James R. Copland
Director of Manhattan Institute's Center for Legal Policy
We're proud to have Ted Frank as a Manhattan Institute adjunct fellow and editor of Point of Law, but most of our readers also know that Ted's primary job these days is running the Center for Class Action Fairness, a non-profit entity Ted founded that challenges class action settlements that, in Ted's view, unfairly compensate plaintiffs' counsel at the expense of the class. Scholars at the Manhattan Institute's Center for Legal Policy (CLP) have long worried about abuses of the modern American class action, which have become ubiquitous since Rule 23 of the Federal Rules of Civil Procedure was changed in 1966 to treat all potential class members as class litigants unless they affirmatively opted out of litigation.
In 2002, CLP visiting scholar Richard Epstein, now of NYU law school, articulated the merits and pitfalls of class action practice in a Civil Justice Report and concluded that "we cannot make a uniform assessment of the overall effects of class action practices," since they are "benevolent in some cases and harmful in others." In his 2010 book Lawyer Barons, CLP visiting scholar Lester Brickman, of Cardozo Law School, discussed in depth the degree to which class counsel, operating without a true client, can collude with defendant companies to expropriate unjust fees in class action settlements, in many cases negotiating away plaintiffs' legitimate legal rights.
Like Professor Epstein, Ted is not opposed to all class actions, but he's particularly concerned about the fee abuses Professor Brickman highlights. Other legal scholars, however, have defended current class action practice, including fee awards, as essential to deterring corporate misconduct. Foremost among these academics is Brian Fitzpatrick of Vanderbilt Law School, who has argued that class counsel should receive as much as 100% of awards as fees in small stakes cases. Ted and Brian have been sparring about this issue recently in many live forums, and we are happy to welcome Professor Fitzpatrick to Point of Law to debate the issue here, with our editor.
As Judge Posner once said, "class actions are rife with potential conflicts of interest between class counsel and class members." This comes into play most often at the class action settlement stage.
Defendants want to minimize the amount they pay. Class counsel wants to maximize the amount they receive, but they're also negotiating on behalf of their clients in the same negotiation. It's all too easy for parties at the table, explicitly or tacitly, to freeze out the absent class members—especially when a professional mediator is pushing them to settlement without considering the interests of parties not at the bargaining table.
For this reason, Rule 23(e) and state analogues requires a fairness hearing where the court ensures that the settlement is fairly treating the class. But judges have their own perverse incentives: approving a settlement (often presented to the court ex parte without any well-crafted objections helpful to the court's consideration) is easy, and reduces a court's workload by taking a complicated case off the docket; scrutinizing or rejecting a settlement requires hard work, and adds to the court's workload.
Brian T. Fitzpatrick
I am honored once again to be paired with Mr. Frank for a discussion of our class action system. As he anticipated, I agree with much of what he had to say. Like all humans, the participants in our class action system--class members, class action lawyers, defendants, judges--are self interested. As in all human endeavors, that self interest can be channeled for good or for bad. Which one we get depends on how carefully we design the system.
One of the biggest design concerns with the present system is the one Mr. Frank has spent so much time trying to ameliorate: the near total absence of adversarial testing of class action settlements. Without such testing, the self interest of all involved, as Mr. Frank noted, can lead to socially-detrimental outcomes. Although I do not agree with all of the objections to class action settlements that Mr. Frank has filed--I do not agree, for example, with his comments about the settlement in the Bank of America Overdraft Litigation, as I note below--I do appreciate the important role he serves as a devil's advocate.
Fitzpatrick points to his study showing $5 billion of fees for $33 billion of recovery. But that analysis is flawed in several ways. First, most acknowledge that fees should be smaller for megafund cases, but when you add megafund cases to tiny cases, the statistical effect of the megafund case is to overwhelm the overpayments in the smaller cases. If attorneys collected $4 billion for a $30 billion settlement, that would be too high: it's not 1000 times more difficult to bring a $30 billion case than a $30 million case; meanwhile the other $3 billion from several hundred cases would result in $1 billion of fees, which is also too high. So "only" 15% recovery may well be too high, depending on what the mix of cases looks like.
Second, the study mixes apples and oranges. Securities cases, which make up the larger share of class action settlements, generally have lower percentage fees than consumer-fraud class actions. That's because securities cases are more likely to have sophisticated lead plaintiffs, and better distribution of settlement funds to class members. That ends up supporting my argument more than Fitzpatrick's: securities cases are harder to bring, and are more likely to lose on a motion to dismiss because of higher pleading standards. Yet, with even the minimal constraints provided by the PSLRA, securities attorneys end up getting a much smaller percentage than consumer-class attorneys, showing how much the consumer-class attorneys are getting overpaid. But the securities attorneys are overpaid, too. First, the PSLRA requires fees to be a reasonable percentage of the amount actually paid to the class, but this statutory language is generally ignored by the settling parties and the courts: in the Franklin Templeton Mutual Fund settlement, the attorneys are asking for almost as much money as the amount that will actually be paid to the class, because they include payments to third-parties such as the settlement administrator in their denominator, against the express language of the PSLRA. Second, the PSLRA forbids courts from using the Vaughn Walker method of requiring class counsel to bid for lead-counsel status, but we know from experience that that market constraint results in multiple bids from experienced counsel that are much lower than what class counsel tend to get in securities cases today. So Fitzpatrick's study hides how much attorneys are being overpaid.
Third, Fitzpatrick's study hides how much attorneys are being overpaid in another way, by exaggerating the denominator. That "$33 billion" figure is fictional: it includes "injunctive relief" that doesn't actually benefit the class. The Fitzpatrick study would count the Blessing v. Sirius XM settlement as worth $180 million, when it actually pays zero to the class. (I'm filing a reply brief in the Second Circuit in that case.) And in securities cases, much of the settlement fund is coming out of the pockets of class members who bought-and-held the defendant's shares: those payments from the right-hand pocket to the left-hand pocket are a loss, not a gain, for shareholders. (Such settlements really raise 23(a)(4) questions when they don't bring in new money from third parties.) But the full amount counts in the denominator, even though it didn't win the class anything.
The cases where the lawyers are abusing the system are not an anomaly. When the Center for Class Action Fairness is deciding whether to take a case, it's almost always deciding between cases where the attorneys are abusing the system a little, or whether they're abusing the system a lot.
Brian T. Fitzpatrick
It is true, as Mr. Frank notes, that courts generally award smaller fee percentages in bigger settlements, but, even still, the data do not support the conventional wisdom that the lawyers are making out with everything: the mean and median fee awards in class action settlements are only 25%, and the highest fee percentage awarded in any case over the two years in my study was 47%. Even 47%--which was an outlier by any measure--is far from everything.
Mr. Frank claims that these numbers are misleading, but I think his criticisms miss the mark. He says that my numbers are driven down by securities fraud settlements because courts tend to award lower percentages in those settlements; he suggests that the numbers are "much" higher in other areas. My study did show that percentages in securities cases are lower than the percentages in some of the other subject areas, but the percentages are not "much" different: as Tables 8 and 12 of my study show, the percentages in other areas are only two or three points higher. (Mr. Frank thinks that even this is perverse because he thinks securities fraud class actions are much more difficult to bring in light of their stricter pleading standards. But Mr. Frank forgets that it is much easier to certify a securities fraud class action than it is to certify most other class actions due, among other things, to the fraud-on-the-market presumption of reliance. Overall, I suspect it is actually less risky to bring a securities fraud class action.)
Mr. Frank also claims that my numbers are misleading because they are based on exaggerated denominators: class action lawyers, he says, ask for a percentage of the value of the injunctive relief they win as well as the cash they recover, and the values they place on these injunctions are not real. Class action lawyers may ask for it, but my study did not give it to them: my study included valuations of injunctions in the denominator only when the valuations were by courts rather than lawyers (and this was not very often). All told, only 4% of the $33 billion denominator in my study comes from valuations of non-cash relief. Even if this amount is thrown out, the share taken by class action lawyers barely budges: it is still right around 15%.
(As an initial matter, I just want to say that I wrote my first two posts before I saw the debate framing asking "rife with abuse or an important legal safeguard?" I'd answer "both": class actions are an important legal safeguard and they're rife with abuse.)
When Professor Fitzpatrick says his study didn't find a single class action in two years where the fee percentage was over 47%, I have to question the methodology of the study. I could double the number of lawyers working for CCAF, and we'd still have to turn away class actions where consumers come to us complaining that the attorneys are collecting more than 47%. In the first thirty cases where we filed objections, twenty-six of them involved cases where the attorneys intended to receive more than their putative clients and, like I said, we're confronted with more unfair class actions than we have the opportunity to object to. Sometimes it's much more, in the 90-100% range of total recovery: I have cases on appeal in the Second, Sixth, and Ninth Circuits where judges rubber-stamped settlements where attorneys ended up with over 90% of the recovery.
I suspect the problem is that Professor Fitzpatrick is not collecting accurate data. Most class actions don't report how much class members are actually collecting in settlements; settling parties suffer no consequences when they exaggerate recovery in their papers. Thus, in the first Classmates.com settlement, it was widely reported in the press that the class would receive $9.5 million; the actual number would have been $0.1 million if the court had not honored our objection. In the pending Brazil v. Dell, the parties reported in preliminary settlement papers that the settlement made $18 million available to the class, justifying a $7 million payout to the attorneys; in fact, enough barriers were placed in the claims-made settlement that the class will receive only $0.5 million, a figure that never would have been made public if we hadn't objected. In McDonough v. Toys R Us, the number that the class is to receive still isn't public after final judgment, other than that we can tell mathematically from what little has been disclosed that it will almost certainly be less than half of what the attorneys are getting. Settling parties suffer no consequences for exaggerating settlement value to courts, and those exaggerated values end up in Fitzpatrick's study's denominators; the real values remain known only to settlement administrators (who will not disclose them if asked) and rarely end up public. Professor Lester Brickman writes about this, too.
Returning to the Bank of America overdraft case, we see right away the difference between reality and study denominators. Professor Fitzpatrick characterizes the settlement as for $410 million; in fact, terms of the settlement call for as much as $60 million of this amount to go to third parties, rather than class members. The $123 million award to the attorneys is not only more than twice the typical percentage for a settlement larger than $150 million (a fact that the district court fails to acknowledge in its opinion), but works out to thousands of dollars an hour for a case that was barely litigated and immediately settled for nine cents on the dollar—and where nearly all of the legal research has application in numerous other cookie-cutter cases. (The main risk the attorneys identify was that the majority of their claims had been waived by previous class action attorneys who had accepted an $8 million payoff.) I have no objection to attorneys getting a multiple of lodestar when they take on risk, but settling a multi-billion dollar lawsuit for pennies on the dollar is the essence of a relatively riskless proposition. (The fact that the attorneys were not even willing to admit to a lodestar amount to demonstrate the fairness of the award in a cross-check is a fairly strong negative pregnant that the lodestar multiplier was in the double-digits.) That's not rewarding attorneys for success, it's rewarding them for going after big defendants. To add insult to injury, the MDL court signed off on the parties' agreement to make it nearly impossible for class members to object to this ripoff, demonstrating class counsel's fear of scrutiny—which is ironic, given their propensity for objecting to other overdraft settlements (and accepting a buyout in the Trombley v. National Bank case, where the victorious class attorneys collected a 28% fee after settling on Docket Entry #6, with class counsel being compensated for paralegals' work at about $3000/hour).
Because the Overdraft MDL was not subjected to a market test by, say, putting out the lucrative litigation for bid, there was a wealth transfer of tens of millions of dollars from consumers below median incomes to very wealthy trial lawyers, who would have been excited to litigate the case for, say, three times their lodestar, even though that would be a fraction of what they actually received. And by all accounts, the settlement would have been even higher if previous class counsel hadn't walked away with $8 million by agreeing to an even more unfair class action settlement. It's hard to see how that's good for anyone.
Brian T. Fitzpatrick
Mr. Frank criticized the methodology in my study because it is based on settlement amounts approved by district courts rather than settlement amounts actually distributed to class members. He's right about that, but, again, it turns out not to make much of a difference to the portion that attorney's take from settlements. The vast majority of the money approved by courts is distributed pro rata--meaning it is all distributed, and how much each class member gets depends on how many others submit claims. Thus, even if we were to ask what portion of distributed money goes to class action lawyers, the answer would be about the same.
This is not to say that there aren't isolated examples where the only ones defendants end up paying are the lawyers. But it is to say that these cases are not representative. It is almost unheard of for undistributed settlements to revert back to defendants these days. If they cannot be distributed to class members, they at least go to third parties like charities; either way, the deterrence gained is the same.
Mr. Frank returned to the Bank of America settlement, but, again, I do not understand why. In one breath, he says the lawyers there did not take on any risk, but in the next he acknowledges that Bank of America had already settled the same claims for a fraction of a cent on the dollar. The fact that the new lawyers managed to persuade Back of America to resettle the case for over ten times the original amount is not just good lawyering, it is remarkable lawyering. They deserved to be paid handsomely. So what if they made a multiple of their hourly rate? The lodestar method fell into disfavor in class action litigation previously because it rewarded lawyers for dragging things out rather than getting results. The percentage method rewards results, and remarkable results should be rewarded with remarkable fees.
I will close on one point on which Mr. Frank and I agree. The optimal method for awarding fees is not to do so ex post by trying to divide some fair percentage of the settlement. It is to do so ex ante by auction. But until courts warm up to that, we will continue with the second best. And that is the percentage method, not the lodestar
FEATURED DISCUSSION ARCHIVE:
Obamacare Decision: Reactions, July 2012
Law School Faculty Diversity, May-June 2012
Class Actions, May 2012
Constitutionality of Individual Mandate, March 2012
Human Rights and International Law, February-March 2012
The constitutionality of President Obama's recess appointments, January 2012
Do caps on medical malpractice damages hurt consumers?, December 2011
Trial Lawyers Inc.: State Attorneys General, October 2011
Wal-Mart v. Dukes, April 2011
Kagan Supreme Court nomination, May-June 2010
Election roundtable, November-December 2006
Who's the boss, September 2006
Medical judgement, July 2006
Lawyer Licensing, May 2006
Contingent claims, April 2006
Smoking guns, July 2004
Center for Legal Policy at the