The U.S. Supreme Court's denial of class certification for 1.5 million female employees in 2011 in Dukes v. Walmart was a clear watershed in federal class action litigation. Its biggest direct impact may be on procedure - when a class certification motion is heard, how much evidence is in play. But it has major strategic implications. Other key developments in class action litigation include the Ninth Circuit's scrutiny of recent class action settlements and rising interest in arbitration after the U.S. Supreme Court's ruling in AT&T Mobility LLC v. Concepcion
. California Lawyer
discussed those and other issues with Fletcher C. Alford of Dentons; Bernice Conn of Robins, Kaplan, Miller & Ciresi; Steven A. Ellis of Goodwin Procter; Layne H. Melzer of Rutan & Tucker; Brad W. Seiling of Manatt, Phelps & Phillips; Ben Suter of Keesal, Young & Logan; and Steven B. Weisburd of Dechert. The roundtable was moderated by California Lawyer
and reported by Cherree P. Peterson of Barkley Court Reporters.
Moderator: The Supreme Court's denial of class certification in the Dukes employment discrimination case (Wal-Mart Stores, Inc. v. Dukes, 131 S.Ct. 2541 (2011)) continues to reverberate in the class action world. Let's start with the way it changes the use of Daubert.
One significant aspect of Dukes
is its suggestion that Daubert
applies at the class certification stage. An open question, with a split among the circuits, remains as to whether courts must do a fulsome Daubert
analysis on class certification, or instead some more limited Daubert
review (Daubert v. Merrell Dow Pharms, Inc.
, 509 U.S. 579 (1993)). Some plaintiffs counsel have argued that the absence of a fully developed factual record at the early class certification stage may, in some cases, warrant a more preliminary assessment of whether the expert has reliably applied that methodology to the facts of the case.
A number of circuit courts, including the Ninth, had suggested for some time that, to one degree or another, Daubert
applies at the class cert stage. For me, the real watershed holdings in Dukes are that class certification cannot abridge the defendant's right to litigate individual issues. Another critical aspect is the notion that it's not enough for the plaintiffs to raise common issues. They have to demonstrate that those common issues are susceptible to proof on a class-wide basis. The other key aspect of Dukes
is the notion that the class certification analysis often necessarily overlaps with the merits.
, the position of defendants was - because these cases are very expensive to litigate - to try to keep the plaintiffs from getting into "merits discovery" before certification. It's an artificial distinction anyway, but Dukes
certainly blows that away and says merits do inform the decision on class certification.
I also think we're really going to have two very, very different systems in California over the next five or so years, with a much more rigorous analysis in the federal courts and something in the state courts that's like the good old days. So the question of how a defendant can get into federal court - or how a plaintiff can keep the case in state court - is going to be critically important.
brought various elements of class action jurisprudence from other federal courts together as a compelling reminder that the "rigorous analysis" to be applied at the class certification hearing is indeed "rigorous." The class certification hearing is now clearly akin to a mini-trial with due process implications. Dispositive issues must be amenable to resolution on common proof. Statistical extrapolations do not trump a defendant's right to litigate unique defenses to any particular plaintiff's claim. Due process considerations render 23(b)(2) certification inappropriate for cases seeking compensatory damages. And expert testimony is subject to Daubert
does tie in more to the atmospherics. After all, the Supreme Court, as we've seen in some of these cases, including Comcast
, has a pretty strong 5-4 split on some of the these class action-related issues (see Comcast Corp. v. Behrend
, 133 S.Ct. 1426 (2013)). But all nine justices believed there was no basis to certify on the record that that was before the district court.
I think Dukes
is going to get plaintiffs to focus more on their pre-lawsuit analysis and strategy, and force them to consider more carefully whether they are creating a potentially removable case or not. I agree we are going to see a divergence in how rigorous the analysis is going to be at the class certification stage in federal court versus state court.
Traditionally, California courts have held that they should look to the federal courts for guidance on class certification matters. But, recently, I've seen plaintiffs' practitioners going to great lengths to avoid Class Action Fairness Act, or CAFA (28 C.S.C. §§ 1332(d), 1453 and 1711-1715), removal and keep their cases in state courts. So it will be very interesting to see the extent to which California courts adopt some of the reasoning and principles of Dukes
and some of the other defense-friendly federal case law that has come down.
I've been handling an antitrust matter in the Third Circuit for the last five years. The Third Circuit has been requiring a rigorous analysis in terms of class since the Hydrogen Peroxide case (In re Hydrogen Peroxide Antitrust Litigation
, 552 F.3d 305 (3d Cir. 2008)). The Ninth Circuit too has been subjecting class actions to greater scrutiny since Dukes. I think it's just a matter of time before state courts also begin adopting a more rigorous approach.
Moderator: So let's talk about the issues that are raised by keeping consumer issues in state courts.
Even within the state court it depends where you are. If you're venued in one of the complex courts, you get more of a rigorous analysis even though the courts may be applying state law only. And if you're not in one of the complex courts, it's very easy for the judges who aren't familiar with class action cases to kind of go back to the easy decision that says, well, this is not a merits decision. I'm just going to certify it and figure it out later.
A vexing problem in state court consumer litigation is the breadth of standing in UCL (California's Unfair Competition Law) cases (Cal. Bus. & Prof. Code 17204). California has endorsed "gatekeeper" standing: if the class representative bought the product or lost money or property as a result of some claimed false ad, then that may suffice, even if many others in the class were not similarly misled or injured. This can sometimes spill over into a less rigorous class certification analysis. Federal law, whether because of Article III or Rule 23, tends to be less forgiving. These differences often lead plaintiffs to prefer state court.
I'll offer a prediction. I think within five years the California Supreme Court will issue an opinion that says essentially the same things that are in Dukes
and similar federal decisions. I think the state Supreme Court will eventually work it out in the medium to long run.
Moderator: Brad [Seiling], you offered the opposite scenario for five years out, where we would have a pretty serious divergence.
Yeah, I still see that as the case. It's so difficult to get the California Supreme Court to take up civil issues. And it really almost requires that there be serious confusion and conflicting results in the lower courts for the California Supreme Court to step in.
You also have to take into consideration California's UCL, which is very, very broad. And you've got to remember that it is a means of redressing a wrong that the attorney general might not otherwise be handling. So when it comes to UCL actions, I think it may be longer than that. I also continue to believe that, because of the UCL, the standing issue may not exactly track federal class action law.
I tend to think the tracks have gotten closer to each other, but I don't see them merging. The UCL is that unique body of law - less unique now that you need a plaintiff with actual standing - but I think that California is going to continue to protect its consumers.
indicted a statistical model that would have presumed some degree of discrimination based on an extrapolation method. California state courts, on the other hand, have a history of using statistical sampling and extrapolation to establish liability, particularly in the employment context. This is another potential point of departure between state and federal courts.
Moderator: Does this divergence in laws and case strategies mean that we're going to have factually different cases in state court and federal court?
I think what it's going to mean practically is an increasing fight over CAFA removal. In Standard Fire Ins. Co. v. Knowles
(133 S.Ct. 1345 (2013)), we saw plaintiffs try to plead around the CAFA "amount-in-controversy" requirement through a stipulation that class counsel was not going to seek more than $5 million on behalf of the class. And the Supreme Court held that class counsel can't make that stipulation before certification.
Another factor is the decreasing resources for state court judges. Contrast that with the Ninth Circuit in the last year directing district courts to oversee the settlement process much more closely, and you're going to see a substantial divergence in practice between the scrutiny given at state court in California versus federal court.
My experience has been to the contrary. It's certainly true that in the Ninth Circuit we've seen a series of decisions starting with In Re Bluetooth Headset Prods. Liab. Litig.
(654 F.3d. 935 (9th Cir. 2011)), Dennis v. Kellogg Co.
(697 F.3d 858 (9th Cir., 2012)), and finally Radcliffe v. Experian Information Solutions Inc.
, 2013 WL 1831760 (9th Cir. 2013), in which the Ninth Circuit has ordered the district courts to give greater scrutiny to attorneys fees awards for class counsel, to incentive awards to the class rep, and to cy pres distributions. But I've seen scrutiny of settlements in state court practice recently, too. I had a case in San Francisco Superior Court that I litigated for five years. And we reached a settlement in which both sides stipulated to an attorneys fee award of up to $2.3 million to class counsel. But the judge sua sponte reduced the award to $300,00, slashing it by $2 million.
What's also alarming about the Ninth Circuit decisions is the things that have been tripping up settlements are not really the material terms of these deals. Incentive awards to class reps and cy pres awards are not the guts of most of these settlements.
So it gives objectors, who are the bane of the existence of both plaintiffs lawyers and defense lawyers, leverage to make mischief after months and sometimes even years of settlement discussions.
Moderator: What would be a better mechanism? Are there other mechanisms out there to deal with what probably the court agrees is the problem - that there are settlements that are missing the boat?
I don't think the main problem is objectors. But there really is a legitimate issue about cy pres funds in class actions. I think there's a tendency among defendants to settle class actions where they are fairly certain that there won't be significant claims submitted by the class, and I don't think there's anything wrong with requiring cy pres awards to bear some relationship to the wrong for which the settlement funds are being paid. I don't think anything the courts do is going to either encourage or discourage objectors.
In the Kellogg
case, they just reached a new settlement where the value of the class settlement dropped from $10 million to $4 million. They did identify new, supposedly relevant cy pres entities to which cy pres funds will be paid, but the attorneys fees in the case went from 19 percent to 25 percent. So the fees for the lawyers remained approximately the same, despite the fact that the value of the class settlement dropped by $6 million. Again, I'm a plaintiffs lawyer, and I don't want to make it harder for plaintiffs, but I also think that the primary concern really ought to be the quality of settlement and the class members - not whether there are going to be objectors.
Well, I think we're really getting to the heart of the problem: Many class actions are attorney-generated and driven by plaintiffs attorneys fees, and that creates incentives to settle that can sometimes yield results not in the best interest of the class members. The reality is that often all the defendant cares about is getting out of the case as cheaply as possible, and that means paying off the plaintiffs attorney and giving some illusory relief to the class in the form of cy pres. What we're seeing is that that's more difficult to accomplish now, and that's a good thing.
A significant cy pres fund may reflect the lack of merit to the case. If you tell the class here's a $10 million pot and you can claim against it for whatever you want and you end up having $8 million left, to me that indicates that the class really didn't care about this and that these really are lawyer-driven issues. If only a half-percent of the people in this class care enough to check a box on a simple claim form, maybe it's not a class at all and never should have been.
This gets to the fundamental question of the role of class actions. To what extent is it to benefit the class members? That's one component. Another component is to force defendants who are alleged to have acted wrongfully to give up the money that they've gained as a result of the unlawful practice or the allegedly unlawful practice. And sometimes those two goals don't line up perfectly well. If a defendant has unlawfully taken $5 from 10 million people, there's a pretty good argument that the defendant should give up the $50 million. But people might not bother to submit the claim form for the $5. So there's a tension between these two goals.
I think that's where the Ninth Circuit was heading in the Inkjet
ruling (In re: HP Inkjet Printer Litigation
, Case No. C05-3580-JF, 2013 WL 1986396 (9th Cir. May 15, 2013)), where the court said go back and look at the number of coupons that were actually redeemed, evaluate the attorneys fees based on the redemption value, and then separately evaluate the injunctive relief. I think this is an instance where the court is taking a much closer look at what the goal of the case was, what it actually achieved, and what its real value is, which is what CAFA requires.
I agree with Steve [Ellis] that there two goals. But I would say they're two goals of consumer protection. And I think the class action mechanism should focus on the class. If there is someone who's going to focus on punishing the defendant and getting the defendant to disgorge profits, that should be the regulators, the attorney general, the FTC, or some other regulatory agency.
Even if we focus on the "class," an important issue is what's the appropriate test for causation of economic injury. If someone buys a product for $5 and uses or consumes it, but then claims some ancillary deception, such as where the product is made, what is the economic value of the deception? It may be zero. I don't think any economist could credibly say it is $5, and after Comcast
, I think we'll see more Daubert
challenges to damage models and theories that unduly inflate the economic value of a deception. The UCL is limited to awards of restitution. The class action procedural device should not be employed to compel monetary awards where there is no causal connection between the particular deception at issue in the case and the sum of money sought.
Cy pres funds strike me as an oddity in UCL cases, where the mandate is to make restitution to those who actually lost "money or property." By definition a cy pres fund is inconsistent with this legislative mandate.
One thing I've seen in my practice is that both courts and class counsel are increasingly pushing back on the idea of a "claims-made" process, and are insisting on some sort of automatic relief to all class members without the need for a claims process.
I'm seeing plaintiffs looking for statutes that have a strict liability component coupled with statutory damages or penalties like the Telephone Consumer Protection Act of 1991 (47 U.S.C. § 227) cases, cases under the Fair and Accurate Credit Transactions Act of 2003 (Pub. L. No. 108-159), and nursing home staffing cases, where California's Residents' Rights Statute (see Cal. Health & Saf. Code § 1430(b)) allows statutory damages of $500 per violation. It doesn't sound like a lot, but when you're talking about hundreds of thousands or millions of potential violations, you have a big club to leverage a settlement.
I do a lot of securities litigation on the defense side, and I'm seeing a concerted effort by plaintiffs to avoid federal law completely, both in terms of how they plead the case under state law at the get-go, and how they strive to avoid removal. In a recent municipal bond underwriting class action, only state law claims were pled, but we successfully removed the case to federal court under CAFA, while also contending that the federal law should apply. We are seeing a lot of cases where plaintiffs are trying to avoid the "good" federal securities law (from the defendants' perspective) in favor of the more plaintiff-friendly state blue sky laws and consumer protection type laws.
Moderator: Let's return to the impacts of Comcast, especially for California cases.
In California, case law has been very problematic for the defendants in holding that variations in damages have little or no relevance to the class certification question. Comcast
gives hope that if the state courts fall in line with the federal courts, they will give increasing weight to variations in damages.
It remains to be seen how impactful Comcast will be. If it is interpreted as requiring class-wide proof of damages - and as affirming that individual variations may suffice to defeat certification - then it will be very significant. It certainly reconfirms that "merits avoidance" at class certification is dead. It also holds up the 23(b)(3) predominance inquiry as one requiring even more demanding scrutiny than the 23(a) analysis. Comcast
certainly continues the mandate of Dukes that the trial court must be analytically rigorous in all respects before concluding certification is appropriate.
The lesson I've come away with is it's not that there are problems with certification that arise from variations in class member damages, but a plaintiffs expert better have a model that can be used to calculate those damages in all their variations. If, instead, you're going to have to, for a class of 10,000 or 10 million people, do individual calculations taking into account idiosyncratic factors for each class member, that is going to cause a problem on certification.
Moderator: Turning now to the Supreme Court's landmark arbitration ruling in Concepcion (AT&T Mobility LLC v. Concepcion, 131 S.Ct. 1740 (2011)), I wonder how you're seeing it applied.
strongly confirms federal preemption under the Federal Arbitration Act (9 U.S.C. §§ 1-9), but many questions remain and are now being litigated. Some courts continue to view arbitration agreements negatively. Since many companies in a variety of consumer-related contexts have recently embraced arbitration in the wake of Concepcion
, I think we will see arbitration issues continue to be hotly litigated in the courts in the years to come.
The courts are looking very closely at whether there's actually an agreement to arbitrate in each case. Certainly, if the question in the case was whether the plaintiff was deceived into purchasing a product or deceived to enroll - and they don't get that arbitration agreement until they get the actual product - you have some interesting questions as to whether the arbitration agreement is valid. The other thing that I'm seeing in the employment and consumer contexts is plaintiffs lawyers will bring many individual arbitrations, if the court upholds the arbitration clause. And then the company says, "Be careful what you wish for. Now suddenly we're defending 50 cases and we've got to win them all because if we lose one, plaintiffs will argue the decision is collateral estoppel against us."
raises a number of interesting questions. First, the arbitration provision at issue in that case included very consumer-friendly provisions for cost shifting, fee shifting, and burden-of-proof shifting, and it will be interesting to see to what extent the courts attempt to limit Concepcion
to those kind of scenarios. On another level, a number of my clients are highly regulated. And they're not yet adopting arbitration provisions because either they're concerned about the reaction of regulators or, frankly, they're not fans of arbitration, which has its own risks: You may avoid a class, but you may get thousands of individual claims that are very expensive to litigate because you have to pay for the arbitrator. Moreover, the arbitrator may not be bound by rules of law, and appellate review is usually very limited. So I'm not sure Concepcion
marks the demise of class action litigation.
I also represent businesses that are rethinking arbitration. Frankly, the last several arbitrations that I've handled have been unbelievably expensive, particularly where there are multiple arbitrators. They haven't produced a terribly expedited result, and, as you say, there are some serious concerns about the limitations of arbitration. My clients who are businesses are rethinking whether arbitration really does have the same advantages today as it once offered. Although, now, as courts become more congested because of budget cuts, you might get a more expedited result with arbitration.
Securities broker dealers do not have a choice. They have to arbitrate regardless of whether or not there's an arbitration agreement if the claimant wants to arbitrate. The hot issue right now, in view of Concepcion
, is whether class action waivers are appropriate. FINRA (the quasi-governmental Financial Industry Regulatory Authority) has a rule that states it will not handle class actions, so one way claimants attempt to avoid arbitration is to bring a class action in court. Yet, after Concepcion
, a FINRA panel ruled in a case involving Schwab that Concepcion
displaces the FINRA rules. So it's a pickle. There's going to be some interesting law that comes out of the Schwab FINRA case.
In the "false advertising" context, particularly with consumer products, Concepcion
has had a limited impact. It has however caused retailers and manufacturers to look more carefully at ways to potentially incorporate an arbitration agreement into the sales transaction with the consumer. As for whether there are disadvantages to arbitrating cases that would have otherwise been aggregated into a class action, I think in most cases the answer is "no."
In my experience the greater risk comes from claims being aggregated. Particularly in the consumer arena, claims tend to be driven by a small group of persons often with idiosyncratic complaints. That is why when certification is denied, the underlying case, even as individual actions, tends to die. If you aren't willing to stipulate to class certification in advance (and I haven't been involved in a case where that ever made sense outside of settlement), then arbitrating individual claims is preferable.
And there's another dichotomy taking place within clients; it's the theoretical impact of the Concepcion
decision versus the practical realities of doing business. After Concepcion
, you had a lot of in-house lawyers looking at an arbitration clause as a way to insulate the company from class action liability. When the marketing people get involved, they do not like the fact that, to make the clause enforceable, they need to put it up front so customers understand it's part of the deal. The marketing people say, "We're not selling an arbitration clause."
One arbitration-related issue recently before the U.S. Supreme Court in Oxford Health Plans v. Sutter
(2013 WL 2459522) involved a pretty direct challenge to the Court's prior decision in Stolt-Nielsen
(Stolt-Nielsen SA v. AnimalFeeds Int'l Corp.
, 559 U.S. 662 (2010)), which held that parties can't be compelled under the FAA to do class-wide arbitration unless they explicitly agree to it. In Oxford, a broad arbitration provision was read by an arbitrator to permit class-wide arbitration, even though it didn't specifically mention agreeing to class arbitration. So a party was forced not only to arbitration, but class arbitration. Given the absence of procedural protections, appellate review, and a lot of other reasons, that is not a good scenario. [Editor's Note: The Supreme Court ruled in Oxford
after the roundtable met. It held that Stolt-Nielsen
did not govern because the Oxford
arbitrator properly based his action on his interpretation of the parties' agreement, whereas the record in Stolt-Nielsen showed that the parties hadn't agreed to class arbitration; the Oxford
court ruled that any errors the arbitrator made in interpreting the contract were not subject to appellate review.]
Moderator: With so many issues of law settling out after Dukes and Comcast, what's the single biggest event for you in class action litigation in the last year or two?
clarified that it is proper to consider at the class certification stage arguments and issues that may overlap with the merits. This eliminates procedural arguments that plaintiffs counsel often made to dodge difficult defense arguments, and it now properly focuses attention on the critical question of whether plaintiffs counsel have satisfied their burden to prove that all Rule 23's class certification requirements have been met.
In addition to Dukes
, it's the greater judicial scrutiny of settlements - and more caution on the part of plaintiffs counsel as a result in negotiating and structuring settlements. Now I think we're going to see more and more contested certification decisions and even class-wide trials on the merits.
is a very interesting decision and gives a lot of good arguments to oppose certification. But I've always had a cynical view that the class action analysis for most judges is a know-it-when-you-see-it proposition. And sometimes, frankly, the difference between state and federal court and the difference between judges in state court is how much experience that judge has had with class actions.
Recent case law developments confirm that due process matters. Defendants have a right to adjudicate individual defenses. This right cannot be ignored by a statistical "trial by formula." Similarly, a 23(b)(2) class cannot be certified when there are claims for compensatory damages because unnamed class members have a right to pursue those individually. This greater due process scrutiny has led to the death of splitting "merits versus class" discovery, and the "merits avoidance" rule. It has empowered defendants to resist being bullied into settlements and instead to fight unmeritorious cases at the certification stage and beyond.
Whereas 10 or 20 or 30 years ago the class action decision was often made primarily on the pleadings, it's now made almost exclusively on the facts. That's positive, but it means class actions are much more expensive for both plaintiffs and defendants, and there's a risk for plaintiffs to invest in an action only to find out three or four years into litigation that it's not going to be certified.
What has come out of the U.S. Supreme Court majority in the last two, three years evidences a fairly strong anti-class action agenda. And whether the justification is to protect due process rights or to simplify or make class actions more fair at various stages, we are witnessing a clear move away from the looser standards we had in the '80s and '90s.
The last few years have demonstrated what I think is the very pro business, pro corporate stance of the U.S. Supreme Court. Personally, I find the Concepcion
decision enormously troubling in that the U.S. Supreme Court has endorsed what are basically adhesion arbitration clauses requiring consumers to individually arbitrate their claims and essentially removing consumer claims from judicial scrutiny.