The Supreme Court recently granted review in a case that involves whether, or in what circumstances, cy pres relief may be used in class action settlements. In Frank v. Gaos, No. 17-961, the Court will review the Ninth Circuit’s decision in In re Google Referrer Header Privacy Litig., 869 F.3d 737 (9th Cir. 2017). The question presented in the petition for certiorari is: “Whether, or in what circumstances, a cy pres award of class action proceeds that provides no direct relief to class members supports class certification and comports with the requirement that a settlement binding class members must be ‘fair, reasonable, and adequate.’”

The case involves claims under the Stored Communications Act and various state common law claims, alleging that Google violated users’ privacy rights by disclosing the search terms used to owners of websites. The district court approved a settlement that called for Google to provide a disclosure of how it shares users’ search terms, and for an $8.5 million settlement fund. $5.3 million of the settlement fund would be distributed to six cy pres recipients, non-profit organizations that would use the funds to promote Internet privacy, and the remaining $3.2 million would go to attorneys’ fees, administrative costs and incentive payments for the named plaintiffs. The Ninth Circuit affirmed. It explained that cy pres-only settlements are “the exception, not the rule,” but that they are appropriate where the settlement is “non-distributable” because it would not be feasible to distribute money to individual class members. Here, the class size is estimated at 129 million class members, who would be entitled to 4 cents each, and the cost of sending out the payments would exceed the benefit. The objectors argued that there should be a lottery system whereby some class members would receive say $5 or $10, and that if the settlement was nondistributable, the court should instead rule that the superiority requirement for class certification was not satisfied. The Ninth Circuit found that the district court’s rejection of these arguments and approval of the settlement was not an abuse of discretion.

The objectors also argued that the cy pres recipients were inappropriate because Google had previously donated to some of them, three of them previously received funds in other Google class settlements, and three of them were affiliated with class counsel’s alma maters. The Ninth Circuit rejected these contentions, noting that the organizations had a strong nexus to the interests of the class, Google had previously donated to hundreds of organizations, there was no fraud or collusion, and the mere fact that class counsel graduated from schools that had connections to some of the organizations did not warrant rejecting the settlement. The Ninth Circuit also found the attorneys’ fees reasonable because they were 25% of the settlement fund and consistent with a lodestar calculation.

Judge Wallace wrote an opinion concurring in part and dissenting in part. He dissented only on the issue involving the relationships between the cy pres recipients and the alma maters of class counsel, concluding that the district court should have looked into that issue further. He wrote that “I would vacate the district court’s approval of the class settlement, and remand with instructions to hold an evidentiary hearing, examine class counsel under oath, and determine whether class counsel’s prior affiliation with the cy pres recipients played any role in their selection as beneficiaries.” Id. at 748 (Wallace, J., concurring in part and dissenting in part). In my view, requiring such a hearing could present substantial concerns about the protection of the attorney-client privilege and work product doctrine – ultimately it is the clients on both sides who are making settlement decisions, and examining counsel as to their thinking (or their advice to their client) behind such decisions seems problematic, although that may not be something the Supreme Court addresses in this case.

From a defense perspective, this case may present a challenge with respect to what position industry organizations may wish to take in amicus briefs. If the Court were to rule that cases where settlements would be “non-distributable” do not satisfy the superiority requirement and should never be certified class actions at all, defendants might welcome that result. But how would such a rule be applied, and where would the line be drawn? In most cases like this, one might be able to calculate a theoretical class recovery (e.g., with punitive damages) that might, in theory, be large enough to be distributable, although such an amount is very unlikely to be awarded. If a “non-distributable” settlement means only that the settlement is rejected and the parties are forced to continue to litigate, that forces defendants to incur large litigation expenses and risk in cases that they would prefer to resolve. The reality is that there needs to be a route to settling cases like this when appropriate. To the extent the Court is granting certiorari to provide some guidance around when to use cy pres relief and how to do it, that may be welcomed by both sides, provided the Court does not make it unduly difficult to satisfy whatever criteria are ultimately adopted.

The Ninth Circuit recently ruled in favor of President Trump. That was not a typo, and this is not fake news. The ruling was not in favor of Trump in his official capacity, but in his capacity as a class action defendant. And it was in a case where the plaintiffs were on his side, seeking to defend the lower court’s approval of a class action settlement involving Trump University. The key practice pointer from this decision for those who litigate class actions is that if a class is certified before you reach a settlement, you don’t have to give the class members a second chance to opt out when you later reach a settlement. You might want to make it explicit that there will not be a second chance to opt out.

In Low v. Trump Univ., LLC, No. 17-55635,2018 U.S. App. LEXIS 2920 (9th Cir. Feb. 6, 2018), an objector challenged the district court’s approval of a settlement of several class actions involving allegations of false advertising and fraudulent practices by Trump University, which provided real estate investment seminars. The objector argued that she was entitled to a second opportunity to opt out of the class after a settlement was reached.

The Ninth Circuit first held that the notice that was sent to class members when the class was certified and before a settlement was reached did not promise them as second opportunity to opt out. The Ninth Circuit adopted a standard focusing on “what an average class member would have understood the notice to guarantee.” Id. at *19. The objector focused on a statement in the notice that if “the Plaintiffs obtain money or benefits, either as a result of the trial or a settlement, you will be notified about how to obtain a share (or how to ask to be excluded from any settlement).” Id. at *16.  The court concluded that the notice, which was based on a Federal Judicial Center form, when read as a whole, made clear that there was a single deadline to opt out, and the decision had to be made when the first notice was sent. Based on the notice as a whole, a class member being excluded from the settlement could refer to simply declining to submit a claim form in a claims made settlement, and it was unclear at that point what type of settlement might be reached.

The Ninth Circuit then held that due process did not require a second opportunity to opt out, applying binding precedent in Officers for Justice v. Civil Service Commission of San Francisco, 688 F.2d 615 (9th Cir. 1982), which reasoned that class members’ rights were adequately protected by the opportunity to object to the settlement, the district court’s fairness hearing, and the right to appeal from the approval of a settlement. The court found no intervening Supreme Court authority on point. In a footnote, the court suggested that a district court might have the power to withhold approval of a settlement that did not provide a second opportunity to opt out, in an appropriate case.

Although large numbers of opt outs are relatively rare in the context of a class notice, regardless of whether a settlement has been reached, it may be advisable not to offer class members a second chance to opt out. If the parties think ahead about this issue, they can address it explicitly when the first notice is issued to class members.

Discovery disputes in class actions often focus on plaintiffs’ requests for computer data regarding putative class members’ claims, and how far defendants need to go in providing such data. An Illinois federal court recently addressed this in an employment class action. The key takeaways from this opinion are that: (1) a defendant was not required to create a dataset that did not exist in order to respond to the plaintiffs’ discovery requests; and (2) a defendant was entitled to create a new dataset as work product for the purpose of assisting its expert, without disclosing the data until the expert report was disclosed.

In Ahad v. Bd. of Trs. of S. Ill. Univ., 2018 U.S. Dist. LEXIS 11248 (S.D. Ill. Jan. 24, 2018), a putative class action alleging gender discrimination in the compensation of physicians, the plaintiff sought sanctions on the grounds that the defendants’ expert had relied on certain data regarding physicians’ compensation that was not produced in response to the plaintiff’s earlier discovery requests. The court first held that “Defendants did not have a duty to create a spreadsheet or database that tracked total compensation when one did not exist” because, although the defendants had the ability to create such data, they “were only required to produce records as kept in the normal course of business,” and “were not under a further obligation to create a new dataset if it was not kept in the usual course of business.” Id. at *14-16. The court noted that the plaintiff could have requested raw payroll data from which a dataset could be created. Id. at *16. The defendants were entitled to create a total compensation spreadsheet for the use of their expert witness, which was protected as work product until the expert report was disclosed. Id.

The court’s approach here allows defendants the flexibility of being able to work with a consulting expert and with their counsel to create a new dataset that may or may not support their position in the litigation, and protect that information unless and until they want to use it. Plaintiffs are not prejudiced because they can request raw data and have their expert manipulate it as they see fit to attempt to support their case.

The Ninth Circuit’s decision this week vacating a class action settlement in In re Hyundai and Kia Fuel Economy Litig., No. 15-65014 (9th Cir. Jan. 23, 2018) is getting a lot of attention in the class action bar. It’s 84 pages long, but the bottom line is that the Ninth Circuit held that the district court failed to analyze whether differences in state law precluded a finding that common questions of law and fact predominated under Rule 23(b)(3). That’s an argument that defendants often make in opposing class certification, and courts often agree with it. In fact, the trial court in this case when a motion for class certification was under consideration strongly suggested that this argument would carry the day. But the parties later settled and agreed to certification, and the Ninth Circuit majority criticized the district court for sidestepping the issue in the settlement approval process. This is because the basic requirements for class certification under Rule 23(a) and (b) must be satisfied for settlement purposes under the Supreme Court’s decision in Amchem Prods., Inc. v. Windsor, except that factors such as manageability do not apply where the case is being settled, not tried. This is nothing new, but parties and district courts sometimes do not give sufficient attention to this in the settlement process.

Some commentators have suggested that this decision could endanger nationwide class action settlements in cases governed by state law (it’s not a problem where federal law applies). Defendants sometimes need to enter into a nationwide settlement to achieve complete peace, such as in this MDL proceeding involving alleged misrepresentations as to the fuel economy of Hyundai and Kia vehicles. Entering into 50 separate statewide settlements would be impractical.

So what can parties do when they want to settle but the issues are governed by state law that in some respects varies among the 50 states? Here are a couple of thoughts:

  • If you need to do a Rule 23(b)(3) settlement, find at least one plausibly common and predominating issue on which state law does not vary and make that your focus in seeking approval of the settlement. The defendant does not have to affirmatively sign onto the argument that predominance is satisfied based on that issue; it can just not oppose the plaintiff’s motion. Defenses should not be an issue because it is a settlement, and thus the defendant is effectively waiving its individualized defenses if the settlement is approved.
  • Consider whether you can seek certification of a settlement class under Rule 23(b)(2) if the settlement includes some declaratory or injunctive relief and the monetary relief goes hand-in-hand with the declaratory or injunctive relief. Rule 23(b)(2) does not require a showing of predominance. If the class receives notice and has the opportunity to opt out, some of the concerns about the propriety of a Rule 23(b)(2) class may fall by the wayside.

In Bristol-Myers Squibb Co. v. Superior Court, 137 S. Ct. 1773 (2017), the U.S. Supreme Court held that the California state courts lacked personal jurisdiction over claims made by out-of-state plaintiffs in a mass action, where the plaintiffs did not claim that they were harmed in California and the defendant’s relevant conduct did not occur in California. (See my June 22, 2017 blog post for a more extensive summary of the decision.) There has been some debate about whether this decision applies to a class action, and whether it applies in federal court.

A Northern District of Illinois decision recently addressed this issue in the context of a nationwide putative class action against a distributor of a dietary supplement. DeBernardis v. NBTY, Inc., 2018 U.S. Dist. LEXIS 7947 (N.D. Ill. Jan. 18, 2018). While noting that there has been a split of authority among district courts nationwide, and that it was a “close question,” the court dismissed the plaintiff’s claims to the extent they were asserted on behalf of putative out-of-state class members, explaining that:

The Court believes that it is more likely than not based on the Supreme Court’s comments about federalism that the courts will apply Bristol-Myers Squibb to outlaw nationwide class actions in a form, such as in this case, where there is no general jurisdiction over the Defendants. There is also the issue of forum shopping, which was mentioned in the Chinese DryWall case as a basis for distinguishing mass torts from class actions, but possible forum shopping is just as present in multi-state class actions.

Defendants should closely monitor developments on this important issue. It is likely worth raising not only in the context of a nationwide class action but also in a multi-state class action. Some plaintiffs’ lawyers are bringing simultaneous or serial statewide class actions individually, which avoids the impact of this decision. Another approach is for plaintiffs’ lawyers to bring a nationwide or multistate class action in a jurisdiction where the defendant is subject to general jurisdiction, such as the state of incorporation or principal place of business.

In a long-running employment class action in California, a California Court of Appeal recently addressed once again the use of surveys of class members. The case was the subject of a prior California Supreme Court decision (see my June 2014 blog post), which provided guidance on the use of statistical evidence by plaintiffs to attempt to prove their claims on the merits in class actions, the need for a trial plan, and the need to allow the defendant to prove its affirmative defenses.

After remand, the trial court ultimately denied class certification on various grounds, including the unreliability of, and discrepancies between, two surveys conducted by plaintiffs’ counsel. In Duran v. U.S. Bank Nat’l Ass’n, 2018 Cal. App. LEXIS 36 (Cal. Ct. App. Jan. 17, 2018), the California Court of Appeal recently affirmed. What I found most interesting was how the court evaluated whether a trial of the affirmative defenses would be manageable:

In the portion of the court’s decision about which plaintiffs complain, the court observed that if “the vast majority of the class sample would credibly testify that most of their time was spent inside the office and the confidence interval was relatively narrow, then one might infer that the number of witnesses the defense would call on its affirmative defenses would be relatively small. But if the results relied upon [by an expert witness] are as unreliable as suggested by a comparison of the 2008 and 2015 surveys, then the court cannot take much comfort in the notion that Plaintiffs’ trial plan adequately manages the affirmative defenses. As with the issue of restitution, the affirmative defenses would appear to require a host of ‘mini trials.’”

What is the best approach for the defendant to take in this scenario? Strategically, a defendant may not want to reveal too much of its trial strategy at the class certification stage. But it may be necessary in order to present the best possible defense to class certification. A court should not need to guess how many witnesses the defense would call on its affirmative defenses based upon the reliability of a survey. The defendant can instead demonstrate how it will prove its affirmative defenses and why that will make a class trial unmanageable.

When a defendant receives an adverse decision on a motion for class certification, whether the court of appeals grants permission to appeal under Rule 23(f) of the Federal Rules of Civil Procedure (or a state-court equivalent, if available) can be a crucial turning point in the case. If the court of appeals will not hear an interlocutory appeal on class certification, the only way to obtain review of that decision is to take the case through trial, to a final judgment. But few class actions are tried. Due to the high stakes and large costs involved, when a class is certified, the case often settles. While the petition for permission to appeal under Rule 23(f) is thus a crucial step, it must be put together under a short timeframe (14 days from the trial court’s order), which may make it difficult for the defendant to select and retain appellate counsel new to the case. And the petition must be concise – 5,200 words (approximately 20 pages).

The short timeframe and limited space for these petitions sometimes results in defendants filing petitions that are focused heavily or even entirely on why they believe the trial court got it wrong. But the court of appeals will also be focused on other factors, such as whether there are significant questions of law involved (as opposed to merely application of established law to the facts of your case), why the issues presented matter for other cases or the development of the law, and why the appeal cannot wait until after a final judgment.

In many instances, a court of appeals’ decision on whether to grant permission to appeal will be only a one-line order without explanation. In In re Marietta Memorial Hospital, 2018 U.S. App. LEXIS 460 (6th Cir. Jan. 8, 2018), however, the Sixth Circuit wrote a short opinion on this, likely to provide parties filing Rule 23(f) petitions with additional guidance. The court began by setting forth the key factors it considers in deciding whether to hear these appeals:

We may, in our discretion, permit an appeal from an order granting class certification. Fed. R. Civ. P. 23(f). This “unfettered” discretion is akin to the discretion of the Supreme Court in considering whether to grant certiorari; thus, we may consider any relevant factor we find persuasive. See Fed. R. Civ. P. 23, advisory committee’s note (1998 amendment); In re Delta Air Lines, 310 F.3d 953, 959 (6th Cir. 2002) (per curiam). Typically, however, we consider (1) whether the petitioner is likely to succeed on appeal under a deferential abuse-of-discretion standard; (2) whether the cost of continuing the litigation for either the plaintiff or the defendant presents such a barrier that subsequent review is hampered; (3) whether the case presents a novel or unsettled question of law; and (4) the procedural posture of the case before the district court. In re Delta Air Lines, 310 F.3d at 960.

In explaining why it declined to grant review, the Sixth Circuit noted that the defendant “has not identified any novel legal question raised by the petition,” and that the defendant’s “mere ‘general assertion[s]’ that certification is dispositive of the litigation” were insufficient. The Sixth Circuit advised that, in order to demonstrate that the costs of litigation would likely preclude review after a final judgment, “the defendant should provide insight into the potential expenses and liabilities.” The Sixth Circuit also noted that the defendant failed to address whether the district court might reexamine the class certification decision, and noted that both parties had filed dispositive motions, which weighed against interlocutory review.

The short opinion in Marietta Memorial Hospital provides a useful reminder for defendants about what should be addressed in a Rule 23(f) petition. It is also important to keep in mind that in most circuits these petitions are decided quickly by a motions panel, which will not delve deeply into the case. A defendant will often be best served by highlighting, where possible, the importance of the issues presented and explaining persuasively why the case is unlikely to reach a final judgment.

The Supreme Court began its new Term yesterday with oral arguments in cases involving whether arbitration agreements permitting only individual (non-class) arbitrations are enforceable under the Federal Arbitration Act, or prohibited by the National Labor Relations Act as an improper restriction on collective action. It is a case that essentially pits one federal statute against another. The decision could impact the enforceability of arbitration provisions with class action waivers in other contexts where there is another relevant federal statute.

The position in favor of the employers was perhaps most persuasively articulated by Jeffrey Wall, the Principal Deputy Solicitor General, when he said “It is a fundamental attribute of arbitration, and this Court said it three times, to pick the parties with whom you arbitrate. And our simple point is this case is at the heartland of the FAA. It is, at best, at the periphery of the NLRA, on the margins of its ambiguity, and you simply can’t get there under the court’s cases.” Justices Ginsburg, Sotomayor and Kagan seemed to push strongly against that position, suggesting that the right to collective action was at the heart of the NLRA.

The justices were so eager to get their questions and comments in that Justices Alito and Kagan had to reserve in advance the opportunity to ask questions and Chief Justice Roberts stepped in to referee that process. Perhaps the Court could develop a new electronic system for that process, with justices pressing a button to create a waiting line of sorts to ask questions.

Justice Kennedy, who is often the swing vote in these types of cases, seemed to be focused on whether it would make a difference if a small number of employees retained the same lawyer and sought arbitration individually. His questions and comments could be read to suggest that as long as that type of action were permitted, there would not be a conflict with the NLRA. Justice Breyer also seemed to focus on whether the answer to the question presented was not tied to a prohibition on class actions, but rather whether prohibiting employees from joining together even in a non-class action context would violate the NLRA, which he seemed to think would be prohibited.  Chief Justice Roberts posed a hypothethical in which the rules of the arbitration forum (selected by the employer) allowed a class arbitration, but only if there were 51 or more people who joined. The NLRB’s lawyer said that would be permissible. Justice Alito then asked about whether it would be permissible if the rules of the arbitration forum prohibited class arbitration, and the NLRB’s lawyer appeared to agree that that would be permissible, just as if there were no class actions allowed in court. If those propositions are correct, an employer could potentially reach the same outcome simply by selecting a a particular arbitration forum with a particular set of rules. Justice Sotomayor pointed out that many union contracts provide exclusively for arbitration, suggesting that the Court may not want to reach a result that could invalidate those. Paul Clement, arguing for the employers, pointed out that the notion of a class arbitration was unheard of when the FAA was enacted in 1925 (and the same would be true with respect to the NLRA).

This week the Eighth Circuit issued its long-awaited decision in a class action against State Farm involving the “labor depreciation” issue that I have covered extensively on this blog (see my August 14 post and others). State Farm prevailed on both the merits of whether its approach to determination of the actual cash value of property damage was proper under Missouri law, and on class certification. This important ruling will not end the labor depreciation class action litigation entirely, but it will be of substantial benefit to insurers litigating this issue in various jurisdictions across the country. It is part of a growing trend of decisions in favor of insurers, including decisions by the Tenth Circuit and Nebraska Supreme Court, and Minnesota Supreme Court.

The Eighth Circuit opinion is captioned as In re State Farm Fire and Casualty Company (in the district court the case was captioned as Labrier v. State Farm Fire & Cas. Co.).  The district court had denied State Farm’s motion to dismiss, finding depreciation of labor costs improper as a matter of law, and granted class certification. The Eighth Circuit granted discretionary appellate review under Rule 23(f) and reversed both the order on the motion to dismiss and the class certification ruling, remanding with direction to dismiss the complaint.  The Eighth Circuit also had accepted a mandamus petition with respect to certain discovery rulings that State Farm had challenged, but found it unnecessary to reach the discovery issues given its ruling on the merits and class certification.

The court of appeals held that, under Missouri law, actual cash value means “the difference between the reasonable value of the property immediately before and immediately after the loss.” (Slip op. at 6.) The court explained that the determination of the difference in fair market value, including the determination of depreciation in connection therewith, is generally a question of fact where there is a dispute over conflicting estimates. (Id. at 8.) The Eighth Circuit concluded that the Missouri Supreme Court would agree with the Minnesota Supreme Court’s decision that the application of depreciation to labor costs is a case-by-case factual question that “may only be determined based on all the facts surrounding a particular insured’s loss.” (Id. at 13.) The court thus concluded that “there are no predominant common facts at issue,” and thus certification under Rule 23(b)(3) was improper. (Id.)

In the interest of full disclosure, I represented the American Insurance Association and Property Casualty Insurers Association of America as amici curiae in this case.


There have been a substantial number of putative class actions filed recently against insurers involving the Medicare Secondary Payer Act (MSPA). These cases are typically filed by assignees of Medicare advantage organizations that have paid for medical services arising from auto accidents. The claim is that under the MSPA, the PIP/MedPay coverage under auto policies is primary and the coverage provided by the Medicare advantage organizations is secondary. Most of these cases are pending in Florida, and I have seen at least one case in New Jersey. I have seen some rulings denying motions to dismiss, mostly focused on whether there was an enforceable assignment of rights. I have not yet seen a decision on class certification. This is definitely an issue worth watching. Please let me know if you learn of significant developments on this issue.