In re: CALIFORNIA PIZZA KITCHEN DATA BREACH LITIGATION
No. 23-55288
UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT
February 24, 2025
D.C. Nos. 8:21-cv-01928-DOC-KES, 2:21-cv-09578-DOC-KES
FOR PUBLICATION
AVIVA KIRSTEN, Plaintiff-Appellant, v. KANSAS GILLEO, Plaintiff-Appellee, JEREMY PITTMAN, individually and on behalf of all others similarly situated, Plaintiff-Appellant, SYDNEY RUSEN; ESTEBAN MORALES; DOUG WALLACE; BRETT RIGAS; EVENCIO DIAZ, individually and on behalf of all others similarly situated, Plaintiffs-Appellees, v. CALIFORNIA PIZZA KITCHEN, INC., Defendant-Appellee.
David O. Carter, District Judge, Presiding
Argued and Submitted June 3, 2024
Pasadena, California
Filed February 24, 2025
Before: Richard R. Clifton, Daniel P. Collins, and Kenneth K. Lee, Circuit Judges.
Opinion by Judge Lee;
Partial Concurrence and Partial Dissent by Judge Collins
SUMMARY*
Class Settlement / Attorneys’ Fees
The panel affirmed the district court’s approval of a class settlement, reversed the attorneys’ fee award, and remanded in a class action brought by California Pizza Kitchen, Inc. (CPK) employees whose personal information was compromised by a cyberattack.
One group of plaintiffs’ lawyers struck a settlement with CPK. The monetary value of the class’s claims were (at most) around $950,000, yet the attorneys sought $800,000 in fees. The district court approved the settlement.
The panel held that district courts may approve claims-made settlements—even those that raise indicia of collusion—so long as they adhere to procedural requirements and find the settlement “fair, reasonable, and adequate” under
The panel reversed the fee award because the district court did not assess the actual value of the settlement and compare it to the fees requested. The panel remanded for the district court to determine the settlement’s actual value to class members and award reasonable and proportionate attorneys’ fees, consistent with this opinion.
Judge Collins concurred in the judgment to the extent that the majority reversed and remanded the district court’s approval of the fee award. He dissented from the majority’s decision to affirm the approval of the underlying settlement because, in approving the final settlement proposal before class certification, the district court provided little explanation as to why it approved this settlement and instead issued a series of perfunctory orders, despite the fact that (1) the final settlement triggers every Bluetooth factor; (2) the settlement’s final value ended up being nearly a fourth of the estimated “conservative” value presented at the preliminary approval hearing; and (3) the settlement’s proposed fee award comprises nearly 46% of the entire settlement.
COUNSEL
Theodore W. Maya (argued), Christopher Stiner, and Tina Wolfson, Ahdoot & Wolfson PC, Burbank, California; Todd S. Garber and Andrew C. White, Finkelstein Blankinship Frei-Pearson and Garber LLP, White Plains, New York; Seth A. Meyer, Meyer Law Firm PLLC, Scottsdale, Arizona; for Plaintiffs-Appellants.
David K. Lietz (argued), Milberg Coleman Bryson Phillips Grossman PLLC, Washington, D.C.; Mason Barney, Siri & Glimstad LLP, New York, New York; Rachele R. Byrd, Wolf Haldenstein Adler Freeman & Herz LLP, San Diego, California; Daniel O. Herrera, Cafferty Clobes Meriwether & Sprengel LLP, Chicago, Illinois; for Plaintiffs-Appellees.
Jon P. Kardassakis (argued) and Michael K. Grimaldi, Lewis Brisbois Bisgaard & Smith LLP, Los Angeles, California, for Defendant-Appellee.
OPINION
LEE, Circuit Judge:
California Pizza Kitchen, Inc. (CPK) is a restaurant chain offering California-style pizza at about two-hundred locations across the country. But in November 2021, CPK was not in the news for its trademark Original BBQ Chicken Pizza or its underrated Thai Chicken Pizza. Rather, CPK revealed that a cyberattack had compromised the personal information of over 100,000 former and current employees. That disclosure spurred lawyers to race to the courthouse and file competing class action lawsuits against CPK to get a slice of the action.
One group of plaintiffs’ lawyers quickly struck a settlement with CPK: the deal offered cash payments and credit monitoring services to class members but CPK would only be required to make payments to class members who submitted valid claims (i.e., a claims-made settlement). Given the low redemption rate for claims, the monetary value of the class’s claims is (at most) around $950,000—yet the attorneys sought $800,000 in fees. A competing group of plaintiffs’ lawyers challenged that settlement, contending that they could deliver a deal for the class that would top it. Despite expressing some reservations, the district court approved the settlement.
We affirm the settlement approval but remand the attorneys’ fee award. District courts may approve claims-made settlements—even those that raise indicia of collusion—so long as they adhere to procedural requirements and find the settlement “fair, reasonable, and adequate” under Rule 23(e). Our review of the record shows
BACKGROUND
I. CPK suffers a cyberattack—and then a deluge of putative class actions.
In September 2021, CPK fell prey to a cyberattack by Conti, a ransomware group. Conti’s business model is straightforward: it hacks into a company’s system, encrypts the company’s files, and leaves a ransom note. In exchange for payment, Conti provides its victims with a decryption key and promises not to leak the stolen data. Conti has an incentive to keep those promises—if it started releasing stolen data, then future victims would be disinclined to pay.
Faced with this sticky dilemma, CPK chose to pay the ransom. Because the breach compromised employee data (including Social Security numbers), CPK notified 103,767 of its former and current employees. Soon, numerous plaintiffs’ lawyers filed a flurry of lawsuits—five in a matter of weeks.1 All these class action lawsuits alleged that CPK failed to safeguard its employees’ personally identifiable information, and brought claims for negligence, breach of implied contract, and violations of business and privacy statutes.
Counsel for the first four cases filed (Gilleo, Morales, Wallace, and Rigas) quickly agreed to cooperate with each other in hopes of splitting the pie among themselves. They made overtures to the Kirsten plaintiffs, trying to bring them into the fold. But those attempts went nowhere: the Kirsten plaintiffs had little appetite for cooperation, as their counsel had divergent views on case strategy and jockeyed for the position of lead counsel. The district court eventually granted a stipulation consolidating Gilleo, Morales, Wallace, and Rigas into a single action—In re California Pizza Kitchen Data Breach Litigation. The Kirsten plaintiffs trundled on alone.
II. CPK and the consolidated plaintiffs reach a deal.
CPK and the consolidated plaintiffs proceeded straight into mediated settlement negotiations. In March 2022, they jointly requested that the district court stay the action pending finalization of a proposed settlement agreement.2 Two months later, CPK and the consolidated plaintiffs presented a proposed settlement to the district court for preliminary approval. The proposed settlement class included all former and current employees potentially affected by CPK’s data breach, with a subclass for all class members residing in California. The class agreed to release all claims arising out of the data breach
First, CPK agreed that all class members could claim: (1) up to $1,000 in reimbursement for ordinary expenses and lost time incurred because of the data breach; (2) up to $5,000 in compensation for monetary loss from actual identity theft; and (3) 24-months’ worth of credit monitoring services, which included $1 million in identity theft insurance. Members of the California subclass could claim another $100 statutory damages award, though that amount would be counted towards the $1,000 reimbursement cap for ordinary expenses. The settlement agreement did not set an aggregate cap on these benefits, so CPK could theoretically be on the hook for massive liability. But the agreement was structured as a claims-made settlement, so class members would only recover if they submitted a timely claims form with supporting documentation—and the financial cost of the settlement to CPK would rest on the number of valid claims.
Second, CPK committed to limited injunctive relief. Right after the cyberattack, CPK had engaged a third-party cybersecurity consultant to provide post-breach “forensics, recovery, and remediation.” As part of the settlement agreement, CPK agreed to “maintain certain recently implemented business practices and remedial measures” for three years after the settlement date.
Finally, the settlement agreement permitted class counsel to request—without CPK’s objection—up to $800,000 in attorneys’ fees and costs. Those fees would be paid by CPK separately from any money disbursed to class members.
The consolidated plaintiffs estimated that, in aggregate, this settlement had “a conservative value of over $3.7 million.” By their calculations, if only 2% of the class—which had around 103,767 members—claimed the full $1,000 in ordinary expense reimbursements, then the settlement would be worth $2,075,340. If only 5% of the California subclass—which had around 30,781 members—claimed the statutory damages amount, then that value would increase by $153,905. And if only 4% of the class claimed the credit monitoring service, which the consolidated plaintiffs valued at $15 per month per person, then another $1,494,244.80 could be stacked onto the settlement’s value.
Counsel for the Kirsten plaintiffs objected to the settlement as collusive and the fees request as excessive. It asked the court to appoint them as lead counsel, claiming that they could negotiate a better settlement for the class.
III. The district court separately approves the settlement and attorneys’ fees.
In June 2022, the district court convened a preliminary approval hearing. At the hearing, the district court extensively analyzed the proposed settlement. It invited Bruce Friedman, the private mediator who oversaw the settlement negotiations, to testify and probed him on the parties’ negotiation process. The district court questioned the consolidated plaintiffs on the many vigorous objections raised by the Kirsten plaintiffs. And it raised its own concerns about whether the settlement sufficiently addressed the class members’ injuries.
The district court preliminarily approved the proposed settlement as “adequate, fair, and reasonable.” But it declined to award attorneys’ fees without knowing the number of submitted claims, and reserved that issue until after the claims process had concluded.
Once the claims deadline had passed, the consolidated plaintiffs returned to seek final
- 176 ordinary expense claims totaling $384,134.77;
- 979 lost time claims totaling $50,320.00;
- 45 identify theft claims totaling $191,354.50;
- 803 California statutory damages claims totaling $80,300.00; and
- 1,264 claims for credit monitoring services, valued at $455,040.00.
They represented that the class had filed 1,828 claims with a face value of $1,161,149.27, for a final claims rate of 1.8%. These claims were unvalidated, meaning that the claims administrator had not yet determined whether they were properly documented and compliant with the settlement’s terms. Even without validation, one glaring error is apparent on its face: because of the $1,000 cap on ordinary expense claims, the 176 ordinary expense claims are capped at $176,000, despite the class members asking for $384,134.77. That means the maximum monetary value of the claims is, at most, around $950,000, not $1.16 million. No matter the final value of the settlement, the consolidated plaintiffs requested $773,632.95 in attorneys’ fees and $26,367.05 in expenses.
The district court observed that the settlement value had dramatically deflated from the consolidated plaintiffs’ earlier estimate, and expressed “tremendous concern” over the scope of attorneys’ fees. After holding two hearings in November and December 2022, the district court issued final approval of the settlement in a sparse written order. It awarded the full $800,000 in attorneys’ fees and costs, noting that attorneys’ fees “would constitute 36.3% of the total class benefit, which is $2,133,719.”3 The written order did not explain the court’s rationale for approving the settlement and the fees.
The Kirsten plaintiffs timely appealed. We have jurisdiction under
STANDARD OF REVIEW
We review a district court’s decision to approve a class action settlement for abuse of discretion. Briseño v. Henderson, 998 F.3d 1014, 1022 (9th Cir. 2021). Our review of a district court’s substantive fairness determination is narrow: we “rarely overturn” settlements for substantive reasons but will do so if the agreement contains “convincing indications” that “self-interest rather than the class’s interests in fact influenced the outcome of the negotiations.” Allen v. Bedolla, 787 F.3d 1218, 1223 (9th Cir. 2015) (citation omitted). Given these substantive limitations, we hold district courts to a stringent procedural standard. Id. “To survive appellate review, the district court must show it has explored comprehensively all factors” relevant to the substantive fairness determination. Dennis v. Kellogg Co., 697 F.3d 858, 864 (9th Cir. 2012) (citation omitted).
We review a district court’s award of attorneys’ fees for abuse of discretion, and the factual findings supporting such an award for clear error. Lowery v. Rhapsody Int’l, Inc., 75 F.4th 985, 991 (9th Cir. 2023).
I. The district court did not abuse its discretion in approving the class settlement.
A. Due process concerns require judicial approval of class action settlements.
Class action settlements, by their very nature, present “unique due process concerns for absent class members.” Hanlon v. Chrysler Corp., 150 F.3d 1011, 1026 (9th Cir. 1998).
Before 2018, this circuit instructed courts engaging in Rule 23(e)’s “fair, reasonable, and adequate” analysis to consider the following non-exhaustive list of factors:
[1] the strength of plaintiffs’ case; [2] the risk, expense, complexity, and likely duration of further litigation; [3] the risk of maintaining class action status throughout the trial; [4] the amount offered in settlement; [5] the extent of discovery completed, and the stage of the proceedings; [6] the experience and views of counsel; [7] the presence of a governmental participant; and [8] the reaction of the class members to the proposed settlement.
Staton, 327 F.3d at 959 (citation omitted) (known as the “Hanlon,” “Staton,” or “Churchill,” 361 F.3d 566 (9th Cir. 2004), factors).
In 2018, Rule 23(e) was amended to prescribe its own multi-factor test. To determine whether a settlement is “fair, reasonable, and adequate,” courts should consider whether:
- the class representatives and class counsel have adequately represented the class;
- the proposal was negotiated at arm’s length;
- the relief provided for the class is adequate, taking into account:
- the costs, risks, and delay of trial and appeal;
- the effectiveness of any proposed method of distributing relief to the class, including the method of processing class-member claims;
- the terms of any proposed award of attorney’s fees, including timing of payment; and
- any agreement required to be identified under Rule 23(e)(3); and
- the proposal treats class members equitably relative to each other.
Another concern lurks in the background of every class action settlement: “class counsel may collude with the defendants, tacitly reducing the overall settlement in return for a higher attorney’s fee.” In re Bluetooth Headset Prods. Liab. Litig., 654 F.3d 935, 946 (9th Cir. 2011) (internal quotation omitted). In Bluetooth, we identified three “subtle signs that class
If these indicia of implicit collusion are present, then a proposed settlement “must withstand an even higher level of scrutiny for evidence of collusion or other conflicts of interest than is ordinarily required under Rule 23(e).” Bluetooth, 654 F.3d at 946. The district court must especially “assure itself that the fees awarded in the agreement were not unreasonably high.” Id. 947 (internal citations and alterations omitted).
But before we can review the district court’s settlement approval order, we first must address a unique wrinkle here: the district court’s preliminary and final approval orders were sparse—almost boilerplate—and memorialized little of the district court’s rationale. Usually, when we cannot determine the district court’s reasoning, we remand the case for the court to explain its rationale. But here, we can reasonably infer the district court’s rationale from the record, which was unusually extensive with multiple hearings and supplemental briefs. See, e.g., Hanlon, 150 F.3d at 1023 (“Although the district court’s [class certification] findings are almost conclusory, the record provides more than adequate foundation upon which to reach our conclusions.”); In re Pac. Enters. Secs. Litig., 47 F.3d 373, 377 (9th Cir. 1995) (the district court’s “conclusory statement[s]” could be supported by the record, which reflected an “extensive settlement hearing”).4 From the district court’s exhaustive questions and statements, as well as the parties’ responses, at the various hearings, we can reasonably deduce what factors the district court considered when approving the settlement.
B. The district court applied the Bluetooth heightened standard to review the settlement for collusion.
We now address whether the district followed the Bluetooth procedural safeguards in approving a class settlement. The Kirsten plaintiffs correctly point out that the settlement raises all three red flags of potential collusion under Bluetooth. 654 F.3d at 947.
First, Bluetooth warns that a class counsel receiving a disproportionately large fee award compared to what the class members received signals potential collusion. See Briseño, 998 F.3d at 1026 (a “gross disparity in distribution of funds between class members and their class counsel raises an urgent red flag”). Here, the parties agreed to $800,000 in fees, even though the
Second, the parties agreed to a so-called “clear sailing” provision in which CPK agreed that it would not challenge the agreed-upon attorneys’ fees. See Bluetooth, 654 F.3d at 947 (these provisions potentially “enabl[e] a defendant to pay class counsel excessive fees and costs in exchange for counsel accepting an unfair settlement on behalf of the class”).
And third, if a settlement has a “reverter” or a “kicker” provision, it could imply collusion. See Bluetooth, 654 F.3d at 949 (reverter provisions “amplif[y] the danger of collusion already suggested by a clear sailing provision”). A reverter in a class action settlement means that any fees not awarded to class counsel would revert to the defendant’s coffers, and not to the class members. Here, it means that if the district court had reduced fees below $800,000, then CPK’s obligation to pay attorneys’ fees would be reduced, with no additional funds given to the class.
CPK contends that claims-made settlements do not contain reverter provisions, and so Bluetooth cannot apply. In a claims-made settlement, the defendant waits until after the claims process concludes—i.e., class members have submitted forms to seek compensation and the claim administrator validates them—to pay out the amount claimed by class members. By contrast, in a common-fund settlement, the defendant establishes a pot of money (i.e., a “common fund”) before the claims process begins. If a common-fund settlement includes a reversionary provision, then the defendant can recoup whatever amount is left in the fund after paying out class members’ claims. Because a claims-made settlement only requires CPK to pay out the amount claimed, the argument goes, no money reverts to CPK, so there technically can be no reverter provision and thus no Bluetooth scrutiny.
That argument fundamentally misunderstands Bluetooth’s purpose. The reverter provision flagged in Bluetooth addresses what happens to unawarded attorneys’ fees. 654 F.3d at 947. In a class action settlement, the defendant has no incentive to differentiate between (i) payments made to class members and (ii) payments made to class counsel—and is only concerned about minimizing the total payment. See, e.g., Staton, 327 F.3d at 964; Briseño, 998 F.3d at 1025. Put another way, a defendant will happily pay more to class counsel at the expense of the class if it means that its total liability—i.e., attorneys’ fees and payment to the class—will be lower. See Briseño, 998 F.3d at 1025.
So when class counsel and the defendant include clear sailing and reverter provisions in the settlement, they might be benefiting themselves—at the expense of the class members. The district court may ultimately award less than the agreed-upon attorneys’ fees, despite a clear sailing provision in which the defendant agrees not to challenge it. But if the court does so, then the unawarded fees would ideally be distributed to the class, as the defendant is already willing to pay that amount to secure the settlement. See Briseño, 998 F.3d at 1027 (“[T]here is no plausible reason why the class should not benefit from the spillover of excessive [attorneys’] fees.”). Yet a reverter directs the unawarded fees to be given back to the defendant instead of the class members. Bluetooth directs courts to be skeptical of such settlements because they hint at collusion between the class counsel and the defendant.
This fear of collusion applies equally to both claims-made settlements and common-fund settlements. Whether money leaves the defendant’s hands is immaterial. In either case, the purpose of Bluetooth’s
Having established that Bluetooth applies, we are satisfied that the district court fulfilled its heightened obligation to ferret out any “evidence of collusion or other conflicts of interest.” 654 F.3d at 946. As we have recognized, the presence of all three Bluetooth factors does not trigger a domino effect that makes a settlement per se collusive. See Kim v. Allison, 8 F.4th 1170, 1180 (9th Cir. 2021) (noting that the presence of all three factors is not a “death knell”). Bluetooth is merely a procedural safeguard to ensure that district courts scrutinize class settlements for collusion that may harm class members. It does not set crusty, rigid rules that mandate a substantive result.
Here, the district court gave the settlement provisions a hard look under the Bluetooth factors. It probed the parties’ disputes over whether CPK’s decision to negotiate with the consolidated plaintiffs amounted to a “reverse auction,” whether the Kirsten plaintiffs’ counsel’s “effort to become class counsel” was improper, and whether a “claims-made settlement” was “inherently collusive.” It questioned Mr. Friedman, the parties’ private mediator, on the parties’ “term sheets” and “settlement proffer[s],” and how many negotiation sessions the parties conducted. In its preliminary approval order, it made an explicit finding that the settlement was “non-collusive.” On this record, we cannot say that the district court ignored the Bluetooth factors or abused its discretion.
C. The district court did not abuse its discretion in finding the settlement fair, reasonable, and adequate.
We next substantively review the class settlement and hold that the district court did not abuse its discretion in finding it fair, reasonable, and adequate. The district court properly evaluated the settlement under the
The district court found the relief for the class adequate after considering a host of concerns.
monitoring services for two years, up to $1,000 for ordinary expenses and lost time, and up to $5,000 for monetary loss from identity theft. The district court also considered the
On appeal, the Kirsten plaintiffs contend that they could have notched a more favorable settlement on behalf of the class. They maintain, for example, that the CCPA claims are possibly worth over $77 million in statutory damages, and that the settlement released them for only a small sliver of that amount. But the accuracy of that valuation is tenuous at best. The Kirsten plaintiffs provide no legal authority for their valuation of the CCPA claims and have not rebutted CPK‘s potential legal defenses against them.
The district court also weighed the “significant risks, expense, and/or uncertainty” that the class would face in litigation.
Lastly, we underscore that courts do not have a duty to maximize settlement value for class members. Briseno, 998 F.3d at 1026. Rather, our inquiry is much more modest and limited to ensuring that the class settlement is fair, reasonable, and adequate. Perhaps the Kirsten plaintiffs’ counsel could have obtained a better deal if they had litigated the case further. But reasonable disagreements on case strategy are not evidence of an unfair deal or collusion. An early settlement may often be beneficial for the class, as it reduces attorneys’ fees, preserves value for the class, and offers immediate compensation for injured class members.5
The most problematic part of the class settlement is its claims-made nature. As evidenced by the 1.8% claims rate here, redemption rates are typically very low because most class members do not bother jumping through the hoops to submit a claim. See, e.g., Roes, 1-2, 944 F.3d at 1053; Briseno, 998 F.3d at 1026. As a result, claims-made settlements often allow defendants to settle on the cheap. But we have never held that claims-made settlements are per se inadequate under
II. We reverse and remand the approval of excessive attorneys’ fees.
Finally, we address the district court‘s fee award of almost $800,000. Because the settlement agreement was not conditioned on attorneys’ fees, we can vacate
Courts may only award “reasonable” attorneys’ fees.
Unlike the district court‘s settlement approval, we cannot infer the district court‘s rationale for its fee award from the record. At the November 2022 hearing, the district court raised “tremendous concern[s]” about the reasonableness of the consolidated plaintiffs’ $800,000 fee request. First, it questioned the lodestar of $687,681. Although the district court noted that the blended hourly rate of $670 seemed reasonable, it asked about the sheer “number of attorneys allegedly making appearances” in the case. Second, the district court appeared disinclined to award fees “clearly above the benchmark of 25%” percentage-of-recovery.
Yet its final award did just that—even though the attorneys’ fees constitute around 45% of the settlement value to the class. In a short, two-paragraph order, the district court signed off on a positive multiplier to counsel‘s lodestar, even though the fees were significantly over the 25% percentage-of-recovery benchmark. The court failed to “provide an adequate explanation” for the fees, so we must reverse and remand. Lowery, 75 F.4th at 992 (citing Stanger v. China Elec. Motor, Inc., 812 F.3d 734, 739 (9th Cir. 2016)).
Even apart from its failure to provide an explanation, the district court erred by approving fees that appear excessive of settlement value. To assess the reasonableness of fees, the district court must first independently “calculate the class action settlement‘s benefit to the class members.” Id. In a claims-made settlement, that obligation requires the district court to wait until at least after the claims deadline to award attorneys’ fees. When the claims validation process is brief or where the settlement presents a significant risk of fraudulent or invalid claims (e.g., settlements offering potentially lucrative compensation), that may even require the district court to wait until after the claims validation process has concluded.
Here, the district court appeared to accept the consolidated plaintiffs’ assertion that the class‘s claims were worth $1.16 million, despite the substantial likelihood that many of these claims would be thrown out post-validation for exceeding the caps or failing to provide supporting documentation. As noted earlier, that $1.16 figure is facially invalid based on the settlement agreement, which caps ordinary expense claims at $1,000. As the consolidated plaintiffs conceded at oral argument, the class‘s 176 ordinary expense claims
On remand, the district should scrutinize the reasonableness of the consolidated plaintiffs’ lodestar. It should then calculate the actual value of the settlement to the class and perform a crosscheck of the lodestar against the 25% percentage-of-recovery benchmark to ensure that fees are reasonable. Based on the record before us, that method will likely favor downward adjustment.
CONCLUSION
We AFFIRM the district court‘s settlement approval but REVERSE the attorneys’ fee award. We REMAND for the district court to determine the settlement‘s actual value to class members and award reasonable and proportionate attorneys’ fees, consistent with this opinion.
COLLINS, Circuit Judge, concurring in the judgment in part and dissenting in part:
I concur in the judgment to the extent that the majority reverses and remands the district court‘s approval of the attorney‘s fee award. I respectfully dissent, however, from the majority‘s decision to affirm the approval of the underlying settlement in this case.
Under our precedents and
This duty is even more stringent, we have held, “when a settlement is negotiated absent class certification.” Allen, 787 F.3d at 1224. “Specifically, such settlement agreements must withstand an even higher level of scrutiny for evidence of collusion or other conflicts of interest than is ordinarily required under Rule 23(e) before securing the court‘s approval as fair.” Roes, 1–2 v. SFBSC Mgmt., LLC, 944 F.3d 1035, 1048–49 (9th Cir. 2019) (simplified). This higher scrutiny is required in order “to ensure that class representatives and their counsel do not secure a disproportionate benefit at the expense of the unnamed plaintiffs.” Id. at 1049 (simplified).
Here, in approving the final settlement proposal before class certification, the district court failed to follow these procedural requirements. It provided little explanation as to why it approved this settlement and instead issued a series of perfunctory orders, despite the fact that, inter alia, (1) the final settlement triggers every Bluetooth factor; (2) the settlement‘s final value ended up being nearly a fourth of the estimated “conservative” value presented at the preliminary approval hearing; and (3) the settlement‘s proposed attorney‘s fees award comprises nearly 46% of the entire settlement. The district court‘s failure to adequately examine any of these issues compels reversal and remand. Because the majority nonetheless upholds the district court‘s approval of the settlement, I respectfully dissent.
I
A
In September 2021, California Pizza Kitchen (“CPK“), a national restaurant chain, suffered a data breach that exposed the names, Social Security numbers, dates of birth, and other personal identifying information of over 100,000 current and former employees. On November 15, 2021, CPK, after investigating the breach, sent notification letters to 103,767 individuals whose personal information may have been compromised as a result of the breach.
Soon thereafter, five putative class actions were filed in the United States District Court for the Central District of California, seeking monetary and equitable relief for a nationwide class of persons whose personal identifying information was accessed by the breach.1 The last of these suits was filed on December 10, 2021 by plaintiffs Aviva Kirsten and Jeremy Pittman (the “Kirsten Plaintiffs“), who are the Appellants in this case. In contrast to the complaints of the first four plaintiff groups (the “Settling Plaintiffs“), the Kirsten Plaintiffs’ complaint explicitly asserted a claim under the
Two days before the complaint in Kirsten was filed, the Settling Plaintiffs agreed to coordinate their efforts. After initiating their suit, the Kirsten Plaintiffs reached out to the Settling Plaintiffs “to meet and confer about streamlining efforts and potentially coming to consensus on leadership.” However, these attempts to reach consensus failed.
Attempts to consolidate all five cases also failed. On January 24, 2022, the Kirsten Plaintiffs drafted and circulated a stipulation to consolidate the five cases and establish a briefing schedule for the appointment of lead class counsel. Fifteen days later, the Settling Plaintiffs circulated their own version of the stipulation, and the Kirsten Plaintiffs sent back redlines of the draft stipulation, which, inter alia, proposed a shorter briefing schedule for contested leadership motions. The next day, the Settling Plaintiffs responded with an edited draft stipulation rejecting the Kirsten Plaintiffs’ proposed leadership motion schedule. After five days, the Settling Plaintiffs, without hearing back from the Kirsten Plaintiffs, filed a stipulation and a proposed order consolidating the four other cases. In a footnote, the stipulation noted the existence of the Kirsten litigation, stating that “the parties to this stipulation were unable to obtain agreement to this stipulation from plaintiffs in that action.” Without comment, the district court then entered the proposed order consolidating the four cases. The deadline for lead counsel motions was set for March 17, 2022.
B
Meanwhile, the Settling Plaintiffs had begun discussing settlement with CPK. After a mandatory pre-motion conference with CPK‘s counsel concerning CPK‘s anticipated motion to dismiss, counsel in the Wallace case had “substantial questions in [his] mind about the viability of Plaintiffs’ claims in this litigation.” He subsequently drafted a settlement proposal term sheet, which he first circulated to his co-counsel and then, on January 26, 2022, to CPK‘s counsel. The term sheet proposed a claims-made settlement and included a demand for compensation for statutory claims under the CCPA.
On February 9, 2022, the Kirsten Plaintiffs and CPK conferred about scheduling in the Kirsten litigation. After this meeting, CPK‘s counsel sent an email to the Kirsten Plaintiffs’ counsel asking, “Would you consider for example a claims made style settlement? Do you have a proposed term sheet?” The Kirsten Plaintiffs responded that the parties would need to exchange information and set a mediation before engaging in any settlement talks.
Following consolidation, the Settling Plaintiffs and CPK began settlement talks in earnest. The parties engaged in informal discovery regarding CPK‘s insurance coverage and the data breach. The parties selected a mediator and engaged in their first mediation session, via Zoom, on March 10, 2022. This mediation session was not fully successful, but the parties agreed to meet again on March 15 for another mediation session on Zoom.
Four days later, on March 14, 2022, CPK‘s counsel and the Kirsten Plaintiffs’ counsel conferred pursuant to
The next day, on March 15, 2022, the Settling Plaintiffs and CPK held their second mediation session, where they ultimately came to agreement on the material terms of settlement, including an award of attorney‘s fees. On March 16, 2022, they informed the district judge‘s courtroom deputy clerk that they had reached a settlement in principle, see C.D. CAL. LOC. CIV. R. 16-15.7, and they also filed a stipulation requesting that the district court stay the consolidated action pending approval of the proposed settlement. Motions for appointment of lead counsel would have been due the next day.
On March 17, 2022, the Kirsten Plaintiffs filed a motion to have their counsel appointed lead counsel, together with an opposition to the settling parties’ request for a stay. Nonetheless, the district court promptly granted the requested stay of proceedings, pending the filing and disposition of a motion for preliminary approval of the settlement.
On April 22, 2022, the district court held a hearing concerning the Kirsten Plaintiffs’ lead counsel motion. The court stated that “[i]t seems just odd on my part that I wouldn‘t consolidate all five cases” and wondered out loud why it had not done so, but the court ultimately decided “to leave the case[s] divided” and to hold the lead counsel motion in abeyance while the district court evaluated the Settling Plaintiffs’ motion for preliminary approval. The Settling Plaintiffs filed a motion for preliminary approval on May 2, 2022, and the court subsequently scheduled a hearing on the motion for June 29, 2022.
C
Before summarizing the preliminary approval hearing, I will first describe (1) the key terms of the proposed class settlement between the Settling Plaintiffs and CPK; and (2) the Kirsten Plaintiffs’ written objections to the settlement prior to the preliminary approval hearing.
1
Under the terms of the proposed settlement, the Settling Plaintiffs agreed that “every Settlement Class member (except those who timely opt out)” would “fully and finally release CPK . . . from any and all claims or causes of action, whether known or unknown, that concern, refer or relate to the Data Security Incident announced by CPK on or about November 15, 2021, and all other claims arising out of the Data Security Incident announced by CPK on or about November 21, 2021, that were asserted, or that could have been asserted, in the Consolidated Cases.” In return, CPK agreed to a claims-made settlement, i.e., one in which only those class members who submitted timely, valid claims could receive any payment. That claims-made settlement would provide the following relief:
- Up to $1,000 per claimant for “out-of-pocket expenses and lost time” incurred as a result of the data breach.
- Up to $5,000 per claimant for “extraordinary losses,” i.e., “proven monetary loss as a result of actual identity theft.”
- A $100 “statutory damages award” for “California Settlement Subclass members,” which could be combined with a claim for out-of-pocket expenses, but “subject to the $1,000 cap on compensation for ordinary losses and lost time.”
-
“24 months of 3-bureau credit monitoring to Settlement Class members who enroll,” which the Settling Plaintiffs valued at $360 per claimant.
In the settlement, CPK also agreed to a clear-sailing provision whereby it “agree[d] not to object to [the Settling] Plaintiffs’ request for combined attorneys’ fees and costs to” their counsel “in an amount not to exceed a total of $800,000, inclusive of costs.” CPK further agreed to “maintain certain recently implemented business practices and remedial measures . . . for a period of three (3) years” to safeguard against future data breaches.
Under the agreement, class members would receive notice of the settlement primarily through notices sent to the respective postal addresses or email addresses “associated with” each class member. There would also be a website and toll-free number for any inquiries regarding the settlement. The agreement provided for a specified class action settlement administrator to administer the claims, including providing notice to class members, with CPK being responsible for all costs of the settlement administration.
2
In their written opposition to the Settling Plaintiffs’ motion for preliminary approval, the Kirsten Plaintiffs raised six objections to the proposed settlement.
First, the Kirsten Plaintiffs asserted that because the settlement is claims-based and allows any unclaimed funds to remain with CPK, the settlement is functionally reversionary, thereby presenting one of the “subtle signs” this court has identified of potential collusion between class counsel and defendant‘s counsel. Allen, 787 F.3d at 1224. The Kirsten Plaintiffs argued that “[r]eversionary clauses are highly disfavored in the Ninth Circuit,” because they promote “perverse incentives” for “defendants to ensure as low a claims rate as possible” and because they “can benefit both defendants and class counsel, and thus raise the specter of their collusion, by (1) reducing the actual amount defendants are on the hook for . . . ; and (2) giving counsel an inflated common-fund value against which to base a fee motion.”2 Roes, 1–2, 944 F.3d at 1058–59 (simplified).
Second, the Kirsten Plaintiffs objected to the settlement‘s inclusion of a “clear sailing agreement,” that is, a provision that CPK would “not object to an amount of $800,000 paid in attorney fees.” In doing so, the Kirsten Plaintiffs relied on caselaw holding that “clear sailing agreements on attorney‘s fees are important warning signs of collusion because the very existence of a clear sailing provision increases the likelihood that class counsel will have bargained away something of value to the class.” Roes, 1–2, 944 F.3d at 1051 (simplified).
Third, the Kirsten Plaintiffs contended that CPK‘s potential liability was $77 million, but that the proposed settlement‘s ceiling for liability was approximately $38 million with “[t]he only guaranteed payment under the [s]ettlement [being] the $800,000 in attorney fees” to the Settling Plaintiffs’ counsel. As a result, the Kirsten Plaintiffs argued, “[t]he amount offered in [the] [s]ettlement is unfair, unreasonable and inadequate.” The Kirsten Plaintiffs
Fourth, the Kirsten Plaintiffs criticized the settlement‘s one-time mail/e-mail notice plan as inadequate. In their view, the notice plan should also have included “(1) a reminder notice program, (2) a targeted social media/internet advertising campaign, and (3) a publication of the notice through CPK‘s employee intranet or other internal employee communication system.” According to the Kirsten Plaintiffs, option (3) would have been free, and the other two would have “cost very little.”
Fifth, the Kirsten Plaintiffs argued that the settlement‘s release was overbroad because it extended to any claims that “relate to” the data breach in question.
Sixth, the Kirsten Plaintiffs also contended that additional circumstances surrounding the settlement talks underscored the potentially collusive nature of the settlement. The Kirsten Plaintiffs highlighted the facts that (1) they were not included in the stipulation for consolidation; and (2) CPK had avoided informing them of the ongoing settlement negotiations with the Settling Plaintiffs. The Kirsten Plaintiffs also observed that the settlement was reached the day before motions for the appointment of lead counsel would have been due and that they had previously rejected any claims-made settlement structure in their discussions with CPK. The Kirsten Plaintiffs further argued that the mere fact that a mediator had presided over the settlement negotiations was no guarantee of its compliance with
D
At the preliminary approval hearing held on June 29, 2022, the mediator took the stand and answered questions from the court. The district court stated that, based on the Settling Plaintiffs’ representations, “the settlement ha[d] a conservative value [of] over $3.7 million,” and the $800,000 in attorney‘s fees amounted to “approximately 21 percent of the value of the settlement.” Nonetheless, the district court discerned “some due process issues with” the fact that the settlement would resolve the claims in the Kirsten litigation, despite the Kirsten Plaintiffs’ lack of involvement in the mediation. Moreover, the district court stated that it had “a tremendous concern about these fees,” due to the “reversionary” nature of the settlement.
After returning from a recess, the district court announced that it would preliminarily find that the settlement was “adequate, fair, and reasonable.” The court then expressly addressed only one of the Kirsten Plaintiffs’ objections, namely, the claims-made structure of the settlement. The district court declined to approve the $800,000 in attorney‘s fees at that time, explaining that it would wait to see how many claims were made in order to “incentivize[]” the Settling Plaintiffs’ counsel “to get as much money for this class” during the claims process. The court indicated that it would “probably” approve the requested attorney‘s fees if the total settlement value amounted to approximately $3.2 million “or even” $2.5 million. The district court offered no comment on any of the Kirsten Plaintiffs’ other written objections, nor did the court set forth on the
On June 30, 2022, the district court issued a written order preliminarily approving the settlement and finding the settlement “to be fair, reasonable, and adequate, and the result of vigilant, informed, non-collusive arms‘-length negotiations overseen by an experienced and neutral mediator.” The court stated, without further explanation, that it “considered . . . the briefs and arguments of counsel.” As the majority puts it, “the district court‘s preliminary . . . order[] w[as] sparse—almost boilerplate and memorialized little of the district court‘s rationale.” See Opin. at 16.
E
Following the preliminary approval of the settlement, the Settling Plaintiffs and the class administrator sent out 103,380 direct postcard notices and 4,349 email notices to potential class members. Thereafter, a follow-up email notice was sent to more than 60,500 class members. There was also a targeted social media campaign on Facebook, Instagram, and LinkedIn. Despite these efforts, which went beyond the minimum notice provisions to which the Kirsten Plaintiffs had objected, the class administrator ultimately received only 1,828 unvalidated claims for a claims rate of less than 1.8%.
On October 6, 2022, the Settling Plaintiffs filed their motion for final approval of the settlement. The Kirsten Plaintiffs objected, essentially re-raising the same concerns they had voiced prior to the preliminary approval hearing. They also pointed out that, after the preliminary approval hearing, their CCPA claims had survived both a motion to dismiss and a motion for reconsideration.3 The Kirsten Plaintiffs further argued that the low claims rate highlighted that CPK would be “securing a full release” from the class “for a very low dollar cost” and that the settlement‘s “notice scheme was woefully inadequate.”
At the final approval hearing on November 7, 2022, the Settling Plaintiffs confirmed that the claims rate was approximately 1.8%, and they asserted that the final settlement value, prior to claim validation, was $1,161,149.4 According to the Settling Plaintiffs, the claims made were as follows:
- 176 claims for ordinary expenses, totaling $384,134.77;
- 979 claims for lost time, totaling $50,320.00;
- 45 claims for extraordinary losses, totaling $191,354.50;
- 803 claims for CCPA statutory damages, totaling $80,300.00; and
- 1,264 claims for credit monitoring, totaling $455,040.00.
The court then asked the Kirsten Plaintiffs to repeat their objections on the record. The Kirsten Plaintiffs recounted that the settlement includes a clear-sailing provision; that the settlement was reversionary and had ultimately gotten “nowhere near the suggested 3 million or in excess of that“; and that, consequently, the Settling Plaintiffs’ counsel were receiving “a disproportionate amount of the settlement“—approximately 40%—in attorney‘s fees. The Kirsten Plaintiffs also pointed out that the Settling Plaintiffs had not undergone any motions practice at all, which, in the
The district court responded that it had already “gone through the collusion arguments” and “rejected those.” With respect to the settlement‘s final value at $1.16 million, the court merely stated that it could not “recall why” it had earlier expected the settlement to amount to $3 million. The district court also noted its “tremendous concern over these attorney‘s fees” of up to $800,000. At the conclusion of the hearing, the district court requested additional briefing on the issue of attorney‘s fees. The district court did not discuss the remainder of the Kirsten Plaintiffs’ objections, nor did it set forth on the record any of its findings.
The district court received the requested briefing concerning the attorney‘s fees and held another hearing on December 5, 2022.5 Thereafter, the district court on February 22, 2023 approved the settlement and awarded the maximum $800,000 in attorney‘s fees to the Settling Plaintiff‘s counsel. The court‘s order contains no explanation for its ruling and merely states that the court “reviewed . . . any objections filed with or presented to the Court” and the “oral argument of counsel and any objectors who appeared.” Like the preliminary approval order, as the majority puts it, “the district court‘s . . . final approval order[] w[as] sparse—almost boilerplate—and memorialized little of the district court‘s rationale.” See Opin. at 16.
II
Given these details in the record, it is clear that the district court contravened its duty to explain why the settlement complied with
A
“To survive appellate review, the district court must show it has explored comprehensively all factors, and must give a reasoned response to all non-frivolous objections.” Allen, 787 F.3d at 1223–24 (emphasis added) (simplified). And because, as the majority correctly recognizes, the settlement has a claims-made structure that resulted in a claims rate of less than two percent, provides attorney‘s fees equivalent to 84% of the settlement‘s benefit to the class, and contains a clear-sailing provision for these fees, the settlement here “raises all three red flags of potential collusion under” our caselaw. See Opin. at 17. The district court therefore had a “heightened obligation” to scrutinize the settlement for “any evidence of collusion or other conflicts of interest.” See Opin. at 19 (simplified).
The district court did not abide by this mandate. To begin with, the district court committed an obvious and critical error by assuming that the settlement was valued at $1.16 million when in fact it was worth no more than $950,000. See Opin. at 11 (referring to this as a “glaring error . . . apparent on [the settlement‘s] face“). Because of the $1,000 cap on per-claimant out-of-pocket expenses, the 176 claimants who sought $384,134.77 are entitled, at most, to $176,000. Yet the district court relied on a patently incorrect valuation of the settlement in determining whether the attorney‘s fees were disproportionate to the settlement‘s benefit to the class. This error is starker still because, as recounted above, the district court‘s chief concern at the settlement approval hearings was the
Moreover, the record is bereft of any discussion of
Furthermore, what little the district court did state on the record was internally contradictory, making the court‘s reasoning all the more confounding and opaque. At the preliminary approval hearing, the district court expressed considerable concern about the potential that the settlement‘s value would end up being much less than $3 million, particularly given that the claims-made nature of the settlement arguably rendered it “reversionary” in nature. In one of its few substantive comments, the court explained that in “incentiviz[ing]” the Settling Plaintiffs’ counsel “to get as much money for this class” during the claims process, the court sought to address its discomfort about “not knowing the amount of the claims” that would ultimately be made. Yet, at the final hearing, the district court heard that the settlement‘s final value was no more than $1.16 million, with a claims rate of less than two percent, and that therefore the court‘s hope that the Settling Plaintiffs’ counsel would achieve the previously hypothesized claims rate had failed. Nonetheless, the district court approved the settlement. The district court never explained in its written orders or on the record why it did so, in spite of its prior stated concerns about the claims-made nature of the settlement and the possibility of a low claims rate.
This failure to properly address the low claims rate was independently an abuse of discretion. In Allen, we vacated and remanded a settlement because, after “class counsel said that he would consider it a success if even 10% or 15% of the class made claims,” the actual “claims rate was less than 8%,” and “the record g[ave] no
Finally, the narrow gap between the maximum value of the settlement and the attorney‘s fees award “ma[de] it all the more important for the district court to closely examine the claimed value of the non-cash portions of the settlement that were used to justify the requested attorneys’ fees.” Roes, 1–2, 944 F.3d at 1051. While the district court asked questions about the two-year credit monitoring program, it never made a finding attaching a monetary value to the program. Again, we have previously reversed district courts for precisely this reason. See, e.g., Kim, 8 F.4th at 1179 (reversing a settlement approval because we discerned “no basis for [the district court‘s]” decision to “accept[] class counsel‘s unsupported representation” of the settlement‘s injunctive relief‘s worth); Allen, 787 F.3d at 1225 (reversing a district court after it “did not make express findings about the value of the injunctive relief“).
The foregoing errors plainly demonstrate that the district court failed to apply heightened scrutiny to the settlement and therefore abused its discretion in approving the settlement. The settlement should be vacated, and this matter remanded to the district court so that it can engage in “a more searching inquiry,” Roes 1–2, 944 F.3d at 1050 (citation omitted), and provide a “clear explanation of why the disproportionate fee is justified” and why this settlement was not the result of collusion, In re Bluetooth, 654 F.3d at 949.
B
The majority agrees with much of what I have said. Specifically, the majority agrees that the district court had a “heightened obligation to ferret out any evidence of collusion or other conflicts of interest” in the negotiation of the settlement. See Opin. at 19 (simplified). The majority also agrees that the settlement here “raises all three red flags of potential collusion under Bluetooth,” see Opin. at 17 (citation omitted), and that the district court missed at least “one glaring error . . . apparent on [the] face” of the settlement, see Opin. at 11. Further, the majority agrees that, despite these red flags and the district court‘s heightened obligation, “the district court‘s preliminary and final approval orders were sparse—almost boilerplate—and memorialized little of the district court‘s rationale,” see Opin. at 16, and that it is impossible to “infer the district court‘s rationale for its fee award from the record,” even though the court “approv[ed] fees that appear excessive of [the] settlement[‘s] value,” see Opin. at 24. The majority nonetheless upholds the approval of the settlement, but its reasons for doing so are flawed.
The majority contends that it “can reasonably infer the district court‘s rationale from the record,” see Opin. at 16, and that “the district court gave the settlement provisions a hard look under the Bluetooth factors,” see Opin. at 20. But this conclusory assertion lacks support in the record. The district court expressly addressed only one of the Bluetooth factors—the disparity between the size of the settlement
The majority cites two pre-2000 cases for the proposition that, in some instances, a district court‘s issuance of a conclusory order will not be reversible, but these cases lend no help to the majority here. See Opin. at 16 (first citing Hanlon v. Chrysler Corp., 150 F.3d 1011, 1023 (9th Cir. 1998), overruled on other grounds by Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011), and then citing In re Pac. Enters. Sec. Litig., 47 F.3d 373, 377 (9th Cir. 1995)). As an initial matter, the majority misreads Hanlon. The “almost conclusory” findings that were upheld in Hanlon exclusively concerned “the requirements of Rule 23(a) and 23(b)(3),” and not “the ‘fair, adequate and reasonable’ preconditions for class settlement pursuant to
At bottom, the majority‘s error stems from its unfounded belief that merely “prob[ing]” and “question[ing],” see Opin. at 20, the parties and a mediator at a hearing is a substitute for judicial reasoning and the careful exercise of judgment. But we have said otherwise. We have held that
*
*
*
I respectfully dissent.