Printer Friendly

Auctioning class settlements.

Abstract

Although class actions promise better deterrence at a lower cost, they are infected with problems that can keep them from delivering on this promise. One of these problems occurs when the agents for the class--the class representative and class counsel--advance their own interests at the expense of the class. Controlling agency cost, which often manifests itself at the time of settlement, has been the impetus behind a number of class action reform proposals.

This Article develops a proposal that, in conjunction with reforms in fee structure and opt-out rights, controls agency costs at the time of settlement. The idea is to allow the court, once a settlement has been achieved, to put the class's claims up for auction, with the settlement acting as a reserve price. An entity that outbids the settlement becomes owner of the class's claims and may continue to pursue the case against the defendant. A successful auction results in more compensation for the class. On the other hand, if no bids are received, the court has evidence that the settlement was fair. The prospect of a settlement auction also deters class counsel and the defendant from negotiating a sweetheart deal that sells out the class.

The Article works through a series of theoretical and practical issues in settlement auctions, including the standards that a court should use to evaluate bids, the limitations on who may bid, and ways to encourage the emergence of an auction market.

TABLE OF CONTENTS

Introduction
I. Potential and Pitfalls in Class Litigation
II. Ensuring Fair Settlements
III. Practical Aspects of Settlement Auctions
     A. The Auction's Purpose
     B. Bidding
        1. Standards for Bids
        2. Partial Bids
        3. Settlements with Injunctive Relief
        4. Making Payment
     C. Bidders
     D. The Emergence of an Auction Market
     E. The Timing of Opt-Outs
     F. Apportioning Fees
Conclusion


Introduction

The cases and literature on class actions contain some radical ideas. Among them are auctioning the class's claims to the highest bidder; (1) auctioning the position of class counsel to the lowest responsible bidder; (2) employing third-party financiers, who take a share of any class settlement in return for providing funds to prosecute the case; (3) and handing over the victims' small-stakes claims to a charity or advocacy group. (4)

All of these ideas have two premises at their foundation. First, class actions are valuable because they can achieve significant deterrence, especially in small-value cases for which individual litigation is not an option. (5) Second, class actions are dangerous because individual plaintiffs have either limited capacity or an insufficient financial stake to monitor the work of their agents--the class representatives and class counsel. (6) Without monitoring, representatives or counsel may engage in self-dealing, reaping a large recovery or fee for themselves while settling the claims of class members for pennies on the dollar. (7)

The trick is to harness the deterrence potential of class actions while slashing agency costs to a minimum. Auctioning the position of class counsel seeks to control agency costs directly by subjecting attorneys' fees to market competition. (8) The other three proposals--auctioning the class's claims, using third-party financing, and giving the right to recover on the class's claims to a charity or advocacy group--control agency costs indirectly by steering the claims into the hands of a single person (a bidder, a financier, or a charity) with a sufficient stake to monitor counsel effectively. (9)

This Article suggests a simple strategy to control agency costs that blends aspects of all four approaches. When the parties in a class action arrive at a settlement, the court should put the settlement up for auction. If third parties bid more for the case than the settlement offer, the proceeds of the highest bid are distributed to the class. Ownership of the class's claims shifts to the winning bidder, who has an incentive to monitor counsel while continuing to press the case against the defendant.

Although the proposal is easy to state, it poses practical difficulties. Part I begins the examination of these issues by framing the problems that a settlement auction is designed to alleviate. Part II outlines the auction process and describes the conditions under which an auction can help to reduce agency costs and to ensure fair settlements. Part III addresses a series of practical concerns that settlement auctions raise and demonstrates how an auction process responsive to these concerns helps class actions to deliver more compensation and provide more deterrence.

I. Potential and Pitfalls in Class Litigation

Aggregating similar claims in a single lawsuit can create significant benefits. Principal among them are forcing defendants to internalize the full costs of their wrongdoing, (10) equalizing the incentives for plaintiffs and defendants to invest in litigation, (11) lowering per-claim transaction costs, (12) providing compensation to victims, (13) and ensuring the like treatment of like claimants. (14) But aggregation also creates the risk of substantial costs. Among these are aggregation's potential to overdeter behavior, (15) deprive individuals of their right to a "day in court," (16) and increase transaction costs.

The most significant transaction costs derive from three sources. The first transaction costs are the additional costs of litigation. In cases in which individual losses are very small, which are often referred to as "negative-value cases," (17) the parties and the court incur no litigation expenses, absent the plaintiffs' capacity to aggregate their claims, because the cases are too expensive to bring individually. (18) In addition, class actions and other aggregate litigation often involve other expenses--such as notice to the class (19) and additional judicial management (20)--that ordinary cases do not. The second set of transaction costs are error costs. The aggregation of claims may distort both the likelihood of recovery and the amount of recovery in relation to individual litigation, thus generating a litigation premium that defendants must pay. (21)

Third, aggregate litigation may generate agency costs, which can arise when an asset owned by a principal is placed in the hands of an agent. The agent may have an incentive to maximize personal profit rather than the profit of the principal. (22) The consequence is agency cost, which has three components: the principal's costs of monitoring the agent, the principal's costs of providing incentives that ensure the fidelity of the agent, and any difference between the asset's value in the hands of a faithful agent and the actual value realized by a self-interested agent. (23)

In aggregate litigation, little monitoring occurs, (24) so the most significant agency costs arise from loss in claim value due to self-serving representation. Loss in claim value arises from two sources. First, the class representatives, or other individuals that lead the group, may use the group's claims to leverage more favorable treatment for their own claims relative to the claims of others in the group. (25) Second, class counsel may sell out the entire group's claims in return for a hefty fee. (26) Agency costs do not occur in every case, but the inevitability of conflicts among group members and between the group and counsel makes these costs a constant risk. (27)

As litigation, error, and agency costs rise in relation to benefits, aggregation by class action or other means becomes less attractive. (28) Therefore, considerable attention has been paid to keeping these costs at their minimum. With regard to agency costs, Rule 23--the federal class action rule--created structural safeguards that can reduce these costs. (29) First, a court must find that both the class representative(s) and class counsel are adequate. (30) Second, a court must approve all class settlements and can do so only when it finds that the settlement is "fair, reasonable, and adequate." (31)

But these safeguards do not end the concern about agency costs. At a legal level, the present law on "adequate representation" is too broad in some ways and too narrow in others to sufficiently ferret out all self-dealing behavior by representatives and counsel. (32) Nor do these safeguards presently pertain to nonclass aggregate litigation. (33) At a practical level, these protections have not eliminated self-regarding behavior. There are enough examples of such behavior--Dexter Kamilewicz's settlement; (34) settlements in which class members received nearly worthless coupons and class counsel walked away with large fees; (35) settlements in which class counsel's own clients received greater benefits than those afforded class members under the settlement; (36) and other settlements in which class members stood to receive nothing of value, but class counsel stood to profit handsomely (37)--to dispel any notion that the doctrine of adequate representation eliminated self-dealing. Likewise, courts' lack of knowledge about the strength of the claims, as well as many courts' desire to move big cases off their dockets, renders the "fair, reasonable, and adequate" check on settlements an imperfect mechanism to avoid agency costs. (38)

A number of proposals designed to cut down on agency costs have stepped into this gap. (39) To start, there are the ideas discussed at the beginning of this Article. (40) In addition, some commentators have suggested creative uses of attorneys' fees to give counsel less incentive to sell out the class. (41) Others have recommended that members of the group be allowed a free right to opt out of a settlement once its terms become known. (42) Broadly, these proposals seek to control agency costs in one of three ways. One is to transfer ownership of, or at least a significant stake in, the claims to an entity with a sufficient financial interest to monitor the counsel's performance. (43) The second is to use fee or opt-out incentives to keep counsel from wandering too far from the pursuit of the class's interests. (44) The third, which is related to the second, is to cap fees by opening up the counsel position to competitive bidding. (45)

There are difficulties with all three approaches. The first does not eliminate agency costs because monitoring expenses still exist. Even more important, the auction, third-party-financier, and charitable endowment approaches all occur at, or near, the outset of the case when the value of the victims' claims is often difficult to determine. (46) As a result, bidders or financiers will expect a significant discount on the true value of the claims to proceed--thus thwarting the delivery of a realistic level of compensation to victims. (47) And public-interest groups may be reluctant to take up the sword against well-armed defendants, assuming that litigation is even a part of the modus operandi of such groups. The second approach requires use of the right fee or opt-out structure to align the interests of class and counsel. But even a proper structure cannot account for all agency costs associated with self-regarding legal representation, (48) nor will providing opt-out rights ensure their exercise when, and only when, they avoid agency costs. (49) The third approach may limit the size of the fee, but that does not eliminate the problem of a misalignment of incentives between class and counsel. Indeed, the third approach may enhance the risk of misalignment if the fee is set so low that the attorney must cut a quick deal in order to make any money on the case.

No perfect solution to the problem of agency costs exists. As a general matter, agency costs arise at the time of, or due to the prospect of, settlement. The differing interests of class and counsel in the outcome of the case are most likely to emerge when a deal is--or is likely to be--on the table. At this point, the fear of selling out some or all members of the class becomes realistic. Because most class actions settle, limiting agency costs at the time of settlement will help to control the problem of sellouts.

II. Ensuring Fair Settlements

The best way to control the risk of diverging interests at the time of settlement is to use a combination of all three of the basic techniques to harness agency costs: establish a proper fee structure, put the claims up for auction, and provide limited opt-out rights.

Establishing the right fee structure is critical to aligning the incentives of counsel with those of the client. The potential for agency costs exists unless counsel has an incentive to work exactly as much as the clients want her to work. In other words, this is when the point at which the clients would have to pay more in fees than the value of their claims matches the point at which counsel's expected cost of devoting another hour to the case is greater than the expected benefit that she receives from the fee. Thus, at this matching point, the clients and counsel both want to cease work. The only method for matching up these points is to use a contingency fee that compensates counsel through a combination of an hourly market rate plus a percentage of the net recovery attained for the class. (50)

But this fee structure alone is insufficient to ensure an alignment of interests. It does not prevent the problem of the "reverse auction" or the "race to the bottom," which can arise when two overlapping class actions are filed. Counsel in each class action has an incentive to underbid the settlement offer of counsel in the other class action, lest she receive no fees for the efforts in her own case. The process of underbidding may continue until the defendant achieves a rock-bottom settlement. (51) Some other mechanism(s) must exist to dissuade counsel from engaging in this practice.

An auction of the class's claims provides such a mechanism. The idea of auctioning the class's claims is not itself new. (52) As originally described in the literature, an auction of a class's claims would be held at the outset of the case. The important tweak that this Article suggests is conducting the auction at the time that the settlement is announced. (53) Anyone willing to bid more for the claims than the settlement amount becomes owner of the class's claims. The class members then receive any excess over the original settlement in proportion to the awards that they would have received under the original settlement or otherwise according to the terms of the bid. (54)

A settlement auction significantly reduces the prospect of a reverse auction. In the event that another counsel tries to undercut the terms of the original settlement by proposing to settle for less, the original counsel's response is not to give an even lower price. The original counsel's response is to bid the settlement back up to the original deal, take ownership of the claims, and continue the case against the defendant. (55) The defendant has little incentive to play the reverse auction game.

The third and final element that assures a fair settlement is the provision that allows class members to opt out of the settlement once its terms and payouts to class members are known. (56) Although the ability to opt out may not be available as a matter of right in all class actions, (57) a court should nonetheless extend the privilege of opting out to all class members who can demonstrate that the marginal expected net loss to the value of the class action from their departure will be offset by the marginal expected net gain as a result of their ability to proceed alone. (58) This final check provides a fine-grained ability to ensure that the settlement is in the best interests of the class and its individual members. (59)

Therefore, the auction process is just one component in a balanced set of measures designed to ensure that class members receive appropriate value for their claims. On its own, the settlement auction advances the goals of increasing aggregate deterrence and lowering agency costs, but the proper fee structure and opt-out rights greatly strengthen its capacity to achieve these goals.

For a settlement auction to work effectively, a number of practical matters described in Part III must be resolved. For now, let me put aside these practical matters and make the basic case for an auction as one component of ensuring a fair class settlement. (60) As I have mentioned, a settlement auction can reduce agency costs by reducing the risk of a reverse auction. (61) But preventing reverse auctions is only one benefit. Another benefit is the settlement auction's ability to move a resource--the class's legal claims--to the person who values it most highly. (62) Third, and perhaps of greatest significance, the settlement auction deters sell-outs from the class and other settlements in which counsel fails to achieve the best realistic outcome for the class due to self-interest in a large fee. Faced with the prospect of being outbid by others, and thus foreclosed from realizing a full fee for the work put into the case, (63) counsel has little incentive to agree to a sweetheart deal. The same is true of the defendant, who cannot be assured of escaping the litigation unless it pays fair value for the class's claims. The emergence of settlement auctions should induce class counsel and the defendant to settle claims fairly in the first instance.

In addition, the auction process provides some assurance to the court that the case is settling at the market rate. The court often has little information whether the settlement is in fact "fair, reasonable, and adequate." (64) At the time of settlement, class counsel and the defendant are aligned in arguing that the settlement is a good deal. Therefore, courts must frequently rely on objectors to create the necessary counterarguments and to point out the settlement's defects. But objectors are not always forthcoming, (65) and their motives are not always pure. (66) Moreover, the objection process still leaves the court with a binary choice: accept or reject the settlement. Unless the settlement is truly awful, the momentum is toward acceptance. (67) A bidding process provides a third alternative that can enhance the class's recovery. On the other hand, if no one outbids the settlement, the court has external evidence that the settlement is appropriately valued. (68)

III. Practical Aspects of Settlement Auctions

Although easy to imagine, a settlement auction raises implementation issues. This Part considers some of the most significant issues, in particular those issues that affect the auction's ability to achieve its goals of increasing deterrence, enhancing compensation, and reducing agency costs. In the process, this Part further develops the arguments for a settlement auction.

A. The Auction's Purpose

An important first issue is to determine the goal of the auction: to enhance social welfare or to maximize the return to the "sellers" (the class). As an initial matter, the goal of the auction appears to be maximizing the return to the victims. But this orientation can result in a loss of social welfare. If bidders overestimate the value of the case, which the original settlement had fairly priced, continuing the litigation is not in society's interests, even though the class stands to benefit. (69)

It is a fair point that settlement auctions should be used only if they enhance social welfare; otherwise, it is best to let the litigation end with the original settlement. For two reasons, this objection is not crippling to settlement auctions. First, the auction process's ultimate goal is to improve the quality of settlements in class litigation. Even if a few cases fail to benefit society, those losses must be balanced against the gains that accrue in other cases in which an auction yields social gains or in which the in terrorem effect of an auction leads to a better settlement than counsel might have otherwise negotiated.

Second, settlement auctions diminish social welfare if bidders systematically overestimate the value of class settlements. Settlement auctions can also cause a net loss in social welfare when the defendant's and the successful bidder's ongoing litigation expenses turn out to be greater than the successful bidder's gains from continuing to press the case. (70) As this last sentence suggests, there are two situations in which a settlement auction can be socially harmful: (1) when the bidder pays more than the claims are worth; and (2) when the bidder pays less than the claims are worth but the continued cost of litigation is greater than the gain to the bidder. The second situation can further be broken down into two scenarios: (a) when the successful bidder's attorneys' fees and other costs are greater than the difference between the amount the successful bidder realized and the amount of the bid; and (b) when the fees and costs the bidder paid are less than this differential, but this differential is less than what the defendant expended in attorneys' fees and costs to continue to defend the case. (71) In the first situation and scenario (a) of the second situation, the successful bidder has lost money; in scenario (b) of the second situation, the bidder has made money, but the defendant's expenditures make the lawsuit a losing proposition from society's vantage point.

Bidders who make bad deals are not a problem in the long run. They will either exit the market or adjust their bidding strategies. In the end, the market will take care of itself. But bid winners who make a good deal from their private point of view, and then cause socially wasteful litigation, present another type of problem; there is no market corrective to stop their behavior. (72) There are, however, two steps that can reduce the likelihood of this behavior.

First, the court should consider the possibility of allowing the defendant to meet the terms of the winning bid. (73) If the defendant matches the bid, no additional litigation costs arise; the class's interest in greater compensation and society's interest in greater deterrence match up.

Second, the court may require that bids exceed the settlement by an amount that, in the court's estimation, includes a successful bidder's and the defendant's fees and costs in continued litigation. For instance, if the case settles for $1 million, and the court believes that continued litigation will generate a combined $100,000 in fees and costs for both parties, it might set the minimum bid at $1.1 million. The court's estimate need not be exact, but the minimum bid should be set with an eye toward the likely future expenditures on both sides in order to make sure that continued litigation is indeed in society's best interest--in addition to advancing the interests of the class.

These two steps balance the interests of the class and society. The second step in particular may discourage some bids that would increase payments to class members, but society also has an interest in avoiding wasteful litigation.

B. Bidding

The next issue is to decide how bidding will be conducted. Among the various types of auctions, four are most common: the sealed-bid auction; the English auction, in which bidders compete by bidding up the last amount offered; the Dutch auction, in which the price is lowered until someone accepts the offer; and the second-price (or Vickrey) auction, in which a sealed-bid auction is used, but the highest bidder pays the amount of the second-highest bid. (74) Each has benefits and drawbacks. (75)

The likeliest form for a settlement auction is the sealed-bid method. Given that the reserve price--the original settlement plus any minimum additional amount that the court specifies--is known, a Dutch auction is less attractive. (76) The second-price auction, which is less common, yields less recovery to the class. (77) The real choice lies between the two most common forms: the sealed-bid and English auctions. A sealed-bid approach is generally more realistic. The number of bidders is not likely to be high, thus limiting the effectiveness of the English approach in driving up bids. Moreover, a class settlement is not a tangible object whose value is measurable simply by price. Different bids may contain different elements: some bids may propose unequal additional compensation for members of the class, while other bids may propose to spread the additional compensation equally across the class. The "apples and oranges" nature of the bids may make it difficult for the auctioneer--the court--to compare the value of bids in the heat of an English auction. Finally, sealed-bid auctions have two important advantages in the settlement auction context. First, they tend to induce risk-averse bidders to offer a price closer to their actual valuation of the claims, enhancing revenue for the class; and second, they limit the possibility of collusive behavior to suppress the bid price. (78)

Thus, sealed bidding should be the presumptive starting point for settlement auctions. In some circumstances, the court might find an English auction a more useful way to obtain greater value for the class. (79) In theory, the choice should not matter. Because the auction is conducted with a reserve price, either a sealed-bid or an English auction should maximize the value of the settlement to the class. (80)

One of the standard difficulties in an auction--which prior proposals for auctioning class actions did not address--is valuing the item up for bid. (81) If bids are to be encouraged, some type of pre-auction process to determine the value of the class's claim's seems essential. For this reason, those who have advocated auctioning class claims have also recommended a judicial process to investigate the merits of the claims with the results of the process made available to all potential bidders. (82) Of course, such a process raises questions, such as where the money for the investigation will come from and whether a judicial adjunct can effectively investigate claims within the confines of the American adversarial system.

A great advantage of the settlement auction is that, in most cases, class counsel has already conducted the pre-auction discovery. As part of the auction process, a court should make this discovery available to all prospective bidders. (83) Of course, in settlement class actions in which the lawsuit seeks approval of a prenegotiated settlement, (84) no discovery on the merits of the case may have occurred. But settlement class actions are almost always negotiated against a backdrop of similar lawsuits that have generated evidence and arguments on the merits. If necessary, a court may order class or defense counsel to provide such information to interested bidders.

If the court receives one or more bids, evaluating bids and selecting the winning bid raises a number of concerns.

1. Standards for Bids

As a rule, courts should accept a bid only when it meets three criteria: (1) the bid nets more value for the class as a whole than the settlement offer; (2) no class members end up worse off than they were under the original settlement, unless the bidder can show that the original settlement gave excessive payments to some class members as a means of purchasing their acquiescence in an inadequate settlement; and (3) like class members are generally treated the same under the terms of the bid.

The three criteria are necessary to make the auction worthwhile. The first criterion is the bedrock. Unless an auction makes the class better off as a whole, the reason for holding an auction collapses. This criterion is already implicit in setting a reserve price, but I single it out here for two reasons.

First, a bid may propose somewhat different terms than the original settlement. If the original settlement proposed to settle the class's claims for $100 and the highest bid proposed to pay class members $125, the superiority of the bid to the original settlement is evident. But now suppose that the original settlement proposed to pay every class member $100, while a bid proposed $125 payments in installments over five years. The bid may seem a better deal, but whether it is in fact better depends on the time value of money. A court must ensure that, overall, the bid achieves greater value for the class; otherwise, the deterrence value of class actions declines, and defendants have a perverse incentive to use surrogate bidders to undercut the original settlement with a glitzy-sounding, yet ultimately less valuable, bid.

Second, the first criterion acts as a reminder to the court to ensure that the bidders are responsible--in other words, that they can make good on the bid. (85) The issue is especially acute in cases in which the bid requires performance over time, because the highest bidder's solvency and capacity to perform in the future must be evaluated.

The second and third criteria protect the individuals within the class. Even if the highest bid is more favorable than the settlement to the class as a whole, the bid might make individuals within the class worse off, or it might create differentials among similarly situated class members. Such a bid puts the court in an impossible spot. Should it award a bid that is admittedly better for all on the backs of some class members who stand to lose? Should it allow differential awards among like claims, especially if the differential is created for the purpose of tamping down opposition among certain class members? Such inequities may also engender ill will among some class members--thus creating the risk of opt-outs, with all the inefficiencies that multiple lawsuits generate. (86)

In essence, the three criteria require that the bid be Pareto superior to the original settlement: no one is made worse off, and at least some people are made better off. (87) But the criteria also limit the way in which Pareto superiority is achieved. As a rule, like claimants must be treated alike; the gains cannot be dropped arbitrarily on just a few of the class members while remaining class members stand in place. (88)

2. Partial Bids

An intriguing possibility is the ability to bid for only a portion of the claims. Assume that each class member has two claims (one federal, one state), and the claims fall into one of three categories: low value, medium value, and high value. The defendant agrees to settle both claims for all class members, paying $25 to the low-value claimants, $300 to the medium-value claimants, and $1,000 to the high-value claimants. A bidder may think that the payment for some of these claims is fair and may wish to pursue only some of the claims or the claims of only some of the class members. The issue is whether the bidder may do so, thus leaving the remaining claims and claimants to be compensated by the defendant, or whether it must bid on the entire settlement.

With respect to claims, the clear answer is that a bidder cannot bid on only some of the class members' claims. If the rule were otherwise, enormous difficulties in valuation would arise. Presumably the defendant's settlement offer already accounted for the variability in the strength of the claims. If a bidder could strip out the most valuable claims, then the value of the original settlement would need to be adjusted; expecting a defendant to pay the same amount in settlement when the most valuable claims are not settled would be unfair. But that adjustment would lead to new rounds of negotiation or would require the court to approximate the contribution of each legal theory to the overall value of the settlement--a task that would likely demand an unappealing trial on the merits. (89)

Whether a bidder can propose to buy the claims of only some claimants is a harder matter. In some settlements, the same valuation issues can arise. If the settlement proposes to settle the class's claims for $5 million in total, with no indication of how the settlement funds are to be allocated, (90) then the problem of determining the allocation of funds among claimants is functionally identical to the problem of allocating funds among claims that the prior paragraph described. At some point, however, an allocation plan

must be developed, and if bidding is postponed until the allocation of funds among claimants is known, a court may sidestep this problem. But bidding on the claims of only a subset of class members raises other issues. A defendant's ultimate concern in settlement is the total price tag; (91) the distribution of the award among class members is a secondary matter. (92) Thus, a defendant may have been willing to settle some claims for too much and others for too little. If a bidder can cherry-pick the less well-compensated claims, the defendant might end up overpaying for the settlement. Finally, it is possible to imagine that some distributional difficulties and inefficiencies might arise in setting up a process that pays some claims that the defendant settled and other claims that the bidder settled. And this would create perceptions of unfairness between those who received higher payments through bidding and those who did not.

On the other hand, strong arguments support the use of partial bids. Because the bid covers only part of the class's claims, the bid price will be lower and more participants might be encouraged to enter the auction. The bid can also be targeted right at the part of the case where the problems of agency cost and insufficient deterrence exist. In addition, if a bidder were required to bid on the entire case, the bidder would be tempted to provide more money to all the claimants in order to avoid ill will, objections, and opt-outs. If some of those claims were already fairly priced in the original settlement, requiring bidding to encompass the entire case provides a windfall to those claimants.

Given the strong arguments on both sides, a categorical rule seems unwise. When the parties announce a settlement, the court should examine the nature of the case and settlement to determine if allowing bidding on just a portion of the case is a potentially useful means to encourage bidding and enhance the value of the class's claims.

3. Settlements with Injunctive Relief

Some aggregate settlements involve either exclusively injunctive relief (93) or a mixture of injunctive and monetary relief. (94) These cases present a particular challenge for the auction method. A bidder cannot command a defendant to provide the injunctive relief to which the defendant consented in the settlement; (95) thus, an auction may cost class members' relief that they value. Moreover, allowing bids on just the monetary portion of the claims ignores the fact that the defendant might not have agreed to as generous an injunction if the defendant was also liable for greater damages. The hybrid of injunctive and monetary claims acts a great deal like the situation involving two claims. In that context, I have argued, auctioning one of the claims while leaving the other in place is impermissible. (96) On the other hand, the potential to game the auction system is evident. Unless bidding on claims involving injunctive relief is permitted, class counsel and the defendant can make a settlement auction-proof by adding some nontrivial injunctive relief in the settlement. (97)

Except in the rare case in which a successful bidder is in position to provide identical injunctive relief, (98) the auction method is poorly suited to cases in which injunctive relief forms a valuable part of the class settlement. At the same time, a court should not allow illusory injunctive relief--perhaps an agreement to obey the law in the future or a modest restructuring of senior management--to immunize a substandard settlement. The difficulty lies in distinguishing these two situations.

Establishing hard-and-fast rules to govern this situation may be impossible. If a court is convinced that injunctive relief is central to the goals of the lawsuit and will provide real value to all class members in the future, it should not permit an auction. One intermediate solution is to separate members whose harm from the defendant's conduct could yet be avoided, for whom an injunction could be valuable, from members who have already incurred virtually all of the harm that could befall them, for whom an injunction is valueless. Akin to its power to allow bids only on some of the class members' claims, (99) a court may invite bids only on the claims of class members who fit into the latter group.

4. Making Payment

An interesting practical issue not found in ordinary auctions involves payment to the class. For simplicity's sake, assume that the original case settled for $1 million, and a responsible winning bid offered $1.5 million. Must the bidder pay the entire $1.5 million--or should the defendant pay the class the agreed $1 million and the bidder the additional $500,000? If the latter, may the defendant recoup the $1 million from the bidder in the event that the case is tried and the defendant is found not liable? (100) And if recoupment is possible, must the successful bidder post a $1 million bond to ensure its ability to repay the defendant?

Requiring a defendant to pay its agreed upon $1 million lowers the effective cost of bidding; prospective bidders need to come up with only $500,000 plus the cost of a bond. As a result, this approach could increase the number of potential bidders. The presence of more bidders generally drives up the auction price, likely enhancing revenue for the class. (101) But there are some thorny problems with split payment: the administrative costs of splitting payment; the cost of posting a bond; the dynamics that arise as a successful bidder tries to achieve a verdict of at least $1.5 million while the defendant tries to drive the value of the case to something less than $1 million; and the defendant's likely response to split payment, which is to offer less by way of settlement in the first instance.

On balance, it makes more sense for a successful bidder to pay the full amount of the bid before allowing the bidder to pursue the class's claims against the defendant.

C. Bidders

The question of who may bid seems to have a simple answer: anyone with the resources to make good on the bid. And that answer is partially correct. As part of the inquiry into a bidder's responsibility, a court should reject a bid from any entity that fails to show the necessary financial wherewithal.

But bidding by some entities poses problems. To explain, consider the five types of potential bidders at an auction. The first are law firms, which would use their internal resources to continue the litigation. The second are private equity firms, which are entities that might otherwise engage in third-party financing of claims. The third are consumer advocacy groups or other nonprofit groups with an interest in regulating the defendant's conduct. The fourth, at least in theory, are the defendant's competitors. The fifth is the defendant itself.

Working backwards, we can reject the defendant as bidder. As I described, under certain conditions, the court may allow a defendant to match the highest bid (on condition that the defendant compensates the winning bidder for its costs investigating and analyzing the claim). (102) If bid matching is allowed, the defendant need not participate in the bidding process. Indeed, a defendant's participation in the auction has the likely effect of driving out potential bidders and driving down the price of the winning bid. (103) The defendant had a chance to "bid" when it agreed to the settlement.

Allowing the defendant to submit another bid frustrates an essential purpose of a settlement auction: to force class counsel and the defendant to make a better settlement in the first instance.

The defendant's competitors are another source of bidders. Competitors may have a legitimate reason to pursue a claim against the defendant. For example, the defendant may be engaged in slipshod practices that are negatively affecting the industry. (104) Or the purpose of the suit may be nefarious: a better financed competitor may see an opportunity to drive up the defendant's costs through prolonged litigation, or even to bankrupt the company. (105) In theory, abuse of process and antitrust laws prevent litigation of the second type, (106) but proving an illegitimate motive is difficult and would consume significant resources. (107) Therefore, the prudent position is not to permit competitors to bid at a settlement auction. (108) Consumer advocacy and other nonprofit groups provide a third source of bidders. Giving a class's negative-value claims directly to such organizations makes some economic sense. (109) Some consumer advocacy groups already appear on behalf of objectors to class settlements. (110) Bidding provides these organizations with a different avenue for attempting to hold defendants accountable. Moreover, to the extent that the groups argue that class members deserve better compensation, bidding gives them a direct avenue to accomplish that objective. A number of these organizations have attorneys on staff, so they can keep the costs of the lawsuit against the defendant to a minimum. The major hurdle for such groups is funding the settlement, should they be the winning bidder.

Private equity funds or other organizations that presently engage in third-party funding might also be bidders. These entities possess substantial capacity to analyze the risks of litigation and assess the likely value of claims, so they already possess the infrastructure to participate in settlement auctions. (111) Moreover, repeat players like third-party financiers can afford to be risk neutral in their investment decisions while class members and counsel are more likely to be risk averse. This difference in the way that they value the class's claims suggests that there may be room for financiers to make a profit after winning an auction. Of course, the financial commitment required to fund the costs of litigation, which is the present focus of third-party financiers, is less than the commitment required to fund a winning bid; maintaining lawsuits in their own names represents a shift in the present business model of such financiers. (112)

Finally, other law firms may be interested bidders. In some cases, the law firms may already represent the same class members in competing class actions. (113) Some of these firms may have contended for, but lost, the class counsel position to the firm that settled the claim. (114) Often these firms have participated in the litigation or in similar litigation, so they have enough of a sense of the risks and value of the claims to make an informed judgment about the appropriate value of a settlement and a proper bid. At present, however, law firms have few avenues to address the perceived inadequacy of a settlement achieved by class counsel, other than to object to the settlement on their clients' behalf or to opt their clients out of the case. But neither of these options typically makes sense from the firm's financial standpoint. Bidding provides a realistic option by which the firm can help class members to achieve better recoveries and assume control of litigation that the class members believe to be more valuable than the settlement class counsel negotiated.

Bidding by law firms also raises some of the concerns raised with bidding by defendants' competitors: law firms may bid on a case to deprive a competing law firm of a full fee for its work, a prospect which may ruin a particular firm or discourage firms in general from entering the class action market. Unlike the situation with competitors, a prohibition on law firm bidding is unwise. Because class counsel receives a quantum meruit fee for the value of the work performed on the class's behalf leading up to the original settlement, (115) another law firm's successful bid on a single case is unlikely to ruin class counsel's firm or drive it from the market. (116) Nonetheless, a court must be sensitive to any anticompetitive motives behind a law firm's bid and should disqualify a bid from a law firm when such motives appear likely.

Others can also enter the settlement auction; perhaps one of the class members is a billionaire who wants to see justice done. Moreover, two or more of the three groups that are most likely to bid--advocacy groups, private equity financiers, and law firms--could form a bidding consortium. Permitting third-party financiers and law firms to bid at auction also lessens some of the least appealing aspects of third-party or law firm financing.

One problem with such financing is the injection of a new principal into the already messy principal-agent relationship between the class and class counsel. (117) Another is the potential for conflicts of interest when law firms enter into agreements with another firm to finance the litigation. (118) A third is the sense that financiers are getting rich at the expense of victims by taking a substantial portion of the class's recovery in return for financing. (119)

Settlement auctions do not eliminate these problems, but they blunt the critiques considerably. The concern for a new principal-agent relationship is gone entirely. (120) The only principal left after a settlement auction is the winning bidder, who has an incentive to tamp down agency costs by closely monitoring the lawyers who continue to prosecute the claims against the defendant. (121) Conflicts between law firms and the class are no longer a relevant issue.

Admittedly, bidding will not occur unless the bidder can anticipate a profit, so class members are likely to receive less than the fair market value of their claims; but class members will receive as much as or more than they would have received under the original settlement. (122) And it bears repeating that the in terrorem effect of settlement auctions should lead to better class settlements to begin with, so class members should capture a larger portion of the value of their claims.

D. The Emergence of an Auction Market

One of the concerns with a settlement auction is whether a robust market for settled class claims will develop. Interest from bidders will most likely arise in cases that are strong on the merits but badly undervalued in the original settlement. Anecdotal evidence suggests that lawyers sometimes sell out classes to enrich themselves, (123) a claim which, if true, suggests that a market might emerge. There is also anecdotal evidence to the contrary: that class actions extort exorbitant sums for meritless claims, (124) a claim which, if true, suggests that class settlements are already overvalued and no market will emerge. Empirical data to prove either of these claims is in short supply. (125)

It is likely true that some class actions settle for too much, some for roughly the right amount, and some for too little. The variance is undoubtedly due, in part, to the risk preferences of litigants and counsel. The settlement auction is designed to deal only with the situation in which the class's claims settle for too little. (126) Should no bidders emerge, a judge has some assurance that the class's claims are settled at market value. (127) Whether a market emerges in settlement auctions will provide a real-world test of the value of class settlements.

Even if class settlements are undervalued, a market may not emerge. As I have argued, a proper attorneys' fee structure and a correctly calibrated opt-out right should already discourage undervalued class settlements. (128) Another practical obstacle to the development of such a market is the availability of financing for bidders. It may be one thing for law firms or third-party financiers to take on the risk of loss when financing the legal expenses in a case, but taking on the risk of losing the bid price is a much larger proposition for which funders are likely to be scarcer.

On this point, however, empirical data does not suggest that financiers must come up with tens of millions of dollars in every settlement auction. (129) The median class action in either state or federal court appears to settle for something in the six-figure range. (130) Thus, the wherewithal to finance a successful bid in many class actions is not out of reach for a wide array of private investors, law firms, and consumer-advocacy groups.

An intriguing idea is an investor's ability to securitize a portfolio of lawsuits won at auction. Such investors might then cut up the claims into tranches based on the level of risk involved. (131) Securitization could be an attractive way to lure investors into the auction market. (132)

There is no guarantee that a settlement auction market will emerge. Even if it did emerge, the threat of settlement auctions may improve the quality of settlements to a point that investors can no longer find attractive class settlements on which to bid, and investors will leave the market for other opportunities. Class counsel and defendants might then return to the present status quo. But that return seems unlikely for two reasons. First, if class settlements again become significantly undervalued, bidders may return to the market. Second, once the prospect or reality of settlement auctions props up the value of class settlements, courts will have a new benchmark for determining whether class settlements are "fair, reasonable, and adequate," (133) and for disapproving settlements that are not.

E. The Timing of Opt-Outs

Along with settlement auctions and an hourly-plus-percentage-of-net-recovery attorneys' fee, an opt-out right helps to ensure that class members receive fair value for their claims. (134) In a settlement auction the timing of the opt-out right matters. If class members opt out before the settlement is put up for bid, then bidders have a better sense of the value of the claims on which they are bidding. But opt-outs might have remained in the case had they received the higher amount promised in the winning bid. On the other hand, if class members opt out after the award of the winning bid, the winning bidder must be allowed to adjust the bid to account for the decrease in value of the class's claims due to the exit of some class members. That adjustment might then set off a new round of opt-out rights, with a new round of adjustments, followed by a new set of opt-out rights--until eventually no one opts out. This process could be time-consuming and delay the settlement, as well as resolution of the underlying dispute against the defendant, for months or years.

As a practical matter, the evils of providing post-bid opt-out rights are greater. Therefore, once the original settlement is negotiated, a court should permit a period for opting out, and then put the settlement up for auction. If a bid higher than the class settlement emerges, the winning bidder should, on request, be given access to the identities of those who have opted out. The winning bidder can then determine whether to contact the opt-outs to request that they opt back into the settlement under the bid's terms. If any claimants opt in, their claims become the property of the winning bidder, just as are the claims of those who did not opt out.

F. Apportioning Fees

There remains a final, and rather tricky, problem: handling the compensation of class counsel when a winning bidder emerges. Some compensation is appropriate. The class counsel's work produced the original--albeit inadequate--settlement, as well as the discovery that served as the basis on which bidders prepared their offers. (135) On the other hand, a court should not award the fee that would have been awarded if the original settlement had been approved; otherwise, class counsel has no incentive to avoid negotiating an inadequate settlement.

Therefore, a court should compensate counsel on a quantum meruit basis for work that benefited the class or bidders. (136) In general, the hourly market rate should determine the fee. A deduction from the fee should be made for the reasonable time that the winning bidder's counsel must spend to come up to speed on the case. The winning bidder need not pay twice for the same work. Limiting the fee in this way deters class counsel from negotiating a settlement so substandard that it will attract potential bidders' attention, while also preventing a windfall to the winning bidder.

The difficult question is the timing of the payment to class counsel. One issue is whether compensation should be made at the time of the bid award or rather when the case concludes. Another issue is whether counsel should receive any compensation if the bidder loses the case or recovers less than the amount of the original settlement.

On the first issue, the better approach is to delay compensation until the case is resolved, so that the court can determine both the contribution of class counsel and the appropriate fee deductions. The answer to the second question is more nettlesome. The failure of the winning bidder to achieve a result as good as the one class counsel negotiated shines a new light on the value of class counsel's work in negotiating the settlement. Moreover, the chance that counsel will receive no compensation due to events beyond their control--the entry of a bidder into the case--may discourage lawyers from becoming class counsel, and may thus deprive society of the deterrence benefits that class actions can provide. But a flat rule that class counsel should receive quantum meruit compensation regardless of outcome may drive up the cost of bidding and discourage entry into the auction market, unless the class's claims are near-certain winners that the settlement badly undervalues. Compensation independent of outcome also converts class counsel's contingency agreement into a guaranteed fee, even when counsel has poorly served the class.

The question's closeness makes establishing a categorical rule difficult. In the early years, when an auction market is still emerging, the presumption should be against compensating class counsel when the bidder fails to recover. Likewise, compensation on a partial, pro rata basis is appropriate when the bidder achieves a result less favorable than the original settlement. Among the factors that could overcome the presumption are bad faith by the bidder (137) and the incompetence of the bidder's lawyers in handling the claim. (138) Creating a presumption against compensation provides an incentive for class counsel to negotiate a bid-proof settlement; it should not discourage ethical lawyers from becoming class counsel.

Conclusion

In class actions, settlement is where the rubber meets the road. The precise level of deterrence that the defendant faces becomes clear, and the cracks between the interests of class and counsel are exposed. If class actions are to do a better job achieving deterrence and compensation, while minimizing agency costs, adequate settlements are essential. Those who wish to improve class actions must concentrate their energy on the settlement process.

No single approach to ensuring the adequacy of settlements is sufficient. To some extent a procedural protection--a limited opt-out right--is a useful device to encourage better settlements. An incentive to class counsel, in the form of an hourly-and-percentage-of-net-recovery fee, is another useful device. Both of these measures seek to achieve substantively adequate settlements through indirect means. Even with these measures in place, a court should also attempt to gauge the substantive quality of the outcome. The present approach to determining whether a settlement is substantively "fair, reasonable, and adequate" is insufficient because a court operates with limited information about the strength of the settled claims. The settlement auction, which tests the value of the settlement in the market, is the best way to determine the quality of the settlement directly.

Standing alone, settlement auctions are not a perfect solution. They generate a series of practical issues that frustrate, to some degree, the auction's intended purpose. Working through these practical issues requires a constant focus on the aims of the auction process and, indeed, of class actions more generally: increasing deterrence and reducing the agency costs that thwart the delivery of adequate compensation to class members. As a means to achieve these goals, the settlement auction deserves its time in the sun.

(1.) Jonathan R. Macey & Geoffrey P. Miller, The Plaintiff s' Attorney's Role in Class Action and Derivative Litigation: Economic Analysis and Recommendations for Reform, 58 U. CHI. L. REV. 1, 6 (1991). Under this proposal, the proceeds of the auction would be distributed to class members whose claims against the defendant are extinguished. The winning bidder is then free to prosecute the claims against the defendant. See id.

(2.) See, e.g., In re Oracle Sec. Litig., 136 F.R.D. 639, 641 (N.D. Cal. 1991) (employing a lead counsel auction). Although never popular, the lead counsel auction was most common in securities litigation. See Jill E. Fisch, Lawyers on the Auction Block: Evaluating the Selection of Class Counsel by Auction, 102 COLUM. L. Rev. 650, 664-65 (2002). As a practical matter, this process screeched to a halt in 2003 when the Third Circuit forbade nearly all lead counsel auctions under the Private Securities Litigation Reform Act of 1995 (PSLRA). See In re Cendant Corp. Litig., 264 F.3d 201, 273 (3d Cir. 2001).

(3.) For arguments that financiers can better monitor class counsel than class members can, see Elizabeth Chamblee Burch, Financiers as Monitors in Aggregate Litigation, 87 N.Y.U. L. Rev. 1273 (2012); John C. Coffee, Jr., Litigation Governance: Taking Accountability Seriously, 110 COLUM. L. Rev. 288, 339-43 (2010). For a positive appraisal of litigation financing, see Jonathan T. Molot, Litigation Finance: A Market Solution to a Procedural Problem, 99 Geo. L.J. 65 (2010). For a thoughtful discussion developed as a structured set of questions that judges must consider, see Bert I. Huang, Litigation Finance: What Do Judges Need to Know?, 45 COLUM. J.L. & SOC. PROBS. 525 (2012). Third-party financing of the kind I describe here--in which the investor agrees to loan money in return for a percentage of the class's recovery, with no recourse to class members or counsel--has been used in some countries. See Burch, supra, at 1301-02.

(4.) See Hughes v. Kore of Ind. Enter., 731 F.3d 672, 678 (7th Cir. 2013) ("A time-saving alternative might be a class action with the stated purpose, at the outset of the suit, of a collective award to a specific charity. We are not aware of such a case, but mention the possibility of it for future reference."). This proposal would apply only when putative class members held claims with a "negative value"--in other words, claims in which individual recovery would be so small that individual lawsuits would not be worthwhile. Id. at 677-78.

(5.) See id. at 677.

(6.) See Macey & Miller, supra note 1, at 14.

(7.) In two instances, some class members may not receive even pennies on the dollar. First, some courts have used a cy pres theory to award funds obtained on the class's behalf to a charity--typically because the number of claimants or the modest size of any unclaimed recovery made distribution to individual class members economically infeasible. See In re Baby Prods. Antitrust Litig., 708 F.3d 163,172 (3d Cir. 2013) ("When excess settlement funds remain after claimants have received the distribution they are entitled to under the terms of the settlement agreement, there are three principal options for distributing the remaining funds--reversion to the defendant, escheat to the state, or distribution of the funds cy pres."); Principles of the Law of Aggregate Litigation [section] 3.07 (2010) [hereinafter Aggregate LITIGATION] (permitting cy pres relief when funds remain "because some class members could not be identified," and when "the amounts involved are too small to make individual distributions economically viable or other specific reasons exist that would make such further distributions impossible or unfair"); see also In re Lupron Mktg. & Sales Practices Litig., 677 F.3d 21, 31-32 (1st Cir. 2012) (permitting a court to give unclaimed funds in a case involving overcharges for prostate drug to prostate cancer research and other related research); In re Bluetooth Headset Prods. Liab. Litig., 654 F.3d 935, 949-50 (9th Cir. 2011) (overturning approval of a settlement that provided some injunctive relief to class members but provided monetary recovery only to charities); Masters v. Wilhelmina Model Agency, Inc., 473 F.3d 423, 435-36 (2d Cir. 2007) (discussing permissible uses of cy pres relief). Second, in rare cases a class member may end up owing money as a result of a class settlement. The most famous example is that of Dexter Kamilewicz, who received a $2.19 payout in a class settlement but was charged $91.33 to pay for class counsel's fees and costs. See Kamilewicz v. Bank of Bost. Corp., 92 F.3d 506, 508 (7th Cir. 1996).

(8.) Cf. Fisch, supra note 2, at 667.

(9.) See supra notes 1, 3-4.

(10.) See Alan D. Miller & Ronen Perry, The Reasonable Person, 87 N.Y.U. L. REV. 323, 346 n.97 (2012) ("If one is truly committed to welfare maximization, one must ensure that potential injurers internalize the costs of negligent conduct.").

(11.) See David Rosenberg, Mandatory-Litigation Class Action: The Only Option for Mass Tort Cases, 115 Harv. L. Rev. 831, 848 (2002) ("[O]ne benefit of class action scale economies is [to] motivate courts as well as parties to invest in litigation so as to achieve optimal deterrence and insurance.").

(12.) See id. at 847 ("Class action scale economies produce the well-known benefit of avoiding duplicative litigation over common questions."); see also Bruce L. Hay & David Rosenberg, "Sweetheart" and "Blackmail" Settlements in Class Actions: Reality and Remedy, 75 Notre Dame L. rev. 1377, 1383 (2000) ("[A class action] enables plaintiffs to exploit the 'economies of scale' the defendant already naturally enjoys from treating separate claims as a single litigation unit.").

(13.) See, e.g., James D. Cox, Making Securities Fraud Class Actions Virtuous, 39 Ariz. L. Rev. 497, 497 (1997) ("Securities class actions proceed with the objective of permitting those separated wrongfully from their wealth to get some of it back.").

(14.) See Alexandra D. Lahav, The Political Justification for Group Litigation, 81 FORDHAM L. Rev. 3193, 3201 (2013) (noting the "role of group litigation in promoting equality" with respect to "two separate functions: access to the law declaration (vindication) and distributive justice (remedies)").

(15.) There are two overdeterrence concerns. One is liability for technical violations that are punishable by damages far in excess of actual loss. For instance, as originally written, the Truth in Lending Act made each violation punishable by a $100 fine, even if the actual loss was less. Truth in Lending Act, Pub. L. No. 90-321, [section] 130, 82 Stat. 157 (1968) (codified with some differences in language at 15 U.S.C. [section] 1640 (2012)). Thus, a minor violation could lead to crushing liability. Cf. Ratner v. Chem. Bank N.Y. Trust Co., 54 F.R.D. 412, 416 (S.D.N.Y. 1972) (holding that a class action seeking a $100-per-violation Truth in Lending Act penalty for the defendant's technical error in the bills of 13,000 credit-card customers was not a superior way to resolve the dispute when no customer lost money). The second overdeterrence concern involves a law-abiding defendant who pays a substantial sum to settle frivolous or low merit claims in order to avoid the risk of ruinous aggregate liability. See In re Rhone-Poulenc Rorer Inc., 51 F.3d 1293, 1299 (7th Cir. 1995) (describing the "concern with forcing these defendants to stake their companies on the outcome of a single jury trial, or be forced by fear of the risk of bankruptcy to settle even if they have no legal liability"); id. at 1298 (describing these "settlements induced by a small probability of an immense judgment in a class action [as] 'blackmail settlements"'). This second form of overdeterrence is a type of error cost. See infra note 21 and accompanying text.

(16.) For the most trenchant criticism of class actions on individual-autonomy grounds, arguing that class actions are inconsistent with the autonomy that undergirds liberal democracy, see generally MARTIN H. REDISH, WHOLESALE JUSTICE: CONSTITUTIONAL Democracy and the Problem of the Class Action Lawsuit (2009).

(17.) See In re Baby Prods. Antitrust Litig., 708 F.3d 163, 179 (3d Cir. 2013) ("Some class actions are based on so-called negative value claims, that is, claims that could not be brought on an individual basis because the transaction costs of bringing an individual action exceed the potential relief.").

(18.) This cost must be weighed against the benefits of providing cost-effective relief and deterring defendants from cheating large numbers of people out of small amounts of money. Indeed, an axiom of present class action practice is the class action's special utility in negative-value cases. See, e.g., Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (noting that although Federal Rule of Civil Procedure 23(b)(3) "does not exclude from certification cases in which individual damages run high," the dominant function of Rule 23(b)(3) has been the "vindication of 'the rights of groups of people who individually would be without effective strength to bring their opponents into court at all'" (quoting Benjamin Kaplan, A Prefatory Note, 10 B.C. INDUS. & COM. L. REV. 497, 497 (1969))); In re Rhone-Poulenc, 51 F.3d at 1299 ("In most class actions--and those the ones in which the rationale for the procedure is most compelling--individual suits are infeasible because the claim of each class member is tiny relative to the expense of litigation.").

(19.) See FED. R. CIV. P. 23(c)(2), (e)(1) (detailing circumstances in which class members receive notice).

(20.) See Fed. R. CIV. P. 16(c)(2) (describing case-management powers of judges); MANUAL FOR COMPLEX LITIGATION (FOURTH) [section] 10.13 (2004) (urging judges to manage complex litigation actively).

(21.) See Irwin A. Horowitz & Kenneth S. Bordens, The Consolidation of Plaintiffs: The Effects of Number of Plaintiffs on Jurors' Liability Decisions, Damage Awards, and Cognitive Processing of Evidence, 85 J. APPLIED PSYCHOL. 909, 914, 917 (2000) (reporting experimental data showing that the likelihood of recovery increases as more plaintiffs are aggregated, but that the average damage award decreases if more than four plaintiffs are aggregated); Irwin A. Horowitz & Kenneth S. Bordens, The Effects of Outlier Presence, Plaintiff Population Size, and Aggregation of Plaintiffs on Simulated Civil Jury Decisions, 12 Law & Hum. Behav. 209, 226 (1988) (reporting experimental data showing that the aggregation of claims increases the likelihood of recovery for weak claims, but suppresses the value of strong claims). The data do not reveal whether the increase in the probability of recovery and the decrease in per-claimant recovery wash out, or whether class actions require defendants to pay more than they should. Excess payments, which are a type of error cost, must then be balanced against the cost that would be incurred if some of the aggregated plaintiffs did not file individual suits, for the failure to provide any remedy to deserving victims is itself an error cost. See generally RICHARD A. POSNER, ECONOMIC ANALYSIS OF Law 593-95 (7th ed. 2007) (discussing error costs).

(22.) Cf. Macey & Miller, supra note 1, at 12 (noting that lawyers as the agents of their clients rarely have interests that are "perfectly aligned" with those of their clients).

(23.) See Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. FIN. ECON. 305, 308 (1976) ("In most agency relationships the principal and the agent will incur positive monitoring and bonding costs (non-pecuniary as well as pecuniary), and in addition there will be some divergence between the agent's decisions and those decisions which would maximize the welfare of the principal." (footnote omitted)).

(24.) See Michael D. Sant' Ambrogio & Adam S. Zimmerman, The Agency Class Action, 112 COLUM. L. REV. 1992,2056 (2012) ("Participants in many class actions lack rational incentives to monitor class counsel when they have comparatively small stakes in the entire enterprise.").

(25.) In some instances class representatives receive valuable recoveries that other class members do not. See, e.g., Radcliffe v. Experian Info. Solutions Inc., 715 F.3d 1157,1164 (9th Cir. 2013) (holding that when the class representative's receipt of a large incentive payment was conditioned on support for the settlement, "the incentive awards here corrupt the settlement by undermining the adequacy of the class representatives and class counsel"); Vassalle v. Midland Funding LLC, 708 F.3d 747, 757 (6th Cir. 2013) (rejecting a settlement in which "the class representatives [had] no interest in vigorously prosecuting the unnamed class members' most important interest--the ability to ... contest their debts in court" because, under the settlement, "the class representatives' debts [were] forgiven"). The representative may sell out the interests of only some class members. See, e.g., supra note 7 (discussing Dexter Kamilewicz case). Not all differential payments between representatives and the group necessarily involve agency cost issues. For instance, courts often approve modest additional payments to class representatives to compensate them for their time and effort in being deposed or otherwise assuming responsibilities not borne by class members. See, e.g., Silberblatt v. Morgan Stanley, 524 F. Supp. 2d 425, 435-36 (S.D.N.Y. 2007) (awarding some, but not all, of the additional compensation the class representative sought).

(26.) Creative Montessori Learning Ctrs. v. Ashford Gear LLC, 662 F.3d 913, 918 (7th Cir. 2011) ("We and other courts have often remarked the incentive of class counsel, in complicity with the defendant's counsel, to sell out the class by agreeing with the defendant to recommend that the judge approve a settlement involving a meager recovery for the class but generous compensation for the lawyers--the deal that promotes the self-interest of both class counsel and the defendant and is therefore optimal from the standpoint of their private interests.").

(27.) For discussions of the inherent conflicts among class members, see John C. Coffee, Jr., Class Action Accountability: Reconciling Exit, Voice, and Loyalty in Representative Litigation, 100 COLUM. L. REV. 370, 385-93 (2000); Jay Tidmarsh, Rethinking Adequacy of Representation, 87 TEX. L. REV. 1137, 1158-75 (2009). The same conflicts exist in other aggregated litigation. See Charles Silver & Lynn Baker, I Cut, You Choose: The Role of Plaintiffs' Counsel in Allocating Settlement Proceeds, 84 VA. L. Rev. 1465, 1468-69 (1998) ("There being no way to eliminate conflicts [of interest] from multiple-claimant representations, the only question is how to deal with them.").

(28.) Cf. AGGREGATE Litigation, supra note 7, [section] 1.04 cmt. E ("Maximizing [the net value of a group of claims] is a central object of aggregate litigation.").

(29.) Fed. R. Civ. P. 23.

(30.) Fed. R. Civ. P. 23(a)(4), (g).

(31.) Fed. R. Civ. P. 23(e)(2). The Supreme Court made clear that the standards of adequate representation and fair settlement are distinct; in particular, a fair settlement does not prove the adequacy of the representation. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 620-22 (1997).

(32.) The Supreme Court has squarely addressed adequacy of representation only twice, in both cases finding conflicts of interest between the class representatives and some of the class members. See Amchem, 521 U.S. at 625-28; Hansberry v. Lee, 311 U.S. 32, 41-45 (1940). The Court has never addressed adequacy in the context of a conflict between the class and its counsel. This focus on conflicts as a measure of adequacy poses problems. Given the inevitability of conflicts among class members, a literal focus on conflicts would doom every class action. See supra note 27 and accompanying text. Conversely, focusing on conflicts fails to catch cases in which the class representative and counsel achieve an inadequate result on behalf of the class.

(33.) Cf. AGGREGATE Litigation, supra note 7, [section] 1.05 (recommending that a concept of adequate representation be applied even in nonclass aggregate litigation); Charles Silver, The Responsibilities of Lead Lawyers and Judges in Multidistrict Litigations, 79 FORDHAM L. REV. 1985, 1987-91 (2011) (arguing that nonclass lead counsel should have fiduciary obligations to aggregated claimants).

(34.) See supra note 7 and accompanying text.

(35.) Two classic examples are: the Playboy Club litigation, in which class members received a chit for a drink, and class counsel received $275,000 in fees; and the GM Pick-Up Truck litigation, in which class members received a $1,000 coupon, usable only within fifteen months, toward the purchase of the same type of truck (or $500 toward any other GM vehicle, with certain restrictions), and counsel received $9.5 million in fees. For two retellings of the Playboy Club litigation, in which the settlement was approved, see Milton Handler, The Shift from Substantive to Procedural Innovations in Antitrust Suits--The Twenty-Third Annual Antitrust Review, 71 COLUM. L. REV. 1,10 (1971); Arthur R. Miller, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the "Class Action Problem," 92 Harv. L. Rev. 664, 679 (1979). For a discussion of the GM Pick-up Truck settlement, in which the settlement was rejected on appeal, see In re General Motors Corp. Pick-Up Truck Fuel Tank Products Liability Litigation, 55 F.3d 768 (3d Cir. 1995). Supposed abuses of coupon settlements led Congress to enact legislation that both required close judicial scrutiny and limited attorneys' fees as mechanisms to curb the practice. See 28 U.S.C. [section] 1712 (2012).

(36.) A side agreement that provided additional benefits to class counsel's individual clients (who were not part of the class) became one of the grounds on which the Amchem settlement's fairness was attacked, although the Third Circuit and the Supreme Court ultimately rejected the settlement for different reasons. See Georgine v. Amchem Prods., Inc., 157 F.R.D. 246, 294-99 (E.D. Pa. 1994), rev'd on other grounds, 83 F.3d 610 (3d Cir. 1996), aff'd, 521 U.S. 597 (1997); see also Susan P. Koniak, Feasting While the Widow Weeps: Georgine v. Amchem Products, Inc., 80 CORNELL L. REV. 1045, 1052-57 (1995) (critiquing the Amchem settlement on this ground).

(37.) See, e.g., In re BabyProds. Antitrust Litig., 708 F.3d 163, 176 (3d Cir. 2013) (noting concern about the limited amount of compensation going to class members); Dennis v. Kellogg Co., 697 F.3d 858 (9th Cir. 2012) (disapproving a settlement in which half of the settlement award was set aside for cy pres relief); In re Katrina Canal Breaches Litig., 628 F.3d 185 (5th Cir. 2010) (disapproving a settlement in which attorneys' fees and costs consumed the bulk of the award, leaving only a modest amount for possible cy pres distribution).

(38.) See Samuel Issacharoff, Class Action Conflicts, 30 U.C. DAVIS L. Rev. 805, 829 (1997) ("No matter how virtuous the judge, the fact remains that courts are overworked, they have limited access to quality information, and they have an overwhelming incentive to clear their docket. They cannot reliably police the day-to-day interests of absent class members." (footnote omitted)); William B. Rubenstein, The Fairness Hearing: Adversarial and Regulatory Approaches, 53 UCLA L. Rev. 1435, 1468 (2006) (noting "the informational problems facing the trial judge" in "assessing] the strengths and weaknesses of the plaintiffs' claims"); cf. Macey & Miller, supra note 1, at 4 ("[J]udicial review of settlements and fee requests ... is often haphazard, unreliable, and lacking in administrable standards."). Federal courts have adopted multifactor tests--often involving nine factors, but sometimes supplementing those with six or even eleven additional factors--to test a class settlement's fairness. See Wal-Mart Stores, Inc. v. Visa U.S.A., Inc., 396 F.3d 96, 117 (2d Cir. 2005) (listing nine factors); In re Prudential Ins. Co. of Am. Sales Practices Litig., 148 F.3d 283, 323 & n.73 (3d Cir. 1998) (listing six factors and offering an additional eleven in a footnote). The factors tend to be weighted toward the risks and costs of continued litigation in comparison to the proposed settlement. As in any multifactor test, however, the weighing of the factors is open-ended and subject to manipulation.

(39.) Some scholars have argued that, at least with respect to negative-value claims that are aggregated in a class action, failing to compensate class members adequately--and by extension failing to reduce agency costs--are not troubling concerns. See Brian T. Fitzpatrick, Do Class Action Lawyers Make Too Little?, 158 U. Pa. L. Rev. 2043, 2044 (2010) (arguing that in small-stakes class actions, "it is hard to see, as a theoretical matter, why the lawyers should not receive everything and leave nothing for class members at all"); Myriam Gilles & Gary B. Friedman, Exploding the Class Action Agency Costs Myth: The Social Utility of Entrepreneurial Lawyers, 155 U. PA. L. REV. 103, 104-05 (2006) ("[T]he so-called 'agency cost' problem is mostly a mirage.... In reality, there is generally no legitimate utilitarian reason to care whether class members with small claims get compensated at all"). These arguments, which see class actions purely in deterrence terms, miss the mark. Private litigation is Janus-faced; it seeks to deter, but it does so by compensating the victims. Cf. POSNER, supra note 21, at 192 (noting that compensation provides an incentive to sue). As long as a class action has a preclusive effect on class members' claims, denying class members the ability to reap whatever value their claims enjoy in an aggregated process through the transfer of that wealth to a third person requires an account of class actions and private litigation different from the one presently in use for Federal Rule of Civil Procedure 23. Furthermore, the point at which an agent is willing to settle may not be the socially optimal settlement point, thus frustrating the goal of adequate deterrence.

(40.) See supra notes 1-4 and accompanying text.

(41.) For two such proposals, which differ substantially, see Kevin M. Clermont & John D. Currivan, Improving on the Contingent Fee, 63 CORNELL L. REV. 529, 546-50 (1978); Hay & Rosenberg, supra note 12, at 1394-402. The Clermont-Currivan proposal was designed for single-plaintiff litigation, but can be applied to class actions or other mass litigation. See Clermont & Currivan, supra, at 584 n.185; see also JAY TIDMARSH, CLASS ACTIONS: FIVE Principles To Promote Fairness and Efficiency 125-29,215,226-33 (2013) (suggesting a modified Clermont-Currivan approach both to ensure optimal class size and to achieve fair distribution of class proceeds). Thus far, courts have not adopted either of these fee approaches, but have tended to compensate counsel either by awarding fees for hours worked (the lodestar approach, in which the court may enhance the market rate by a multiplier if the case is highly successful) or by awarding a percentage of the recovery. See, e.g., Wal-Mart Stores, Inc., 396 F.3d at 121 (adopting a "percentage of the fund" approach); Fischel v. Equitable Life Assurance Soc'y of U.S., 307 F.3d 997,1006 (9th Cir. 2002) ("In a common fund case, the district court has discretion to apply either the lodestar method or the percentage-of-the-fund method in calculating a fee award."). Both the lodestar and the percentage-of-the-fund approaches create mismatches between the incentives of counsel to maximize the attorneys' fee and the goal of the class to maximize its recovery. See McDaniel v. Cnty. of Schenectady, 595 F.3d 411,418-19 (2d Cir. 2010) (noting that in many situations, the lodestar method leads counsel "to bill as many hours as possible, to do unnecessary work, and for these reasons also can create a disincentive to early settlement," while the percentage-of-the-fund approach "can create perverse incentives of its own, potentially encouraging counsel to settle a case prematurely once their opportunity costs begin to rise").

(42.) For a seminal article examining the right of class members to exit, see Coffee, supra note 27, at 417-28. At present, courts must give class members an opportunity to opt out in class actions "wholly or predominately for money judgments." See FED. R. Civ. P. 23(c)(2)(B) (requiring an opt-out right in (b)(3) class actions); Phillips Petrol. Co. v. Shutts, 472 U.S. 797, 811-12 & n.3 (1985) (requiring such a right as a matter of due process). In a previously certified (b)(3) class action, a court may, but need not, provide a second opt-out right at the time of settlement. See Fed. R. Crv. P. 23(e)(4). In nonclass litigation, each client can in theory "opt out" by ending the attorney-client relationship and retaining different counsel. But they must compensate the lawyer for time spent; moreover, in recent years, lawyers have increasingly crafted client contracts that make it difficult for individual clients to extract themselves from group litigation. See AGGREGATE LITIGATION, supra note 7, [section] 3.17 (recommending that under certain conditions including prior consent, clients are bound to accept a settlement as long as a substantial majority of other clients represented by the same lawyer on the same matter vote to accept the settlement). But see Tax Auth., Inc. v. Jackson Hewitt, Inc., 898 A.2d 512 (N.J. 2006) (rejecting such an approach).

(43.) This idea underlies prior proposals: to auction the class's claims to a single bidder, see supra note 1 and accompanying text; to rely on a third party to finance the litigation, see supra note 3 and accompanying text; and to give class members' negative-value claims to a public-interest organization, see supra note 4 and accompanying text. Because a defendant can bid on the claim at auction, the auction proposal does not invariably result in ownership passing to an entity with an incentive to monitor counsel; a defendant who wins the bid will drop the case. See Macey & Miller, supra note 1, at 108.

(44.) See Coffee, supra note 27, at 421.

(45.) Cf. Fisch, supra note 2, at 667.

(46.) A recent example of the difficulty of estimating the value of claims at the outset of a lawsuit occurred during the Madoff bankruptcy, in which a sophisticated claimant traded its claims in the bankruptcy estate for thirty-two cents on the dollar, and after an aggressive recovery effort by the bankruptcy trustee, the claims were eventually worth seventy cents on the dollar. See Nathan Vardi, The Great Baupost Madoff Claim Trade that Made a Big Madoff Feeder Fund a Loser Again, FORBES (Jan. 16, 2013, 10:55 AM), http://www.forbes.com/sites/ nathanvardi/2013/01/16/the-great-baupost-madoff-claims-trade-that-made-a-big-madofffeeder-fund-a-loser-again [http://perma.cc/5MH6-C5QZ]. I thank Adam Zimmerman for this point.

(47.) See Randall S. Thomas & Robert G. Hansen, Auctioning Class Action and Derivative Lawsuits: A Critical Analysis, 87 Nw. U. L. REV. 423, 448 (1993) (noting that the auction price increases when uncertainty over the value of the auctioned good is lower).

(48.) For instance, the approach recommended by Professor Clermont and Mr. Currivan is superior to the approach recommended by Professors Hay and Rosenberg because it gives counsel an incentive to continue to work until (and only until) another dollar spent on legal representation does not yield another dollar in gain to the client. In contrast, the proposal by Professors Hay and Rosenberg eliminates the incentive to engage in a sweetheart settlement but does not otherwise align the interests of class and counsel. Compare Clermont & Currivan, supra note 41, at 530, with Hay & Rosenberg, supra note 12, at 1381. Even under the Clermont-Currivan proposal, attorneys with private agendas within a firm may thwart the delivery of an optimal level of legal services. See Morris Ratner, A New Model of Plaintiffs' Class Action Attorneys, 31 Rev. LITIG. 757, 761-62 (2012) (describing the effect of these agendas on the quality of a firm's legal representation of a class). Moreover, these fee structures can neither prevent an attempt to undercut an optimal class by filing a dueling class action, nor thwart a defendant's efforts to drag the case on beyond the optimal end point from the standpoint of class and counsel. See TIDMARSH, supra note 41, at 129-30,223,230-31 (describing the limits of the Clermont-Currivan approach in avoiding agency costs).

(49.) On the one hand, few class members exercise their right to opt out. See Vassalle v. Midland Funding LLC, 708 F.3d 747, 753 (6th Cir. 2013) (reporting an opt-out rate of 0.3 percent in a class action containing 1.4 million members); THOMAS E. WILLGING ET AL., FED. Judicial Ctr., Empirical Study of Class Actions in Four Federal District Courts 52 (1996) (reporting that the median rate of opting out in class actions contained within the study was between 0.1 percent and 0.2 percent). On the other hand, class members may choose to opt out even when it is socially optimal for them to remain in the class. In particular, some class members may strategically opt out in order to extract rents from the defendant, creating a different type of cost. See Michael A. Perino, Class Action Chaos? The Theory of the Core and an Analysis of Opt-Out Rights in Mass Tort Class Actions, 46 EMORY L.J. 85, 97 (1997) ("[S]ome individual claimants may impose significant externalities on other litigants through pursuit of disaggregative strategies.").

(50.) See Clermont & Currivan, supra note 41, at 546-50 (using a marginal-utility analysis to describe the reasons why this fee creates an alignment of incentives). To align incentives properly, the fee must also be subject to some side constraints: the fee must be capped at the amount of recovery to the class; the fund on which the percentage-of-the-net recovery portion of the fee is calculated must be the award given to the class itself, thus excluding money given to other entities through cy pres relief; and counsel must know at the outset that this fee structure will be used. TIDMABSH, supra note 41, at 227; Jay Tidmarsh, Cy Pres and the Optimal Class Action, 82 Geo. WASH. L. REV. 767, 788-89 (2014).

(51.) For a discussion of reverse auctions, see John C. Coffee, Jr., Class Wars: The Dilemma of the Mass Tort Class Action, 95 COLUM. L. Rev. 1343,1370-73 (1995). One apparent example of a reverse auction involved litigation that spilled out of a hostile takeover. See Matsushita Elec. Indus. Co. v. Epstein, 516 U.S. 367 (1996); Epstein v. MCA, Inc., 126 F.3d 1235 (9th Cir. 1997), opinion withdrawn and superseded by 179 F.3d 641 (9th Cir. 1999); In re MCA, Inc., 598 A.2d 687 (Del. Ch. 1991). Although reverse auctions appear to occur infrequently, the problem should be addressed as long as the benefits of preventing reverse auctions outweigh the costs.

(52.) E.g., Macey & Miller, supra note 1 (writing a seminal article); see Thomas & Hansen, supra note 47 (providing a sympathetic critique and refining the plan); see also Jonathan R. Macey & Geoffrey P. Miller, Auctioning Class Action and Derivative Suits: A Rejoinder, 87 Nw. U. L. Rev. 458 (1993) (responding to these suggestions). The idea of an auction was previously discussed but rejected in John C. Coffee, Jr., Understanding the Plaintiff's Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions, 86 COLUM. L. Rev. 669, 691-93 (1986).

(53.) In a subsequent article, Professor Miller proposed an auction at the time of settlement. See Geoffrey P. Miller, Competing Bids in Class Action Settlements, 31 HOFSTRAL. Rev. 633 (2003). Unlike the proposal in this Article, Professor Miller's proposal would have auctioned only the position of lead counsel. Once a settlement was announced, an objecting class member who believed that the case had settled too low could bid on the case; if the highest bid guaranteed that class members would be no worse off than they were under the settlement, the highest bidder then could choose different lead counsel to continue the prosecution of the case. Id. at 639-40. If new counsel eventually achieved a better result, the excess would be distributed pro rata among the class; but if a worse result was achieved, the bidder would have to make good on the guarantee. Id. at 640. The purpose of this proposal is to eliminate the problem of reverse auctions. Id. at 649-50.

The proposal in this Article is different. First, class members immediately receive payment in the amount of the highest bid, rather than waiting for the case to end. See infra Part III.B.4. In addition, bidders are not limited to objecting class members. See infra Part III.C. Third, and perhaps most important, the successful bidder retains the full amount of any excess achieved in subsequent litigation--a fact that should create a broader market for bidders. See infra Part III.B.4. Finally, although this Article's proposal also helps to prevent reverse auctions, its basic purpose is to be one aspect of a set of reforms that achieves the broader goal of ensuring fair settlements and limiting agency costs. See infra notes 61-68 and accompanying text (describing the benefits of a settlement auction).

One court in a derivative action attempted a lead counsel auction akin to the one that Professor Miller proposed. See Forsythe v. ESC Fund Mgmt. Co. (U.S.), C.A. No. 1091-VCL, 2012 WL 1655538, at *6 (Del. Ch. Apr. 3, 2012) (giving objecting class members sixty days to post a bond guaranteeing to the class the amount of the present settlement in the event that the objectors were to recover less than the settlement amount). In the end, objecting class members failed to convince the court that their financing arrangement was reasonable from the viewpoint of the corporation on whose behalf the derivative action was brought. As a result, the court approved the original settlement. See Forsythe v. ESC Fund Mgmt. Co. (U.S.), C.A. No. 1091-VCL, 2013 WL 458373, at *1 (Del. Ch. Feb. 6, 2013).

(54.) For further discussion of the basic constraints on the bid, see infra Part III.B.I.

(55.) Admittedly, counsel must have access to means for funding the settlement. See infra Part III.D (discussing this practical problem).

(56.) If the settlement occurs in a nonclass aggregation, the equivalent to an opt-out would be the individual's refusal to consent to the settlement. Because a court can control the right to opt out more easily than the circumstances under which an individually represented client refuses to accept a settlement, this part of the discussion does not apply to other aggregate litigation.

(57.) No opt-out right exists in class actions brought under Rules 23(b)(1) and 23(b)(2). Compare FED. R. Civ. P. 23(c)(2)(A) (not mentioning an opt-out right for (b)(1) and (b)(2) classes), with FED. R. Civ. P. 23(c)(2)(B)(v) (providing an opt-out right in (b)(3) classes). In extraordinary circumstances, courts can nonetheless allow members in (b)(1) and (b)(2) classes to opt out. See Eubanks v. Billington, 110 F.3d 87, 94-95 (D.C. Cir. 1997); Cnty. of Suffolk v. Long Island Lighting Co., 907 F.2d 1295, 1302-05 (2d Cir. 1990). But see Thomas v. Albright, 139 F.3d 227, 235 (D.C. Cir. 1998) (holding that the district court abused its discretion in permitting plaintiffs to opt out). In addition, a court can allow an additional opt-out opportunity at the time of settlement in a (b)(3) class action. Fed. R. Civ. P. 23(e)(4).

(58.) See TIDMARSH, supra note 41, at 180-86 (providing an analysis of the circumstances that fulfill this requirement). A class member could also opt out by signing a release disclaiming any interest in pursuing the claim on an individual basis. Although it seems odd that a rational individual would prefer no recovery, some people who do not want to be bothered may prefer to opt out.

(59.) Although this opt-out right should be available, it should not be the sole mechanism by which courts determine the fairness of a settlement. See Christopher R. Leslie, The Significance of Silence: Collective Action Problems and Class Action Settlements, 59 Fla. L. REV. 71, 121-23 (2007) (arguing that courts often misinterpret the significance of opt-outs, mistakenly believing that those class members who choose not to leave the settlement believe that the settlement is fair).

(60.) Although this Article has focused on achieving fair class settlements, the same three criteria, including an auction, could in principle apply to other aggregate settlements. See AGGREGATE Litigation, supra note 7, [section] 1.02 cmt. a (noting that class actions are one form of aggregate litigation). In practice the difficulty of extending these criteria is finding a source of power by which a court in nonclass litigation can control the fees, order an auction, and permit opt-outs. All of these powers can be found in Rule 23. See FED. R. Crv. P. 23(h) (permitting a court to award "reasonable attorney's fees"); Fed. R. Civ. P. 23(e)(2) (giving a court the power to approve a settlement); FED. R. Civ. P. 23(e)(4) (permitting an opt-out right at settlement). See generally FED. R. Civ. P. 23(d)(1)(A) (giving courts the power to "determine the course of proceedings").

(61.) See supra text accompanying note 55.

(62.) See William Vickrey, Auctions, Markets, and Optimal Allocation, in BIDDING AND Auctioning for Procurement and Allocation 13-14 (Yakov Amihuded., 1976) (noting that an "open auction ... assures Pareto optimality"); see also Lucian Arye Bebchuck, The Case for Facilitating Competing Tender Offers, 95 HARV. L. Rev. 1028,1030 (1982) (noting that in the context of tenders offers for a company, an auction "has several socially beneficial effects, such as the allocation of targets' assets to their most valuable use").

(63.) See infra Part III.F. (discussing the fee that counsel earns when the original settlement loses to a higher bid). For now it is enough to know that counsel will typically receive a fee that does not fully compensate for the risk of undertaking the litigation.

(64.) FED. R. Civ. P. 23(e)(2).

(65.) See WILLGING ET AL., supra note 49, at 57 (noting that among four federal district courts, there were no objections filed to class settlements in 42 to 64 percent of the cases); Theodore Eisenberg & Geoffrey Miller, The Role of Opt-Outs and Objectors in Class Action Litigation: Theoretical and Empirical Issues, 57 VAND. L. REV. 1529,1532-33 (2004) (reporting that on average, fewer than 5 percent of the class objected to a settlement in civil rights cases, and objections in securities and consumer cases were lower); id. ("The overall impression across the range of cases in the study is that ... objections are extremely uncommon.").

(66.) See, e.g., In re Law Office of Jonathan E. Fortman, LLC, No. 4:13MC00042AGF, 2013 WL 414476, at *1 (E.D. Mo. Feb. 1, 2013) ("[W]hen assessing the merits of an objection to a class action settlement, courts consider the background and intent of objectors and their counsel, particularly when indicative of a motive other than putting the interest of the class members first."); In re Oil Spill by Oil Rig Deepwater Horizon, 295 F.R.D. 112, 159 & n.40 (E.D. La. 2013) (finding no merit in an objection presented by an attorney who had been "deemed a serial objector" and who "had been found to have engaged in bad faith and vexatious conduct") (internal quotation marks omitted).

(67.) See WILLGING ET AL., supra note 49, at 58 (noting that no changes in the settlement were made in 90 percent of the cases in which an objection was filed).

(68.) Cf. Michael C. Jensen, Corporate Control and the Politics of Finance, 4 J. APP. CORP. Fin. 13, 32 (1991) (arguing that one advantage of auctioning off companies in bankruptcy is that auctions "separate the task of assessing the firm's value from that of dividing that value among creditors and equityholders, effectively assigning the first to capital markets and the second to the courts").

(69.) Cf. Peter Cramton et al., Using Spectrum Auctions to Enhance Competition in Wireless Services, 54 J.L. & ECON. 167, 187 (2011) (noting that "[e]fficient auctions raise substantial revenues" but that "focusing more on revenues likely distorts the outcome away from social welfare maximization"); Thomas & Hansen, supra note 47, at 456-57 (noting the potential divergence between yielding revenue for class members and enhancing social welfare under a system that auctioned class actions).

(70.) A successful bidder who presses the case but either loses or recovers less than the bid has not necessarily made a bad deal; the expected value of the claims at the time of the bid may have been high enough that the bid was a good value. The text assumes that the successful bidder made a bad deal.

(71.) I assume that the American rule applies, so that each side bears its own attorneys' fees and expenses.

(72.) This circumstance is a specific example of the broader principle that private incentives can lead parties to make litigation choices that are not socially optimal. For a general analysis, see Steven Shavell, The Fundamental Divergence Between the Private and the Social Motive to Use the Legal System, 26 J. LEGAL STUD. 575 (1997).

(73.) Whether a defendant should be allowed to meet the terms of a winning bid is not as simple a question as it might seem. First, if matching is allowed, the court must require the defendant to compensate the winning bidder for the bidder's reasonable costs of due diligence investigation and bid preparation, plus (perhaps) a modest premium to compensate the bidder for the risk that its bid might not have been accepted. Second, allowing a defendant to bid on the claims against it could drive down the price at auction. See Thomas & Hansen, supra note 47, at 448-49. The same result might be expected if the defendant can match the winning bid.

On the other hand, bid matching can significantly reduce transaction costs. Thus, before holding the auction, the court must weigh the tendency of bid matching to suppress other bids against the potential for bid matching to reduce litigation costs, and then determine whether bid matching is appropriate.

(74.) See Yaad Rotem & Omer Dekel, The Bankruptcy Auction as a Game--Designing an Optimal Auction in Bankruptcy, 32 Rev. LITIG. 323, 354-57 (2013) (discussing all four forms of auction, as well as hybrid auctions and other forms).

(75.) See id. at 358-70 (comparing the processes and their relative merits and weaknesses).

(76.) Cf. id. at 373 ("[A] non-disclosure policy regarding the reserve price is understandable and justifiable because disclosure reduces the number of auction participants by eliminating all of the potential bidders whose valuations are lower than the reserve price.").

(77.) See Michael H. Rothkopf et al., Why Are Vickrey Auctions Rare?, 98 J. POL. ECON. 94, 99 (1990) (stating that with risk-averse bidders, "the bid taker can expect more revenue with a first-price auction than with a second-price auction"). The second-price auction is also subject to the risk that the class, in collusion with some other entity, could fraudulently place a bid slightly below the winning bid, thus forcing the winning bidder to pay a higher price than justified. See id. at 102 (noting that "actual cheating" is not required, but that "the mere fear of it ... will suffice"). In addition, a second-price auction provides additional information to the defendant about the actual value that the winning bidder assigns to the claim--information that may have undesirable effects on the shape of subsequent negotiations between the winning bidder and the defendant, as well as on negotiations between the winning bidder and its counsel, experts, and other consultants. See id. at 102-03.

(78.) In an English auction, prospective bidders can conspire to limit bidding and split any gain from obtaining the item. Because the auction is conducted in public, the members of the conspiracy can monitor each other's behavior and better enforce the collusive pre-auction agreement. With sealed bids, conspirators cannot be sure of each other's behavior. See Rotem & Dekel, supra note 74, at 366-67.

(79.) See id. at 363-66 (describing circumstances in which an English auction would be preferable to a sealed-bid auction).

(80.) See John G. Riley & William F. Samuelson, Optimal Auctions, 71 Am. Econ. Rev. 381, 382 (1981) ("[F]or a broad family of auction rules, expected seller revenue is maximized using either of the two common auctions if the seller announces that he will not accept bids below some appropriately chosen minimum or 'reserve' price.").

(81.) See Macey & Miller, supra note 1, at 111; Thomas & Hansen, supra note 47, at 450-53.

(82.) See, e.g., Macey & Miller, supra note 52, at 460-61.

(83.) Because a successful bidder will be required to compensate class counsel for the fair value of this work, there is no free-rider problem. See infra Part III.F.

(84.) For two settlement class actions, both of which were ultimately held not to comport with the terms of Federal Rule of Civil Procedure 23, see Ortiz v. Fibreboard Corp., 527 U.S. 815, 864 (1999) (rejecting a mass-tort settlement class action under Rule 23(b)(1)(B)); Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 622 (1997) (rejecting a mass-tort settlement class action under Rule 23(b)(3)).

(85.) In government contracting, a "responsible bidder" is one with "[t]he ability ... to properly perform contract work." News Printing Co. v. United States, 46 Fed. Cl. 740, 745 (2000). Relevant factors include "financial strength, prior track record, and organizational skills," and other criteria specified in the bid. Id. at 746. In settlement auctions, financial strength--the capacity to make good on the bid--will be a

paramount consideration, but other factors may also enter the picture. To ensure performance, a court may require the highest bidder to post a bond before award, although the court should weigh the costliness of the bond against the risk of non-performance.

(86.) Recall that, as a rule, a class member can opt out when she can show that her expected gain from proceeding outside of the class action is greater than her loss to the class. See supra notes 56-59 and accompanying text. If a class member can show, for instance, that the defendant would have been willing to pay $100 for her claim, but the bid would provide her with only $75, she has made a prima facie case for opting out--especially if enough other class members in a similar position signal the same desire so that a new class action composed of these opt-outs is a possibility.

(87.) A Pareto improvement occurs when at least one person is made better off by an action and no one is made worse off. See POSNER, supra note 21, at 12.

(88.) For instance, assume that a class of 90,000 members involves only negative-value claims. Further assume that one-third of the class (30,000 members) have weak low-value claims, one-third of the class have strong moderate-value claims, and one-third of the class have strong high-value claims. The original settlement called for a total payment of $9 million, with $0 going to the weak low-value claims, $100 to the moderate claims, and $200 to the higher-value claims. If the bidding process results in a winning bid of $13.5 million, or fifty percent more than the original settlement offer, each class member would be entitled as a rule to a payment fifty percent larger--or $0 to the low-value claims, $150 to the moderate-value claims, and $300 to the high-value claims. Such a bid would meet the three bidding criteria. But the bid may in its terms specify a different distribution, perhaps $50 to the weak claims with the remaining $3 million in excess funds distributed proportionally between the moderate-value and high-value claims. This bid also meets the three criteria.

(89.) In unique circumstances the defendant may agree to settle each claim for a specified amount--perhaps $25 for the first legal theory and $75 for the second. Here, partial bidding on one claim is possible. But such settlements are rare.

(90.) Such settlements are not uncommon in practice, and put enormous ethical pressure on counsel to determine the allocation of funds among the clients. See John C. Coffee, Jr., Conflicts, Consent, and Allocation After Amchem Products--Or, Why Attorneys Still Need Consent to Give Away Their Clients' Money, 84 Va. L. Rev. 1541, 1542, 1544-50 (1998) (demonstrating that conflicts of interest exist at the stage of allocating settlement funds among class members).

(91.) See Macey & Miller, supra note 1, at 25.

(92.) The defendant's only distributional concern is to make sure that the distribution is not so unfair that it is likely to invoke a large number of opt-outs or objections, which might cause the court to disapprove the settlement. Cf. Eisenberg & Miller, supra note 65, at 1563 (discussing whether a low rate of opting out acts as an indicator of a strong settlement and concluding that "the signal is a weak one at best").

(93.) See, e.g., Ass'n for Disabled Ams., Inc., v. Amoco Oil Co., 211 F.R.D. 457 (S.D. Fla. 2002).

(94.) See, e.g., Sullivan v. DB Invs., Inc., 667 F.3d 273, 287 (2011).

(95.) Cf. Nanette L. Stasko, Comment, Competitive Bidding in the Courthouse: In re Oracle Securities Litigation, 59 BROOK. L. REV. 1667, 1697 (1994) (noting the difficulties of adapting the lead counsel auction to cases seeking equitable relief).

(96.) See discussion supra Part III.B.2.

(97.) Such an attempt is hardly fanciful. In present class action practice, counsel sometimes agrees to a combination of injunctive and monetary relief, and then justifies a fee that constitutes a significant percentage of the monetary relief by arguing that the injunctive relief is also valuable to the class. See, e.g., In re LivingSocial Mktg. & Sales Practice Litig., 298 F.R.D. 1, 15-22 (D.D.C. 2013) (rejecting counsel's effort to include the value of the settlement's injunctive relief--allegedly worth $54 million--in considering the proper attorneys' fee, and instead awarding a fee based only on the $7.5 million in monetary relief obtained for the class); id. at 16 ("[T]he value of injunctive relief can be so difficult to quantify.").

(98.) One example is an injunction to establish medical clinics to monitor the health of class members exposed to the defendant's product. In theory, a successful bidder could establish and operate the same clinics.

(99.) See supra Part III.B.2.

(100.) Similarly, if the trial results in a $750,000 judgment, the issue is whether the defendant can recoup the excess $250,000 it paid.

(101.) See Thomas & Hansen, supra note 47, at 451 ("[W]ith more bidders, the higher the expected price and the lower the bidders' expected profits.").

(102.) See supra note 73 and accompanying text.

(103.) See Thomas & Hansen, supra note 47, at 448-49.

(104.) Cf. Cal. Motor Transp. Co. v. Trucking Unlimited, 404 U.S. 508, 510-11 (1972) ("We conclude that it would be destructive of rights of association and of petition to hold that groups with common interests may not, without violating the antitrust laws, use the channels and procedures of state and federal agencies and courts to advocate their causes and points of view respecting resolution of their business and economic interests vis-a-vis their competitors.").

(105.) A common criticism of class actions is their potential to bankrupt the defendant. See, e.g., In re Rhone-Poulenc Rorer, Inc., 51 F.3d 1293, 1298-300 (7th Cir. 1995). Although little evidence exists that class actions presently have this effect, allowing competitors to bid increases this potential because competitors would have an incentive to impose a crushing liability.

(106.) As Judge Posner noted in Grip-Pak, Inc. v. Illinois Tool Works, Inc., 694 F.2d 466, 472 (7th Cir. 1982) (citation omitted):

   The existence of a tort of abuse of process shows that it has long
   been thought that litigation could be used for improper purposes
   even when there is probable cause for the litigation; and if the
   improper purpose is to use litigation as a tool for suppressing
   competition in its antitrust sense, ... it becomes a matter of
   antitrust concern. This is not to say that litigation is actionable
   under the antitrust laws merely because the plaintiff is trying to
   get a monopoly. He is entitled to pursue such a goal through lawful
   means, including litigation against competitors. The line is
   crossed when his purpose is not to win a favorable judgment against
   a competitor but to harass him, and deter others, by the process
   itself--regardless of outcome--of litigating.


(107.) See id. (noting that "the line between effective and abusive resort to legal remedies is indistinct" and that "the evidentiary problems of disentangling real from professed motives would be acute").

(108.) In limited instances allowing a competitor to bid is appropriate. For instance, as we have seen, the auction proposal should not generally be used in cases involving injunctive relief. See supra Part III.B.3. In cases in which a competitor could provide better relief than the defendant would be required to provide by injunction, and could then charge the defendant for the cost of providing this relief, a court may consider a bid from a competitor

as long as the competitor's legitimate motive was clear and the provided relief was proportional to the case. Even here, though, the risk of a competitor providing "Cadillac relief' that pushes a competitor into financial difficulty raises concerns. I thank Simona Grossi for pointing out the prospect of using competitors in this situation.

When a court forbids competitors to bid, it also must take steps to ensure that the source of funding for bids from other parties, such as consumer advocacy organizations, private investors, or law firms, does not derive from one or more of the defendant's competitors.

(109.) See Hughes v. Kore of Ind. Enter., Inc., 731 F.3d 672, 678 (7th Cir. 2013).

(110.) See, e.g., In re Domestic Air Transp. Antitrust Litig., 148 F.R.D. 297, 359 (N.D. Ga. 1993) (describing consumer organizations whose work on behalf of objectors strengthened a class settlement).

(111.) See Molot, supra note 3, at 73 ("[B]y enlisting the help of a third party that holds a diverse portfolio of litigation risk and is better able to bear the risk, the weaker [i.e., risk-averse] party could bolster its negotiating position and secure a settlement that reflects the merits of the lawsuit rather than the relative bargaining positions of the parties.").

(112.) For a discussion of the issues surrounding third-party financing of litigation, see generally Huang, supra note 3; Molot, supra note 3.

(113.) Competing class actions are common. Rhonda Wasserman, Dueling Class Actions, 80 B.U. L. REV. 461, 462 (2000).

(114.) When related class actions are pending in different federal districts, they are often consolidated in one district under the multidistrict transfer power of 28 U.S.C. [section] 1407. See, e.g., In re: Toys "R" Us--Del., Inc., Fair & Accurate Credit Transactions Act (FACTA) Litig., 581 F. Supp. 2d 1377, 1377 (J.P.M.L. 2008) (ordering MDL consolidation when, among other things, "there is a risk of inconsistent rulings on class certification"). After consolidation, the MDL judge often chooses one counsel to lead the litigation, thus cutting out other putative class counsel. See MANUAL FOR COMPLEX LITIGATION (FOURTH), supra note 20, [section] 10.22 (2004) (discussing the power of a federal judge in complex litigation to appoint lead counsel); id. [section] 10.224 (noting the "often intense competition for appointment by the court").

(115.) See infra note 136 and accompanying text.

(116.) Cf. Grip-Pak, Inc. v. 111. Tool Works, Inc., 694 F.2d 466, 472 (7th Cir. 1982) (noting that some scholars regard the use of litigation to achieve anticompetitive ends as "fanciful," especially when litigation involves only a single lawsuit).

(117.) See Huang, supra note 3, at 527.

(118.) See In re Agent Orange Prod. Liab. Litig., 818 F.2d 216, 225-26 (2d Cir. 1987) (discussing the conflicts of interest that can arise when law firms enter into fee-sharing agreements in which one firm provides financing in return for a share of the fees).

(119.) See Molot, supra note 3, at 112 (" [Transaction costs may lead funders to offer plaintiffs modest amounts for their claims, in the hope of turning around and settling with the defendant for more, ... [a result that] would undermine one of the central goals of a litigation risk-transfer system: just compensation.").

(120.) See Huang, supra note 3, at 527 (explaining principal-agent relationship concerns).

(121.) See Macey & Miller, supra note 1, at 108-09 (describing the benefits associated with a single entity's ownership of class members' claims).

(122.) See supra Part III.B.1.

(123.) See S. REP. No. 109-14, at 4 (2005), reprinted in 2005 U.S.C.C.A.N. 3, 5-6 ("[S]tate court judges are readily approving class action settlements that offer little--if any--meaningful recovery to the class members and simply transfer money from corporations to class counsel."); id. at 14-20, reprinted in 2005 U.S.C.C.A.N. at 15-20 (providing examples of cases in which counsel appeared to receive much of the value of the settlement).

(124.) See id. at 20-21, reprinted in 2005 U.S.C.C.A.N. at 21 (arguing that the size of a potential class judgment provides "leverage [that] can essentially force corporate defendants to pay ransom to class attorneys by settling--rather than litigating--frivolous lawsuits").

(125.) See, e.g., Ellen M. Ryan & Laura E. Simmons, Cornerstone Research, Securities Class Action Settlements: 2012 Review and Analysis (2013), available at http://perma.cc/WBQ2-DVS3 (noting that, among securities-fraud cases settled between 2003 and 2012, median settlements ranged from 1.6 percent to 3.3 percent of estimated damages).

(126.) See supra Part III.A.

(127.) See supra note 68 and accompanying text.

(128.) See supra Part II.

(129.) See, e.g., Thomas E. Willging & Shannon R. Wheatman, Attorney Choice of Forum in Class Action Litigation: What Difference Does It Make?, 81 NOTRE Dame L. REV. 591 (2006) (providing data as to settlement amounts).

(130.) For class actions removed to federal court and then remanded back to state court, one study found that the median settlement or recovery was $850,000, or $350 per class member. Id. at 639 tbl.15. The median settlement or recovery for class actions remaining in federal court was $300,000, or $517 per class member. Id.

(131.) Cf. Adam J. Levitin, The Paper Chase: Securitization, Foreclosure, and the Uncertainty of Mortgage Title, 63 DUKE L.J. 637, 670-76 (2013) (discussing the process of securitization in mortgage markets). The recent financial crisis, which was caused largely by the collapse of mortgage-backed securities, gave securitization a black eye. But securitization has significant benefits, including the ability to raise more funds. See id. at 671. Nor would the amount of money at stake in securitized lawsuits approach a level likely to cause widespread financial collapse. For the story of one case in which a bank holding a valuable legal claim securitized the claim and even traded it on a market, see Benjamin C. Esty, The Information Content of Litigation Participation Securities: The Case of CalFed Bancorp, 60 J. FIN. ECON. 371, 372-74 (2001). In this instance, a bank's value increased by 4.3 percent after securitization, likely because the bank's ability to securitize and value its risky asset made it a more valuable takeover target.

(132.) See Molot, supra note 3, at 96 (noting that investments in litigation are attractive to hedge funds because litigation outcomes do not depend on market fluctuations, as a result of which "hedge funds seeking a diverse portfolio of investments have tried to earn returns by betting on litigation").

(133.) Fed. R. Civ. P. 23(e)(2).

(134.) See supra notes 51, 56-58 and accompanying text.

(135.) See text accompanying supra note 83 and accompanying text (discussing pre-auction discovery).

(136.) As a rule, lawyers are entitled to a fee based on quantum meruit when their representation falls below ethical or professional standards--thus disentitling them to their full fee--or their representation is properly terminated. See RESTATEMENT (THIRD) OF THE LAW GOVERNING Lawyers [section] 37 (2000) ("A lawyer engaging in clear and serious violation of duty to a client may be required to forfeit some or all of the lawyer's compensation for the matter."); id. [section] 37 cmt. d (suggesting that a quantum meruit recovery for the lawyer might be appropriate); id. [section] 40 (suggesting a fee of "the lesser of the fair market value of the lawyer's services" or "the ratable proportion of the compensation" under the lawyer's contract when the attorney-client relationship was terminated). If the actions of counsel amount to collusion with the defendant, even this fee may be too high and a complete forfeiture of class counsel's fee may be appropriate. See id. [section] 37 cmt. b (discussing the principle of agency that "willful and deliberate breach disentitles agent to recover in quantum meruit"). I assume in this discussion that class counsel's conduct does not merit a complete forfeiture of the fee.

(137.) For a discussion of some of the ways in which bidders may use the bidding process to harm the defendant, see supra Part III.C.

(138.) The ground of incompetence is a slippery one. As in a legal malpractice case, the theory is that incompetence resulted in a worse outcome than competent lawyering would have achieved. To make this argument, class counsel may need to concede their own incompetence (in other words, admit that competent lawyering by the bidder's counsel would have resulted in a greater award than the award class counsel achieved). In such a circumstance, a substantial limit on the recoverable fee is likely.

Jay Tidmarsh, Professor of Law, Notre Dame Law School. I thank Clarence Wilson for research assistance and helpful conversations. I also thank Simone Grossi, Aaron Kaplan, Geoff Miller, David Rosenberg, Georgene Vairo, Adam Zimmerman, and the participants in a workshop at Loyola Law School for their insightful comments.
COPYRIGHT 2014 College of William and Mary, Marshall Wythe School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Tidmarsh, Jay
Publication:William and Mary Law Review
Date:Oct 1, 2014
Words:19487
Previous Article:Incapacitating the state.
Next Article:Tortured prosecuting: closing the gap in Virginia's criminal code by adding a torture statute.
Topics:

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |