CLASS ACTION ROUNDTABLE

Major class actions involve a number of complex questions: What law should apply when a case involves state law claims and a nationwide class? How should a nationwide settlement be structured when there are significant variations in state laws? To what extent must you be able to identify class members? These are just a few examples of the questions lawyers and courts must grapple with in class action litigation.

Needless to say, major class actions are fraught with uncertainty. In this roundtable discussion, members of both sides of the bar come together to examine how courts and lawyers are dealing with these issues in recent cases, including an ongoing antitrust class action against Qualcomm involving a quarter-billion-member class -- the largest in history. Participants include, from the defense side, Steven Ellis of Goodwin Proctor LLP and Michael Geibelson of Robins Kaplan LLP, and from the plaintiffs side, Dena Sharp of Girard Sharp and Steven Williams of the Joseph Saveri Law Firm. The discussion was moderated by Ben Armistead of the Daily Journal. This transcript has been edited for length and clarity.

MODERATOR: How should courts address choice of law issues in nationwide class actions such as the In re Qualcomm Antitrust Litigationand In re Hyundai and Kia Fuel Economy Litigationcases?

MICHAEL GEIBELSON: If there are careful findings and a reasoned decision about the reasonableness and fairness of the settlement, you shouldn't have any problems.

The settlement process should not be viewed as a game. It involves real businesses, and real dollars are at stake, real people with real damages at issue. And if the old adage is true that bad cases make bad law, then get those cases resolved as quickly as possible with sufficient evidence to make findings, that can be protected on appeal. But the result in one state is just as uncertain as the result in another state. And so, it's incumbent upon the lawyers to figure out what evidence is necessary to make those settlements stick with sufficient findings.

STEVEN ELLIS: The result in Hyundai is not surprising. But something was bothering the judges. First, the panel rejected the settlement In re Hyundai and Kia Fuel Economy Litigation, 2018 DJDAR 767 (9th Cir. Jan 23, 2018); then, three judges dissented from the en banc decision. In re Hyundai and Kia Fuel Economy Litigation, 2019 DJDAR 4888 (9th Cir. June 6, 2019) (en banc). Some of the judges were troubled by variations in state law. The majority made short shrift of that by saying the issues regarding variations of state law weren't adequately preserved in the record.

The lesson is not to assume that the court will approve a nationwide settlement based on state law, even though most do get approved.

DENA SHARP: Under Rule 23, in the settlement context, manageability really isn't an issue, and shouldn't be something courts glom on to.

There has been sloppiness in some decisions grappling with the differences in state laws. It relates to federalism. To me, Hyundai got it right. I don't see the case materially changing the way cases get settled, other than lawyers on both sides being more careful about how they craft the settlement and the release.

STEVE WILLIAMS: I agree. As for how should it be done in a litigated class, Qualcomm is different -- a single defendant, a California corporation. In re Qualcomm Antitrust Litigation, 17-md-02773 (N.D. Cal., filed April 6, 2017)]. The decision to apply California law is a more traditional choice-of-law analysis. That should not be controversial. But there is substantial concern in the 9th Circuit about multistate class settlements that give short shrift to other states' claims, and I think that's going to continue to be the case.

In the settlement context, when the defendant and class counsel are proposing a settlement, the issue of who is watching out now that everyone is on the same side is heightened. The analysis then turns to issues like collusion, and was the settlement fairly negotiated.

Both sides have to be more careful. On the plaintiff side, in not offering a release that they can't support; on the defense side, by being realistic about how about far can you go -- no defense counsel wants to recommend a settlement that is thrown out two years later.

GEIBELSON: It strikes me that one of the things that we're talking about then is whether it's a manageability issue or a predominance issue. There is a spectrum between complexity on one side and simplicity on the other, particularly in the context of settlements with seemingly more complex structures with different allocations under a particular law or a different statute of limitation, or different types of claims.

How do you approach the comparison of complexity in the settlement with payment allocation issues and liability issues?

ELLIS: In a nationwide class action settlement I had years ago, we put the states in buckets -- if you were in one state, you got 50% of your damages; if you were in another with weaker laws, you got 20%. It wasn't done with mathematical certainty, but we tried to come up with a fair way to reflect the fact that it was not even. We didn't have 50 different variations. That would have been too complex. Was it perfect? No, but I think it was fair, adequate and reasonable, and the judge agreed.

WILLIAMS: I break things down into two sets. First is the question of the aggregate liability. Second is the determination of what any particular class member is entitled to.

That makes it easier to focus on liability, because that is an issue you can litigate. Differences among the rights of each class member in terms of how much they recover shouldn't defeat class treatment. That can be subject to allocation processes and special masters. And the quantum of proof necessary to establish a claim is not as burdensome as proving liability because that is how much you're going to pay.

When you get to the distribution process, there's more flexibility. You're trying to avoid having everyone do their own trial. If you force everyone into a second series of proceedings where they are putting on their own cases, you haven't really gained all the benefits of the class device.

SHARP: I would add one or two further dimensions. First, the landscape is materially changed when there is a single unifying claim for the whole class. Something familiar, like a federal claim under RICO, would make a baseline damages allocation easier. Something more ethereal, like unjust enrichment, could be a tie that binds the whole class together. That is one way to deal with those complicated allocation issues.

Second, that kind of procedure would give courts comfort that the settlement is being tailored to the claims, which is the fundamental question that a lot of these decisions raise.

In the antitrust context, there's a real question of, are we going to only try to settle it on behalf of those Illinois Brick repealer states [that allow indirect purchasers to have standing to bring antitrust claims], or are we going to try to settle nationwide? Doing the latter is a risky move, especially here in the Northern District where we have these settlement guidelines. The first question the guidelines ask is, "What are your claims, and is your proposed settlement different than the claims that you plead in your complaint?"

The sliding scale approach, if administered well, would give the court comfort that the lawyers are focused on what the claims are, as opposed to just getting a pot of money and trying to whack it up, and then give the lawyers a piece of it.

GEIBELSON: That is exactly the reason that I asked the question, though. I don't know that I have had a case where a defendant was up in arms about an allocation between one or another; it was always about just throwing a lump of money that they were willing to pay to get rid of this problem, and leaving to a separate discussion, and, frankly, oftentimes a mediator, the decision about how that should be allocated.

SHARP: I think it's very relevant. It's one of the go to arguments for us on class certification when one of the arguments we see in opposition is, "But there are uninjured people in the class! And you're not able to identify every single one of them!" The common sense response is, since when did the defendants care about how the money is allocated? What they really care about is exposure.

Once a class cert order has been entered, I don't think any of us have really seen defendants truly concerned about whether this or that class member gets paid.

WILLIAMS: In almost every class settlement agreement, there is a provision stating, in effect, "I have no further interest in what you do with this money. You're paying for the release I'm getting."

ELLIS: If there is an allocation made, with or without the defendant's involvement, that's so unfair that a class member could say some subgroup of a class wasn't adequately represented -- that could be the basis to bring a collateral attack on the settlement as matter of due process. In that extreme situation, defendants are going to care. Otherwise, you're right, defendants ultimately don't care.

MODERATOR: Qualcomm recently submitted a brief that seemed to argue that the class -- 250 million people, supposedly the largest and most diverse ever in the United States -- is simply too big to certify. Is that ever a valid argument?

ELLIS: If there ever was a case that was just too big to certify, perhaps it's the Qualcomm case. But it's unlikely that the 9th Circuit or the Supreme Court is going to say there are some classes that satisfy all the requirements of Rule 23, but are just too big to certify.

Dukes v. Wal-Mart Stores Inc., 131 S. Ct. 2541 (2011), comes to mind. There was a sense -- maybe fair, maybe not -- that the plaintiffs were overreaching. It might happen in Qualcomm. Plaintiff's counsel won in district court, but the question I would ask is, why was it important to sue on behalf of everyone in the country? It might have been safer to focus on just California, or just repealer states.

Twenty-six or 28 states have departed from the Illinois Brick Co. v. Illinois, 431 U.S. 720 (1977), rule, which under federal law bars indirect purchasers from having standing to bring an antitrust claim. Why not sue just in those states? Why risk the whole class failing by suing on behalf of everyone?

WILLIAMS: Even if it has only been pled in repealer states, you might have cut the class from 100 to 70 million, and it wouldn't have made a big difference. Also, California law governing a California corporation that does things from California that affect people around the country -- that is an instance where choice-of-law favors California.

On the other side of the scales is that the conduct at issue is the same conduct as to everyone. The Department of Justice and the Federal Trade Commission have starkly different views of whether there's an antitrust violation. But the conduct at issue affected everyone in the same way.

We're not going to see a decision that says we are going to not approve certification simply because of the size, but I think this case has issues that make it likely to go to the Supreme Court.

Rule 23 says that if someone causes harm to everyone, they shouldn't get a pass because that's just too many people.

SHARP: Wal-Mart is a case that comes to a lot of people's mind when they look at Qualcomm. There are some distinctions, though. The employment context in Wal-Mart couldn't be a lot more different than the antitrust claims in Qualcomm.

In Wal-Mart, individual inquiries were permitted to be made. The court focused on the due process concerns and the individualized challenges to each class member's rights to recover. It was under Title VII, where an individualized inquiry is basically a rebuttable presumption. That is wildly different than Qualcomm.

I think we can agree there are individual differences in Qualcomm, but are they significant enough to swamp the predominating issues? There are two issues here. First, was there an antitrust violation, and I don't think there is any dispute that's common to the whole class. The second issue is injury, which may implicate some individualized issues. But we come back to this concept that the class remained certified, there is an aggregate damage award or settlement, and that ultimate exposure is what Qualcomm is facing. So, I don't think the individualized inquiry formWal-Mart can be applied in the antitrust context.

This appellate Qualcomm brief came across as a hit parade of all of the top issues in class actions right. There is the, "you're too big to not fail" new giant class. The, "you have 100 million uninjured class members." We have arguments that the plaintiff's failure to provide a plan for the defendants to contest liability should undermine class certification.

As a kicker, you have Qualcomm questioning whether the plaintiff's economist conducted a substantial enough inquiry. And that is interesting, because what Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036 (2016), says is that a battle of the experts shouldn't be determinative at the class certification stage. The question is just whether "no reasonable juror" could buy the pitch that the expert is giving.

I'm concerned the size of this class could provide a platform for any one of those hot issues to become something that either the 9th Circuit or the Supreme Court focuses on.

GEIBELSON: As demonstrated by our discussion here, any time you have an argument that you need to put air quotes around, because of allusions to "too big to fail," that makes the argument suspicious.

If we give the judges the opportunity to do what I think we're all talking about, which is a rigorous predominance analysis, then they'll do their job. Some cases will turn out to be too big to certify, but not just on the numbers alone; and some cases will be totally appropriate, despite their numbers.

MODERATOR: Earlier this year the Northern District issued new settlement guidelines. How have those been playing out?

SHARP: We haven't found that the guidelines cause us to do things much differently. We might include another chart or two in our papers. But these are kind of in line with the 2018 amendments to Rule 23. Also, frankly, they're there to give guidance to those who are maybe less experienced in the class action realm.

WILLIAMS: That's right. The court sought the input from both sides of the bar before they were enacted. Not everyone knows how to get a settlement approved. And they're not one size fits all. In antitrust cases, there never is a reversion; the defendants get a sum, and it's distributed to the class. That is one example.

They allow the courts to have uniform information. Sometimes we'll say to a judge, "You should approve this because it's just like that other settlement." Well, how does the judge know? In prior practice, there was less ability to compare apples to apples. The guidelines make that easier, in particular, by virtue of the "nutritional chart." Much like you see on food labels, we now put in the settlement papers a chart of total settlement fund, total expenses to be repaid, total attorney's fees, so that judges can take a quick look and get a snapshot of what this deal really is about.

ELLIS: True, the guidelines are useful for people who are less experienced. They are consistent with the practice of experienced lawyers on both sides. They are designed to be guidance, which I think is an important word. There are certain details that I might agree or disagree with in a particular settlement. If judges are going to use these as helpful suggestions for streamlining the approval process, that's great.

There certainly will be cases where a settlement is not going to fit perfectly within the guidelines. And that should be fine as long as you have the opportunity to explain why. If it gets a too rigid, I would have more issues with it. But if these are going to be the equivalent of rules, we'll follow them.

MODERATOR: Northern District Judge William Alsup has a somewhat unique practice of not allowing settlement negotiations prior to class certification. Is that helpful? Is it fair?

ELLIS: I haven't had a class action before Judge Alsup that has been affected by this. But, as a general principle, this causes me some concern.

There's a policy of the courts to encourage fair, reasonable and adequate settlements. What makes cases settleable is uncertainty. If you tell the parties they can't negotiate until certain issues are resolved, that makes it harder.

If Judge Alsup were to apply this rule rigidly, that would make it harder to settle cases.

I don't think the rule ends up accomplishing the intended goals.

WILLIAMS: It's a difficult balance. Judges should have the ability to control the cases that are in their courtroom.

I know that Judge Alsup permits settlement discussions when interim counsel is appointed, but also that in the case the 9th Circuit considered, he denied that appointment, which made it impossible to discuss settlement. [Logitech, Inc. v. U.S. Dist. Court for N. Dist. of Cal. (In re Logitech, Inc.), 19-70248 (9th Cir., Sept. 12, 2019).]

There are certainly cases in which parties wanted to discuss settlement early. Take the Volkswagen case. [In re: Volkswagen "Clean Diesel" Marketing, Sales Practices, and Products Liability Litigation, 15-MD2672 (N.D. Cal., filed Dec. 8, 2015).] No one doubted that would go into settlement mode very quickly. Judge Charles Breyer made that happen. But, even when the judge is enforcing it, the parties have reasons that they want to get to that phase quickly. When both parties really want to settle but are prohibited from doing so, that could cause time and resources to be spent that wouldn't otherwise be spent.

I will say that in Judge Alsup's court, you won't waste a lot of time, because it's done quickly. Also, the 9th Circuit judges didn't say any of them would do this, but that it wasn't to a sufficient level of legal error that they could tell Judge Alsup not to.

GEIBELSON: There is a tension between what is early and practicable in light of the stage of discovery. And there certainly is some merit to saying, "if we don't know enough to reach the certification question, how can we know enough to say whether a settlement is fair and adequate?" But that greater level of uncertainty can foster settlement. In balancing the concerns settling parties and the court, I'd love to be in his court to hash these issues out in an appropriate case where everyone knew all of the facts. But it seems that the amount of information submitted along with a settlement to propose as being fair, adequate and reasonable is going to be more difficult if you are required to do it earlier than you might otherwise in the usual case where class certification takes some substantial period of time to get to.

WILLIAMS: It is worth noting that class actions in federal court can take years and can cost tens of millions of dollars. So, there are great deals of resources going into cases that might be avoided.

GEIBELSON: The idea that we're going to make an interim determination can sometimes be troubling. I tried a case where we got decertification at trial, which was good for my client, but it took a lot of time and resources to get there. And it was because there was a certification decision that was made too early in the case, without the benefit of more fulsome discovery. And that prevents settlement and expends resources unnecessarily.

MODERATOR: The DOJ recently announced its appellate amicus initiative. What is the proper role of these briefs?

GEIBELSON: If they're on my side, they're fantastic.

WILLIAMS: DOJ actually submitted them in both Qualcomm cases. [In re Qualcomm Antitrust Litigation, supra; Federal Trade Commission v. Qualcomm, Inc., 5:17-cv-00220-LHK (N.D. Cal., filed Jan. 7, 2017)];. Federal antitrust laws were designed to provide concurrent government enforcement, both civil and criminal, and private enforcement. And private enforcement was deemed by the drafters, and has been deemed by the courts, to be on par with criminal enforcement.

If I was doing a federal case, my view would be the DOJ does not have any more interest in the case than anyone else. The concern I have is that courts treat the DOJ with great weight.

When the attorney general comes into the Qualcomm case, which is based purely on state law violations, that's even more concerning. The DOJ Antitrust Division does not have a justifiable position on how California law should be interpreted. That is not their window to wash.

ELLIS: It sounds like the solution is to have the California DOJ weigh in. And certainly, in the current political environment, if the federal DOJ says something is up, we know the California DOJ will say it's down.

SHARP: Right. There's also a pretty interesting interplay between U.S. DOJ and the FTC. We're used to seeing the FTC in antitrust cases; we're used to seeing some of the states' attorneys general; but to say there is some intra agency tension that is bubbling to the surface in these amicus briefs would be an understatement. There's a significant shift in the dynamic when either of those agencies expresses interest.

MODERATOR: Is the Class Action Fairness Act threatening to displace state law with federal law?

SHARP: Yes. I have a good example in the context of attorney's fees. It's not just courts, but also lawyers who are litigating in federal court, who may not switch gears to think about the application of state law as it relates to attorney's fees in certain cases.

There is a school of thought -- including one set out by one of my partners, Jordan Elias ["Cooperative Federalism in Class Actions," 86 TENN. L. REV. 1 (2018)] -- suggesting we should be more mindful about the application of state law, and to consider certifying novel outcome-determinative questions under state law to the state supreme court, and make decisions that down the road could result in a more of a consistent body of law between state and federal courts.

WILLIAMS: I agree. CAFA was a tort reform bill. It was focused on just a few areas of law, but it sucked in virtually every type of class action. As a consequence, state courts lost much of their ability to develop their own law. Decisions are now being made in complex cases by judges, sometimes thousands of miles away, who are asked to rule on too many things. Think of cases where you have the laws of 37 different states and three different areas of law; you started seeing decisions that are homogenizing the differences simply as a matter of trying to come to a decision. That is troubling. Consumers' ability to sue companies for harms they caused when there was no privity -- that came out of the state courts.

All of the developments in law that I can think of that gave consumers greater rights and abilities to use the courts to vindicate their rights came out of state courts. Those decisions are going to be going to federal courts now.

GEIBELSON: Already via CAFA being applied to lots of different kinds of the class actions, there is a residue of no-injury class actions that have been left in the state courts, where people are pursuing statutory penalties and different sorts of civil penalties, such as Private Attorneys General Act claims.

Concerns about the state courts making more law that a federal court from farther away might rely upon become worse if you're having courts decide those issues in class actions that relate to things that would never get into federal court because there's no injury. And while the statutes say what the statutes say, different judges will react differently, and make decisions based, in part, upon whether something could be managed, and is more appropriate to be brought, as a class action where nobody's been harmed.

ELLIS: These issues are difficult to address in the abstract. CAFA was a reaction to a specific set of problems that were perceived as giving the plaintiff's counsel in certain states too much power. Right or wrong, those were the animating concerns.

If you go back to the founding of our country, there was a notion that state and federal court systems could exist side by side. And through most of our history, there was a sense that it shouldn't matter. I think this is an area that may well be ripe for some further reform.

MODERATOR: Moving on to the California Supreme Court. In Noel v. Thrifty Payless, 7 Cal. 5th 955 (2019), the court addressed to what extent a class must be ascertainable. What are the takeaways from that decision?

ELLIS: We've seen in federal courts over the last 10 or 15 years a rise and fall of ascertainability as an independent requirement for class certification. It was all the rage for years. That has subsided, although ascertainability is still a very real requirement in certain federal courts, including the 11th Circuit, and I believe the 3rd Circuit, and maybe a couple of other circuits.

Noel was trying to grapple with whether we want a strong or weak version of the ascertainability requirement. The court went back toWeaver v. Pasadena Tournament of Roses, 32 Cal.2d 833 (1948), to find precedent for requiring some sort of ascertainability. But it landed on this notion that the class should be defined in objective terms; should be based on common transactional facts that make the ultimate identification of class members possible. But the court was clear that at class certification, the plaintiff didn't have the burden of being able to either identify all class members, or necessarily explain how all class members would be identified.

I don't agree with everything in the court's opinion, but it landed pretty near the majority view in federal courts.

WILLIAMS: I agree.

GEIBELSON: We try a lot of cases. And viewed through the lens of trial, Noel really just kicks the can down the road. I don't think ascertainability will ever die, but rather will be considered or continue to be considered, as the Supreme Court noted, as an issue of manageability.

The outgrowth of this will be that more California state court judges will demand trial plans in order to figure out how this case is actually going to be tried, and to determine whether there is a community of interest that people can actually prove with common proof.

In the absence of that, you don't have to call it ascertainability, but it's another issue that's going to have to be considered as part of the class certification decision to see whether you're ever going to be able to identify the people who actually are members of the class. Otherwise, we're going to wind up being back at fail safe classes, which I don't think anybody likes.

ELLIS: There are some classes in which you can define the class using objective words, but you're never really going to know who is in the class. Not for purposes of notice; not for purposes of a distribution; not for subsequent litigation.

Although I don't necessarily agree, I think the court's resolution was sensible. Certainly sensible to view problems of identifying class members as something that should be dealt with under the notion of manageability in a Rule 23 sense, as opposed to some absolute requirement. It does kick the can down the road.

I was concerned that the court basically said, "this is a consumer class action; there are small amounts at stake. The proper question is not, is there going to be a class action versus individual lawsuits, but is there going to be a class action versus no litigation at all?" Looking through that lens, you're going to certify any class. That was not helpful.

WILLIAMS: I'm sure that is concerning for lots of people. I think the court has clarified the meaning of ascertainability. That standard is something that's not in Rule 23; it's not in the federal code; it's something that the courts developed, and that's in federal law.

California law has a presumption in favor of the class device where it's appropriate. And ideas, such as the choice between a class action or no litigation at all, can maybe be attributed to that unique aspect of California law.

What seemed to be the issue in the trial court and the appellate court was that the plaintiff had undertaken a tremendous amount of work to figure out how to identify the class, but that's an issue of notice. There's no requirement that every class member receive direct notice. There never has been. Elevating that is like saying "You can't find every class member, therefore you can't have a class." That shouldn't be the reason that you don't certify a class. It's best notice practical under the circumstances.

The court said the due process concern isn't so huge here because no one can bring an individual case for such a small amount. But the clarity in that this is not some separate requirement is helpful.

ELLIS: I found this troubling. There is a little difference between how federal and state courts deal with this. It is true that courts agree that if you have a properly certified class, and there are some people in that class who you can't find, you can do some other form of notice. You can't do the impossible.

The court in Noel seemed to suggest that if you have the person's name and address, you don't have to give them individualized notice. That is consistent with the California rules of court. Federal courts, including the U.S. Supreme Court, have said, however, that as a matter of due process, best notice practicable means individual notice when it's possible.

It seemed as if the California Supreme Court was either not aware of this potential disconnect between U.S. Supreme Court precedent and what it was saying, or was intentionally suggesting that a different rule applies in state court.

GEIBELSON: Again, there is an administrative feasibility problem, particularly in false advertising and unfair competition law class actions, where you're just never going to be able to identify the individual people. And we talked about it before as potentially having a lump sum determination, but that creates due process problems that are separate and apart from notice. For example, was there any interaction between a salesperson and the customer that vitiates the alleged false advertisement? Without having that person before the court, or providing some declaration that can be the subject of cross examination on an individual basis, there's a whole lot that we're presuming that never actually gets proven with evidence. And without some sort of administratively feasible way to identify who the consumers are and that may be a low threshold if there are records identifying purchasers, in lots of cases where there is not always a written representation that someone is relying upon, it can be hugely problematic.

SHARP: That's valid. I think the 9th Circuit grappled with that issue inBriseno v. ConAgra Foods, 844 F.3d 1121 (2017). That decision -- and the many 2nd Circuit decisions that have said you don't have to have some perfectly ascertainable class -- carries significant weight. Noel is consistent with those.

There's no disagreement that declining to make ascertainability a specific requirement or baking it into Rule 23 kicks the can down the road. But we sometimes lose sight of the facts, too. Class certification is a procedural motion. Since In re Hydrogen Peroxide Antitrust Litigation, 552 F.3d 305 (3d Cir. 2009), there's been a substantial conflating of the merits with class certification. Cases likeIn re Amgen Securities Litigation ratify some foray into the merits, but not too much. Lately, in both the evidentiary context and in the ascertainability context, there has been a pretty significant body of law developing that says the degree of evidence that you have or the degree to which you can establish with perfect granularity who is in the class, isn't necessarily something that needs be decided at class certification. I know that's a bit unsatisfying. But we need to come back to the bedrock principle that the degree of evidence required to clear any given threshold in the lawsuit depends on the stage of the litigation.

Summary judgment standards and trial standards are pretty different than class certification. Ascertainability is fading into the background. I agree that it seems to be trending back towards a manageability construct. But I think it is an opportunity for plaintiffs to say, to the extent that we don't have perfect transparency into every single person or how we're going to identify them, that if we can pick up a whole lot of these class members, and not swamp the efficiencies and the benefits of the class action, that should count.

MODERATOR: Does evidence need to be admissible to be considered in the class certification stage?

ELLIS: Depends on which judge you ask.

SHARP: I think Sali v. Corona Regional Medical Center, 889 F.3d 623, 632 (9th Cir. 2018), is pretty clear. In the 9th and the 8th Circuits, it's pretty clear we don't need evidence to be admissible. In the 7th and the 3rd Circuits, it's a little bit different. But Sali makes the point that there is a sliding scale of the quantum of evidence, and the probativeness of the evidence, and the attendant admissibility question that should be taken based on the stage of the litigation.

To the extent that this raises questions about Daubert motions filed at class certification -- what's emerging is that it's a tactical question about whether to actually move to exclude experts at all. By filing those motions, we're to some degree embracing the notion that there needs to be some quantum of admissibility. And it's just not clear to me that that's the law.

WILLIAMS: In the Northern district, judges are starting to think that Daubert motions are wasteful; they're granted so rarely. It's creating a negative reaction among the bar. The 9th Circuit has spoken in Sali; they've been clear that it doesn't have to be admissible.

I'll go back to General Telephone Company of Southwest v. Falcon, 457 U.S. 147 (1982), which is where the whole rigorous analysis concept came from in the first place. There, the court basically said, "Sometimes, you can just look at the pleadings to decide class certification."

We've certainly seen that. It's a case by case thing. If you have a case where class and merits issues overlap and a judge has to figure them out in order to make a decision, that might be an instance where the judge is more inclined to put more weight into admissible evidence than inadmissible evidence. But I think few judges will impose a strict admissibility standard.

SHARP: There are few instances in which the judge needs to be the finder of fact on class certification. The operative question is whether the proposed methodologies are capable of class wide proof. To the extent there are factual issues to be resolved, they must be left to the finder of fact, and not cause class certification to become a summary judgment motion.

ELLIS: I disagree that 9th Circuit law is clear. I know the opinions that have gone back forth on this issue, but the requirements in the text of Rule 23 need to be satisfied. The judge is the one who is making findings based on materials submitted to the court. Whether it's admissible evidence or not, findings need to be made that each of the requirements for certification is satisfied.

When it comes to expert testimony, I'm puzzled by this notion that the court should consider inadmissible evidence. If it's not admissible to prove anything, why should it be admissible to prove that the requirements of Rule 23 are satisfied?

To be sure, you don't have the same concerns that animated Daubert with regard to confusing juries with junk science, but if the expert's opinion is not going to be admissible at trial, how could a judge make a determination at class certification that a reasonable jury could accept what's in the expert report?

If it is truly an interim decision subject to modification at some point later, it may be more acceptable to some jurists that a lesser quantum of evidence earlier on may be sufficient, knowing that there may be some opportunity later to reexamine the issue.

SHARP: Interesting. On the admissibility point, you're right, it raises some thorny issues. But the question that's being asked may be a little different than the one you posed. I think it's less, "This is clearly inadmissible! How are we going to use this?," but rather, does the court need to make that determination at the time of class certification? We would be insane to have an expert put on a model that's based on a bunch of clearly inadmissible evidence.

It is true that it is incumbent upon the plaintiff in the class certification context to establish that each of the elements of Rule 23 is met. But it's important to distinguish the plaintiff showing by a preponderance of the evidence that the Rule 23 factors are met, from this question of the battle of the experts, which often does result in a circumstance in which the party opposing class certification is saying to the judge, "We don't like their model, and you need to decide which one is better." Courts do not need to engage in that type of fact-finding at the class certification stage.

WILLIAMS: The focus in terms of the judge requiring admissible evidence should be pretty narrow. The question really is whether common issues predominate. Requiring only admissible evidence for that is not what anyone had really intended at the class certification stage. There's going to be further proceedings; there's going to be an opportunity to test that and prove summary judgment at trial. So, to require that, at that point, doesn't seem to be part of what was intended.

ELLIS: Where the case is pretty mature, discovery is done or substantially done, and the expert has worked up a full report, and that's what is submitted for class certification eventually, the judge is going to need to decide, "Is this model admissible?" What's the harm in having that determination made at class certification?

Judges might prefer to kick the can down the road, but why is that proper?

WILLIAMS: Perhaps the requirements that class certification be done at the earliest and most practicable time is not always honored. It does typically come well before you have completed your merits, but not always.

I think that we have argued more and more to apply for class merits -- let's do it all together because we know the response for a motion is going to be in someplace that's merits based, so we don't have to get into it. I'm not sure there's anything wrong with what we proposed there. Because the jury is not, at that time, considering the expert report, and there's no fear of them considering junk science, the judge tends to do a shorter analysis to determine this is really not junk. If it is, I think the judge is going to throw it out. To get hung up on that at that point because of what the subsequent merits report might show doesn't seem to serve the purposes of Rule 23.

SHARP: It feels like we're converting the rigorous analysis into an evidentiary shooting match. It continues to feed into this notion that class certification is a mini trial on the merits. The Sali decision has some interesting language and I'm stealing "shooting match" as it turns out: "Transforming a preliminary stage into an evidentiary shooting match inhibits an early determination of the best manner to conduct the action."

Class certification has some relatively basic threshold criteria. We have an expert report that we have every interest in having pass muster all the way down the road. We don't want to spend $5 million on an economist report only to have it thrown out.

WILLIAMS: In those cases where the plaintiff expert's court certification does have some serious defects, I don't think judges are hesitant to take that on directly.

SHARP: Right.

MODERATOR: How does the California Supreme Court's decision in ZB v. Superior Court, 2019 DJDAR 8825 (Sept. 12, 2019), affect the future of PAGA and class action waivers?

GEIBELSON: This one got decided on the grounds that weren't very helpful. At a minimum, it signaled the court's unwillingness to dig into the issue that everybody was waiting for. It left the claims for wages to go to arbitration, and looked at penalties as being something different. If it signals anything for the courts of appeal, it's a need to engage in close statutory analysis about what will and won't be allowed to continue in arbitration versus in superior court.

It brings to mind the Song Beverly causes of action that had so much notoriety over the last several years because there's a provision in it that says that the penalty will be paid to the person paying with a credit card. Well, in other sorts of penalty cases, the penalties may or may not be specified to be paid to a particular person.

If anything, it refines the scope of what is going to be inside and outside of what is agreed to be arbitrated in some consumer contracts.

WILLIAMS: I agree. I think what we are going to see is a lot of laws going to the courts of appeal on this. And I would anticipate other aspects of this going back to the Supreme Court for further clarity.

ELLIS: For over a decade, there has been some tension between California state courts and federal courts, particularly the U.S. Supreme Court, about the enforceability of arbitration clauses, specifically class action waivers.

I viewed this case as an effort by the California Supreme Court to try to tamp down this conflict. There was clearly an opportunity for the court to expand upon Iskanian v. CLS Transportion, 59 Cal. 4th 348 (2014), and establish, at least under California state law, a broader exception to the enforceability of class action waivers. The court didn't choose that path; it chose a less adversarial one.

It's interesting to me thatNoel, ZB, and Southern California Gas Leak cases, 2019 DJDAR 4671 (Cal., May 30, 2019), which we haven't discussed, are all unanimous decisions. I don't think we would have seen that from the California Supreme Court even three years ago. And it's not just because there's been a turnover of the personnel, although that's obviously part of it. I think these justices are making a conscious effort to try to steer down a more middle path. It will be interesting to see if that continues, but I think it's good development for the courts as a whole.

GEIBELSON: You're really giving a little sugar after the medicine of earlier.

WILLIAMS: You know, it made me think of probably, to me, the most important class action case ever, which was a unanimous decision: Brown v. Board of Education. And it raises the questions made that unlike our state court, is that what will go to our Supreme Court now, assuming it were resolved the same way? I don't think that would be a unanimous decision.

So, yes, I found this fascinating that these decisions were unanimous in the California Supreme Court.

MODERATOR: To wrap up, what about other lingering class action issues we haven't discussed, such as settlements involving cy-près-only relief?

WILLIAMS: Cy près is a big class issue. I thought there was a sentiment on the U.S. Supreme Court to get rid of it, but there were issues in the Facebook case that prevented that. I think that is going to get back to the court soon.

ELLIS: I agree. If the issue gets squarely presented to this court, it's hard to imagine any result other than the court taking a pretty negative view of a settlement if the only relief that the class members get is cy près relief.

There's a distinction between a situation where 90% of the relief goes to the class members, and then there's some residue that goes to a cy près. What the court seems concerned about are class action settlements where the individual class members get nothing, there is some sort of cy près award, and the plaintiff's counsel gets a third of that as an attorney's fee award.

WILLIAMS: Technology is advancing more quickly than the law. That was a data privacy case. There was an injury, but it's very difficult to quantify the injury from an invasion of privacy. And that's the challenge you have. It's an enormous case and an enormous class, but how do you determine if Facebook is going to be required to pay somebody? How do you value it? We're in uncharted waters.

SHARP: Does the lack of obvious injury mean that a class case is just not the right vehicle? And if so, is there a vehicle?

Pure cy près relief cases have gone the way of the dodo. Talk about people overreaching. Anybody who has a pure cy près relief case and insists on taking it all the way up hopefully has somebody who can see the writing on the wall, and suggest that maybe there is a better case to stick a stake in the ground on.

ELLIS: That sounds right.