the “factors which will impact upon the appropriate percentage to

was not required to do a lodestar cross-check, we cannot say it erred in not providing the “lodestar material” to the class members.

56

be awarded as a fee in any particular case will undoubtedly vary”). For starters, we question the value of this consideration. Such a factor may “lack[] rigor because it provides no direction to courts about when to start decreasing the percentage award, nor by how much.” 5 Newberg § 15:80. Requiring consideration of the economies of scale could also create “perverse incentives,” as it may encourage class counsel to pursue “quick settlements at sub-optimal levels.” Id.; see also 5 Newberg § 15:81 (detailing the “rough justice” of limiting attorney’s fees in megafund cases). But we need not (and do not) ultimately decide the virtues (or vices) of such a factor because the District Court ultimately considered factors that reasonably capture many considerations related to the economies of scale in a megafund case.

Finally, to the extent Mr. West challenges the amount of attorney’s fees awarded in this case, he has failed to show an abuse of discretion. The District Court awarded $77.5 million in attorney’s fees, which it found is 20.36 percent of the $380.5 million settlement fund. (We note that the $380.5 million figure does not even account for the additional funds Equifax may be required to pay into the settlement fund.) 20.36 percent is well within the percentages permitted in other common fund cases, and even in other megafund cases. 27 See Camden I, 946 F.2d

27 Also, we continue to note that our Circuit does not limit attorney’s fees in megafund cases as a matter of law.

57

at 774–75 (“The majority of common fund fee awards fall between 20% to 30% of the fund,” with 25 percent as “a ‘bench mark’ percentage fee award.”); In re Checking Account Overdraft Litig., 830 F. Supp. 2d 1330, 1367 (S.D. Fla. 2011) (“[C]ourts nationwide have repeatedly awarded fees of 30 percent or higher in so- called ‘megafund’ settlements.”); Manual for Complex Litigation § 14.121 (“Attorney fees awarded under the percentage method are often between 25% and 30% of the fund.”); see also Wolff v. Cash 4 Titles, 2012 WL 5290155, at *5 (S.D. Fla. Sept. 26, 2012) (average percentage award in the Eleventh Circuit is “roughly one-third”); In re Anthem, Inc. Data Breach Litig., 2018 WL 3960068, at *2, *15 (N.D. Cal. Aug. 17, 2018) (data breach case involving 27 percent award, which “appears to be in line with the vast majority of megafund settlements”).28 The District Court thus properly applied the percentage method in this common fund settlement case, considered the appropriate factors, and did not abuse its discretion in the amount it awarded in attorney’s fees.

28 Mr. West says we should view the $77.5 million in attorney’s fees as 25 percent of a $310 million settlement fund (not $380.5 million) because that is what Plaintiffs’ counsel initially secured and agreed to in the original term sheet with Equifax. The additional $70.5 million of the $380.5 million fund was added after further negotiations with regulators. But even assuming Plaintiffs’ counsel did not play a role in securing the additional funds, such that 25 percent really is the operative figure, 25 percent is also well within percentages approved in other common fund cases. Camden I, 946 F.2d at 775 (25 percent is the “bench mark”). And again, the $310 million figure does not account for the additional funds Equifax may be called upon to pay into the settlement fund in the future.

58

G. Incentive Awards After awarding attorney’s fees and expenses to Plaintiffs’ counsel, the District Court approved incentive awards (sometimes called service awards) for the class representatives in order to compensate them for their services and the risks they incurred on behalf of the class. The District Court recognized that courts “routinely approve” such awards, and it found the awards “deserved” in this case because the class representatives “devoted substantial time and effort to this litigation working with their lawyers to prosecute the claims, assembling the evidence supporting their claims, and responding to discovery requests.” “But for their efforts,” the District Court found, “other class members would be receiving nothing.”

While the parties were briefing this case, a panel of this Court recognized that incentive awards are “commonplace in modern class-action litigation,” yet held that two Supreme Court cases from the 1880s “prohibit the type of incentive award that the district court approved here—one that compensates a class representative for his time and rewards him for bringing a lawsuit.” NPAS Sols., 975 F.3d at 1260. In light of NPAS Solutions, Plaintiffs acknowledge that “service awards are prohibited as a matter of law” in this Circuit. It is true that NPAS Solutions binds us here. So the question is how to proceed. Mr. Davis says the “incentive awards likely compromised Named Plaintiffs’ representation of the

59

Settlement Class’s interests” because the class representatives might have been “tempted to accept [a] suboptimal settlement[]” in order to obtain the incentive awards.

We reject Mr. Davis’s view that the prospect of incentive awards infected the entire settlement, so we decline his invitation to vacate the settlement as a whole. The record indicates the class representatives’ representation of the class was not affected by the possibility of receiving incentive awards. The settlement agreement expressly stated that it remained in effect even if the District Court declined to approve the incentive awards. Plaintiffs likewise filed two separate motions in the District Court: one for approval of the settlement and certification of the class, and one for attorney’s fees and expenses and incentive awards for class representatives. The motion for approval of the settlement was not contingent on the District Court approving the incentive awards. Given these facts, the class representatives’ decision to agree to the settlement and to seek its approval was not influenced by the possibility of receiving incentive awards. Cf. Radcliffe v. Experian Info. Sols. Inc., 715 F.3d 1157, 1164–67 (9th Cir. 2013) (invalidating settlement that “explicitly condition[ed] the incentive awards on the class representatives’ support for the settlement”).

Plaintiffs argue that the best approach is to simply reverse the District Court’s decision approving the incentive awards and remand solely for the limited

60

purpose of vacating the awards, and we agree. This approach is administratively feasible, as the settlement agreement expressly provides that the agreement does not terminate due to “modification or reversal or appeal of any decision by the Court[] concerning the amount of Service Awards.” And this approach makes the most sense as a matter of judicial economy. Specifically, as set out above in great detail, the District Court, before ever approving incentive awards, independently assessed the proposed settlement and the class and did not abuse its discretion in finding that the settlement was fair, reasonable, and adequate and that the class representatives adequately represented the class. No purpose would be served by forcing the District Court to repeat the entire process anew.

To be clear, this is the only issue on which we reverse the District Court’s decision. On remand, the District Court is instructed to vacate the incentive awards and to otherwise leave the settlement agreement intact. We expect the District Court will be wary of any attempts to expand this mandate or to otherwise delay or prevent the settlement from taking effect, and we encourage that approach. H. Appeal Bonds At last, we arrive at the final issue in this case. After the Objectors appealed the District Court’s approval order, the court granted Plaintiffs’ motion for appeal bonds and imposed appeal bonds of $2,000 on each Objector. The District Court

61

noted that “[c]ourts routinely require objectors who appeal final approval of a class action settlement to post a bond to ensure payment of costs on appeal.” The District Court considered a number of factors and found an appeal bond of $2,000 was “appropriate.” Mr. Cochran and Mr. Frank challenge the District Court’s decision to impose the appeal bonds for a variety of reasons.

Federal Rule of Appellate Procedure 7 states, “[i]n a civil case, the district court may require an appellant to file a bond or provide other security in any form and amount necessary to ensure payment of costs on appeal.” Fed. R. App. P. 7. 29 Mr. Cochran, Mr. Frank, and Plaintiffs do not cite any cases from this Court concerning when an appeal bond is permitted under Rule 7, nor have we found any ourselves. Although our Court has addressed when attorney’s fees under a fee- shifting statute can be included in an appeal bond as “costs” under Rule 7, those cases don’t deal with the issue presented here: when a run-of-the-mill appeal bond is permitted. See Young v. New Process Steel, LP, 419 F.3d 1201, 1207–08 (11th Cir. 2005); Pedraza v. United Guar. Corp., 313 F.3d 1323, 1333 (11th Cir. 2002).

Although we review a district court’s decision to impose appeal bonds for abuse of discretion, we review de novo the proper interpretation of federal rules of procedure, including Rule 7. Young, 419 F.3d at 1203; see also SEB S.A. v.

29 Although Plaintiffs moved for appeal bonds under Rules 7 and 8, it appears the District Court imposed them only pursuant to Rule 7. As such, our discussion of this issue is limited to appeal bonds imposed under Rule 7.

62

Sunbeam Corp., 476 F.3d 1317, 1319 (11th Cir. 2007). In interpreting a federal rule, we examine its text and give effect to its plain meaning. See Sargeant v. Hall, 951 F.3d 1280, 1283 (11th Cir. 2020) (interpreting the Federal Rules of Civil Procedure). The plain text of Rule 7 is clear. Again, Rule 7 says a district court may impose an appeal bond “in any form and amount necessary to ensure payment of costs on appeal.” Fed. R. App. P. 7 (emphasis added). The word “ensure” means “to make sure, certain, or safe.” Ensure, Merriam-Webster’s Unabridged Dictionary, https://unabridged.merriam-webster.com/unabridged/ensure (last visited June 2, 2021). Therefore, a Rule 7 appeal bond is appropriate when the bond is imposed to make sure costs on appeal are paid. See also, e.g., 4 Newberg § 14:15 (risk of nonpayment is “arguably the only pertinent factor” and thus “ought to be the primary focus” for appeal bonds); 16A Catherine T. Struve, Federal Practice and Procedure § 3953 (5th ed. 2021) (“The court should require a bond only if ‘necessary to ensure payment of costs on appeal.’”).

The District Court in this case considered several factors when deciding to impose the appeal bonds: “(1) the appellant’s financial ability to post a bond; (2) the merits of the appeal; (3) whether the appellant has shown any bad faith or vexatious conduct; and (4) the risk that the appellant will not pay the costs if the appeal is unsuccessful.” Other courts in this Circuit have applied similar factors. See, e.g., Aboltin v. Jeunesse LLC, 2019 WL 1092789, at *3 (M.D. Fla. Feb. 15,

63

2019). While most of these factors do not appear to be relevant based on our reading of Rule 7, the last factor certainly is. Specifically, the District Court considered “the risk that the appellant will not pay the costs if the appeal is unsuccessful” and found there was a “substantial risk that the costs of appeal will not be paid unless a bond is required.” This consideration squares with Rule 7: if there a risk the appellant will not pay the costs on appeal, then an appeal bond helps make sure the costs are paid.

Although the District Court considered other factors that may not be relevant, we need not ultimately decide this issue of relevance under Rule 7 because the record indicates the District Court independently imposed the appeal bonds based on a proper factor. Specifically, the District Court found that the “substantial risk” of nonpayment “warrant[ed] an appeal bond.” The District Court did not therefore abuse its discretion when it imposed the appeal bonds based on its finding that there was a “substantial risk that the costs of appeal will not be paid unless a bond is required.” And because we hold that the District Court did not abuse its discretion in imposing the appeal bonds on this independent basis, we need not consider Mr. Cochran’s and Mr. Frank’s arguments, which challenge the District Court’s ruling to the extent it was also based on other grounds.30

30 Mr. Cochran appears to briefly challenge the amount of the appeal bond. Rule 7 simply states that an appeal bond should be in an “amount necessary to ensure payment of costs on appeal.” Fed. R. App. P. 7. This means courts should look to 28 U.S.C. § 1920 and Federal

64

III. CONCLUSION

We affirm the District Court’s rulings in their entirety, except as to the narrow issue of incentive awards. As discussed, because NPAS Solutions now prohibits the incentive awards approved for the class representatives, we must reverse the District Court’s decision to approve the incentive awards. We remand this case to the District Court solely for the limited purpose of vacating those awards. On remand, we instruct the District Court to leave the rest of the settlement agreement intact.

AFFIRMED IN PART, REVERSED IN PART, AND REMANDED.

Rule of Appellate Procedure 39(e), which provide for taxable costs on appeal, when determining an appropriate amount to cover “costs on appeal.” See, e.g., Tennille v. W. Union Co., 774 F.3d 1249, 1257 (10th Cir. 2014) (looking to section 1920 and Rule 39(e)); see also Federal Practice and Procedure § 3953 (“Costs on appeal for which a Rule 7 bond can be required include the costs authorized in [section] 1920 to the extent those costs relate to the appeal.”); 4 Newberg § 14:16 (“Given that an appeal bond is meant to ensure the availability of funds for cost-shifting on appeal, the amount of the bond should have some relationship to the costs that a losing appellant would have to shoulder.” (footnote omitted)). The costs under section 1920 and Rule 39(e) include fees of the clerk, fees for printed or electronically recorded transcripts, fees for printing, costs of making copies, docket fees, costs of preparing and transmitting the record, and costs for the reporter’s transcript. See 28 U.S.C. § 1920; Fed. R. App. P. 39(e). Here, the District Court found that $2,000 was “appropriate” to cover the taxable costs listed in section 1920 and Rule 39(e). Mr. Cochran has not shown that this amount was an abuse of discretion.