CLASS ACTIONS – REPRESENTATIVE PROCEEDINGS
Cecil D. Redford Professor
University of Texas Law School
© Copyright 1999 Charles Silver
A class action is a representational lawsuit that binds all persons with
related claims, including those who were not parties to the proceeding. The
procedure enables parties and court systems to enjoy certain benefits that are
lost when claims are heard separately. These include reduced litigation costs
per claimant, cost-internalization, and improved litigation incentives. The
procedure generates these benefits by eliminating the need for collective
action and giving class counsel control over claims held by non-parties.
Because market discipline is lacking and judges regulate class actions
inappropriately, this transfer of control entails serious risks of agency failure
and opportunism that are often dealt with inadequately. The contemporary
challenge is to reform the procedure in ways that preserve its advantages
while reducing its risks and costs.
JEL classification: K41
Keywords: Class Action, Consolidation, Representative Litigation,
Collective Action, Agency Costs
Most litigation is representative in nature because it is handled by agents,
typically though not always attorneys, who advocate and otherwise act on the
parties’ behalf. Ordinarily, parties hire their agents directly, using market
pressures and contracts to encourage good performance and to mitigate the
risks and costs that are predicted to beset any relationship of principal and
agent (Stiglitz, 1987).
In a percentage of the cases, judges appoint attorneys in the absence of
market transactions. This happens mainly when a party is a child, an
incompetent, or a person in need of charity. It also happens in certain
situations where bargaining impediments prevent lawyers and clients from
coming to terms despite perceived opportunities for mutual gain or systemic
needs to aggregate claims. The class action is the paradigm of judicial
appointment of representation in this situation.
Numerically, class actions make up a small part of an average court’s
civil docket. Of the 272,027 civil lawsuits filed in all federal district courts
in the United States in 1997, only 1,475 or one-half of one percent were
class actions. If all 647 authorized judgeships had been filled, federal district
court judges would have received an average of only two new class actions
apiece in 1997 (Administrative Office of the United States Courts, 1997).
Though comparatively rare, class actions are disproportionately
important. They consume far more of judges’ time than other civil suits.
Judges have been estimated to spend eleven times as many hours on certified
class actions and twice as many hours on non-certified class actions as they
do on non-class civil cases (Willging, Hooper and Niemic, 1996). Because
class actions also involve far more claimants than other civil suits, these
judicial investments can yield enormous economies of scale. Class actions
brought to remedy economic harms often encompass thousands or even
millions of consumers. Personal injury class actions can also be quite large,
especially those that attempt to resolve claims held by persons who may
become ill in the future. When judges’ time is amortized across claimants,
the economy of the class action is revealed (Bernstein, 1978).
Class actions also involve unusually large financial stakes. A study of
securities class actions reported that the 105 lawsuits in its sample for 1995
generated settlements averaging US$10,619,174 and totaling
US$1,115,013,250 (Martin et al., 1996). Many personal injury class actions
have generated settlement funds or settlement offers of US$100,000,000 or
more. In 1998, a single antitrust class action yielded a proposed settlement
in excess of US$1,000,000,000.
Not all class actions seek money damages. Many are brought to achieve
social, political, or legal reforms. Targets of reform lawsuits include mainly
public and private institutions, such as school districts, large employers,
manufacturing plants, welfare agencies, prisons and legislatures, whose
operation affects large numbers of individuals.
Given their potential to transfer massive amounts of wealth and to
reorganize important institutions, it should surprise no one that class actions
are politically controversial. Although the federal rule that governs the
conduct of these lawsuits in the United States retains the basic structure it
received in 1966, there have been many attempts to reform the rule and to
reduce the rate at which class actions are brought. Statutes now restrict
prisoner class actions, securities cases, consumer cases and lawsuits by legal
aid organizations that receive public support. Federal common law also
deals with matters like subject matter jurisdiction, notice, and attorneys’ fees
in ways that make class actions more difficult to bring.
Reflecting the importance of class actions, the scholarly literature
discussing them is enormous. The history of the procedure, which has its
roots in England, its trunk in the United States, and new branches in other
countries, has been carefully studied (Yeazell, 1987). Developments in the
case law are followed closely. A large and growing empirical literature
studies trends in filings, dismissals, settlements, and fee awards (Alexander,
1991; Garth, Nagel and Plager, 1988; Hensler et al., 1985; Martin et al.,
1996; Rosenfield, 1976; Willging, Hooper and Niemic, 1996b). The impact
of class actions on issuers of securities and securities markets has also been
studied (Strahan, 1998; Bohn and Choi, 1996). Finally, a literature
emphasizing microeconomics and game theory considers normative
justifications for class actions, examines agency problems that arise in them,
and suggests avenues for reform (Coffee, 1987b, 1995; Hay, 1997a; Macey
and Miller, 1991; Perino, 1997; Rosenberg, 1984, 1987).
A. Legal Basics: What a Class Action is and How One Works
2. The Class Action is a Representational Device, Not a Joinder Device
The class action is a procedural device that expands a court’s jurisdiction,
empowering it to enter a judgment that is binding upon everyone with
covered claims. This includes claimants who, not being named as parties,
would not ordinarily be bound. A class-wide judgment extinguishes the
claims of all persons meeting the class definition rather than just those of
named parties and persons in privity with them, as normally is the case.
Judges and scholars sometimes treat the class action as a procedure for
joining absent claimants to a lawsuit rather than as one that permits a court
to treat a named party as standing in judgment on behalf of them. This is a
mistake (Hutchinson, 1983). Class members neither start out as parties nor
become parties when a class is certified.
3. How a Class Action Commences
A named party begins the process of having a class recognized by filing a
pleading in which a class is preliminarily described. Later, a named party
files a motion for certification. The motion refines the class definition and
requests entry of an order allowing a class action to proceed. In theory, a
named plaintiff can ask to have a class of plaintiffs or of defendants
certified, and a named defendant has the same two options. In practice,
requests by named plaintiffs to certify plaintiff classes greatly outnumber all
other requests. Defendant classes comprise fewer than 1 percent of all
certified class actions, and named defendants rarely ask judges to certify
classes of any kind (Willging, Hoper and Niemic, 1996b).
A named plaintiff who seeks to stand in judgment for a class is called a
putative or actual class representative, depending upon whether the motion
for certification has been granted. An attorney appointed to act for a class is
called class counsel. Usually, the attorney representing the named plaintiff is
appointed class counsel in the order that certifies the class. A person who is
not a named party but who falls within a class definition is called an absent
class member or an absent plaintiff.
4. The Certification Decision
For a lawsuit to proceed on behalf of a class, a judge must certify that certain
criteria are met. The named plaintiff must be a member of the class. It must
be impracticable for the named plaintiff to join all absent plaintiffs to the
lawsuit as named parties. Common questions of law or fact must connect the
named plaintiff’s claims with those of the absent plaintiffs. The named
plaintiff must also be a typical claimant who, having no significant conflicts
of interests with the absent plaintiffs and being assisted by competent
counsel, can be relied upon to represent adequately all members of the class.
These criteria are repetitive and overlapping.
At least one of the following additional criteria must also be met to
secure class certification. There must be a risk that an opposing party will be
subject to multiple, duplicative, or inconsistent obligations unless a class is
certified. There must be a chance that an absent claimant’s ability to protect
his interests will be impaired unless a class is certified. The defendant must
have subjected all class members to a common practice or course of conduct,
justifying a class-wide grant of declaratory or injunctive relief. Or, questions
of law or fact common to all class members’ claims must predominate over
any individual questions that may arise and the class action must be superior
to other available procedures for handling the block of complaints.
Judges certify common question class actions far more often than they
find any other criteria apply. Approximately two-thirds of all certified class
actions fall into the common question category (Willging, Hooper and
Niemic, 1996b), which is used for all manner of economic harm cases,
including securities and insurance fraud and other deceptive trade practices.
These cases typically involve large numbers of claimants who incurred small
losses and require mainly evidence about the conduct of defendants
The law encourages judges to decide certification motions ‘as soon as
practicable’ (Federal Judicial Center, 1995). Even so, judges usually decide
dispositive substantive motions first. A plaintiff who prevails on a
substantive motion will find the chances for certification greatly improved
(Willging, Hooper and Niemic, 1996b). Judges also frequently consider class
certification and settlement at the same time. This occurs in about one-third
of all lawsuits where classes are certified (Willging, Hooper and Niemic,
5. Post-Certification Procedures
Certification usually triggers additional procedures that are specific to class
actions. In common question cases, absent plaintiffs are entitled to notice of
the lawsuit and to an opportunity to exclude themselves from it. They also
may appear in the lawsuit if they wish. When certification and settlement
occur concurrently, absent plaintiffs are entitled to notice, to a complete
description of the settlement including attorneys’ fees and expenses that may
be paid, to file objections, and usually to exclude themselves. Any settlement
that is proposed also must be judicially approved for a judgment to bind all
A class action that has been certified for all purposes proceeds toward
trial much as any other lawsuit does, and trial rates for certified class actions
are comparable to those for other civil suits (Willging, Hooper and Niemic,
1996). The named plaintiff uses ordinary discovery tools to gather
information about the merits of class members’ claims. Admissions and
dispositive motions are used to remove factual and legal issues from dispute
and to identify any remaining factual issues that must be tried. At the trial
itself, the named plaintiff must offer evidence sufficient to prove all elements
of his individual claim and to support favorable findings on all class-wide
issues, including damages (Issacharoff, 1991). The stakes being larger in
class actions than conventional lawsuits, the parties devote greater effort to
these matters and incur greater expense, but the differences are matters of
degree not of kind.
Settlement rates for certified class actions also are generally comparable
to those for conventional lawsuits, with securities class actions settling more
frequently than class actions involving other complaints (Martin et al., 1996;
Willging, Hooper and Niemic, 1996b) . Empirical studies have not borne out
the charge, often heard in policy discussions, that a judge’s decision to
certify a class forces even an innocent defendant to surrender (Priest, 1997).
As explained above, judges must review and approve class action
settlements after giving absent plaintiffs notice and an opportunity to object.
Historically, objections have been infrequent. Few absent plaintiffs take the
trouble to attend fairness hearings. More file written objections, but the
number who write in also is small. As might be predicted, these meager
efforts rarely prevent settlements from proceeding. In the past, judges
approved without alteration 90 percent of all proposed class settlements to
which objections were filed (Willging, Hooper and Niemic, (1996b).
Recently, certain abusive or allegedly collusive class action settlements
have received widespread press coverage and extensive academic scrutiny
(Coffee, 1995a; In Camera, May-June 1993, July-August 1993; Koniak,
1995; Koniak and Cohen, 1996). As a result, judges, who are charged to act
as guardians for absent class members, have begun to examine settlements
more closely. They have paid special attention to proposed settlements that
resolve immature claims of personal injury, that offer class members
coupons or discounts instead of cash, or that abridge the right of dissenting
class members to opt out. Although these settlements are not impermissible
per se (Miller and Singer, 1998), they often signal that class counsel and a
settling defendant colluded to the detriment of a class.
B. The Basic Microstructure of the Class Action
6. Separation of Ownership from Control
Class actions contain far more class members than class representatives.
Consequently, most of the economic value of a class action is attributable to
absent plaintiffs’ claims. If a single class representative were to sue on
behalf of ninety-nine absent plaintiffs with identical claims, the absent
plaintiffs collectively would account for 99 percent of the value of the case.
Because representatives with minority interests manage class actions, it
is useful to think of classes as litigation groups in which ownership and
control of assets are in different hands (Jensen and Meckling, 1976). In this
respect, classes resemble stock companies, mutual funds, and innumerable
other economic undertakings in which investors play relatively passive roles
(Coffee, 1998; Shapiro, 1998; Silver and Baker, 1998). Seeing this
resemblance, judges have subjected named plaintiffs to fiduciary duties like
those that customarily apply to managers of other enterprises.
Judges have also subjected class counsel to fiduciary duties. From an
economic perspective, this makes sense. The regulation and incentivizing of
counsel affects the likely success of class actions far more than anything
having to do with named plaintiffs. As explained, a named plaintiff holds a
tiny fraction of the entire economic interest possessed by a class. Often, a
named plaintiff’s claim is far too small to justify the expense of litigation,
and it is always too small to motivate a named plaintiff to invest resources at
a level that is optimal for a class. By comparison, class counsel’s interest is
usually far larger. Counsel is paid on a contingent basis an amount that
reflects the value of the relief recovered by the entire class. In most cases,
the fee is a fraction - between 20 and 40 percent - of the common fund
(Lynk, 1990, 1994; Martin et al., 1996; Willging, Hooper and Niemic,
1996). Because class counsel holds the largest single stake, it is not
surprising that class counsel is the primary decision-maker and the moving
economic force in class litigation (Kane, 1987; Miller, 1998).
The concentration in class counsel’s hands of economic interest and
managerial control accounts both for the attractiveness of class action
lawsuits and the risks associated with them. The attraction derives from the
possibility of placing a large block of claims, including many that may be
too small individually to warrant conventional lawsuits, under the control of
someone who is in a position to maximize their value. The risks derive from
the possibility that control will be abused for the benefit of the agent in
7. Comparing Class Actions and Other Group Lawsuits
Class actions are involuntary or nonconsensual group lawsuits. They begin
and, for practical purposes, end without class members’ consent.
Other group lawsuits are consensual. They involve multiple clients all of
whom contract for representation with the same attorneys. Small voluntary
groups greatly outnumber large ones, but entrepreneurial attorneys have
built enormous groups with tens of thousands of plaintiffs by using
mass-marketing techniques and by actively working referral markets
(Resnik, Curtis and Hensler, 1996; Silver and Baker, 1997; Spurr, 1988).
Entrepreneurial lawyers play the same role in large voluntary group lawsuits
that organizers play in other collective actions (Hardin, 1982). They also use
the same techniques as organizers to reduce their costs. For example, they
use existing associations, such as labor unions, trade associations, and
homeowner groups, to recruit individual plaintiffs and to serve as
Still other group lawsuits straddle the consensual/non-consensual divide.
These lawsuits are called consolidations. Consolidation occurs when
pending cases are routed to a single court for coordinated management.
Usually, this occurs when lawsuits involve common facts. For example,
claims arising out of airplane crashes, explosions, and other disasters are
frequently consolidated. The perception is that such cases can be processed
efficiently in tandem (Steinman, 1995a, 1995b).
Consolidations begin consensually. Plaintiffs employ attorneys and
authorize them to file lawsuits. Thereafter, plaintiffs lose control. Judges
frequently consolidate cases at defendants’ request and over plaintiffs’
objections. In effect, they force together plaintiffs who would rather remain
apart. In federal multi-district litigation, a consolidation order may transfer
cases to distant jurisdictions where plaintiffs’ chosen lawyers rarely practice.
Judges then subject transferred cases to the control of managing committees.
These consist of several lawyers who are selected to handle most tasks for all
plaintiffs, including the preparation of master complaints and motions,
discovery and settlement negotiations. These lawyers receive supplemental
compensation, usually as a result of judicial orders entitling them to shares
of other lawyers’ fees (Resnik, Curtis and Hensler, 1996; Silver, 1991b).
From an economic perspective, all three kinds of group lawsuit can be
analyzed in similar terms. In all, control is separated from ownership and
placed in the hands of lawyers with large financial interests. From a
practical perspective, however, the three cases differ. When litigation groups
form voluntarily, plaintiffs use contracts to govern them. They select
attorneys, set their compensation, allocate responsibility for expenses,
establish procedures to govern individual and group-wide decision making,
create incentives to keep the collective action going, and otherwise take care
of the group’s constitutional needs. They also waive conflicts of interests,
retain withdrawal rights, and can fire attorneys who perform poorly (Silver
and Baker, 1997). By contrast, class actions and consolidations are created
by regulation. Judges form them and decide how they will operate. Class
actions and consolidations thus operate outside the reach of market forces.
Absent plaintiffs cannot fire their appointed representatives when they are
dissatisfied with their performance. They cannot sell their interests or join
new groups when they disagree with fundamental strategies or decisions, as
shareholders in publicly traded corporations can. Nor can they rely on
anything analogous to the takeover market to police inefficient management
of their claims. Consequently, agency problems can be far more serious in
class actions and consolidations than in voluntary group suits (Silver and
Baker, 1998; Silver, 1991b).
C. Advantages of Aggregation
Litigation groups offer plaintiffs several important advantages. These
include economies of scale in litigation costs, greater leverage in settlement
negotiations, and conservation of defendants’ assets, all of which can
increase plaintiffs’ recoveries.
8. Economies of Scale
Group lawsuits enable plaintiffs to reduce litigation costs per capita by
sharing the burden of litigation services that can be used by many plaintiffs.
Plaintiffs who sue individually must separately pay for legal and factual
investigations, for representation at trial and in settlement negotiations, and
for litigation support services such as document preparation, expert witness
testimony, and computer simulations. A group of plaintiffs need pay for
these services only once, the services being non-rival goods insofar as
members of the group are concerned. Plaintiffs who sue together also gain
by taking advantage of the specialization of labor when performing tasks
such as monitoring counsel’s performance, answering interrogatories, and
attending hearings and depositions. By assigning these tasks to claimants
who can best handle them, performance can be improved and savings
generated that make all plaintiffs better off than they would be if each were
responsible for performing every task.
9. Enhanced Bargaining Leverage
Most group lawsuits settle. Bargaining power is therefore as important in
group lawsuits as in conventional cases. In theory, plaintiffs can often gain
leverage in settlement negotiations by joining forces. This happens, most
straightforwardly, because groups of plaintiffs will often find it rational to
litigate more intensively than individuals.
The impact of economies of scale on bargaining power is clearest when
individual claims are too small to justify the expense of litigation. In this
situation, claims pressed by individual plaintiffs have little settlement value
because plaintiffs cannot make credible threats to try their cases. By
comparison, the threat value, and therefore the settlement value, of groups of
small claims may be relatively great, owing to the feasibility of trying small
claims en masse.
Economies of scale can increase the settlement value of large claims too.
On the standard economic model of the decision to settle or sue, a plaintiff’s
minimum demand falls as litigation costs rise. Because aggregation reduces
plaintiffs’ litigation costs per capita, it should increase their minimum
demands and, correspondingly, the settlement value of their claims.
Aggregation also can increase plaintiffs’ demands by increasing the
expected value of claims at trial. A claim’s expected value is partly a
function of the probability that a plaintiff will prevail in a trial against a
defendant. In turn, the probability of winning is a function of the level and
quality of litigation support services obtained. If members of plaintiff groups
save money by sharing the cost of non-rival litigation services, they can
afford to purchase more services or services of better quality than plaintiffs
who sue by themselves. They can thereby raise the expected value of their
claims at trial and improve their prospects in settlement. This may explain
why class members have won sizeable settlement payments on claims that
failed to generate payments in conventional cases. The class action can bring
‘compensation and closure’ to persons who are otherwise poorly served by
the tort system (Tidmarsh, 1998).
Even when plaintiffs sue separately, their fortunes often intertwine.
Plaintiffs’ attorneys share information about trials and settlements. Factual
findings, legal rulings, and verdicts in decided cases inform assessments of
the settlement value of cases yet to be resolved (Hensler and Peterson, 1993).
Predictably, these connections between separate lawsuits give defendants an
edge in litigation. A defendant facing a test case knows that the stakes are
greater than just the amount likely to be won by the litigating plaintiff. The
amount paid in settlement or lost at the trial of the test case will affect the
value of claims held by other plaintiffs who are waiting in the wings. The
defendant may therefore find it rational to invest far more in the lawsuit
than the litigating plaintiff’s claim alone would warrant. Manufacturers of
tobacco products maintained an unbroken record of success in personal
injury litigation for four decades by outspending the opposition.
When claim values intertwine, a defendant almost certainly will find it
economically rational to outspend an attorney who represents a single client.
Ordinarily, a plaintiff’s attorney’s fee is a percentage of the client’s recovery
Clermont and Currivan (1978). This compensation arrangement gives an
attorney no incentive to spend even an amount equal to the expected value of
the plaintiff’s claim, the expected fee being far smaller. By itself, the
possibility that the trial or settlement of a test case may affect the value of
other plaintiffs’ claims will not induce the attorney to spend more. Unless
other plaintiffs or their attorneys agree to share litigation expenses, these
external effects will be ignored.
Aggregation brings the parties’ stakes more nearly into balance and
improves the incentives of plaintiffs’ counsel as well. When all claimants
participate or are represented in a single lawsuit, the defendant’s exposure is
the sum of the expected values of all the claims (or the defendant’s total
assets, whichever is less). As a group, the plaintiffs’ stake is the same
amount. The investment incentives are still unequal because the attorney
representing the plaintiff group stands to earn only a fraction of the
plaintiffs’ recovery, usually one-third (Martin et al., 1996). Even so, the
plaintiffs’ attorney can credibly threaten to invest significant resources in
litigation services that, from the plaintiffs’ perspective, are joint goods. If, as
seems likely, litigation investments have declining marginal value, the
threat to spend heavily will give the defendant pause.
Group counsel’s predominant financial interest is and must be the
economic force that drives group litigation. The expectancy of receiving a
fee that far exceeds the typical class member’s recovery motivates attorneys
to make the enormous financial commitments these lawsuits require and to
incur the correlative risks. Even so, the size of counsel’s interest has sparked
political controversy. Because counsel’s interest dwarfs that of all but the
largest class members, for example, institutional investors in securities fraud
cases, it often seems that group lawsuits have more to do with paying
attorneys than compensating victims.
From an economic perspective, this imbalance is not a source of concern.
Attorneys who manage large litigation groups are not distinguishable from
managers of widely held companies or mutual funds. It would be foolish to
limit a manager’s compensation to the amount any individual investor earns
in a given period. It would be equally unreasoning to limit class counsel’s
fee to the amount recovered by the largest claimant. Such a limitation would
undermine counsel’s incentive to invest. This would harm group members,
not help them.
Of particular concern to a defendant facing a group lawsuit is the
possibility that a single trial will adjudicate its entire liability. Ordinarily, a
defendant can handle risk far more easily than a plaintiff can. The plaintiff
will have an undiversified risk of loss associated with a claim that may be
his single largest asset. By contrast, the defendant will face hundreds or
thousands of claims, some of which will succeed, some of which will fail,
and none of which taken individually will pose a serious threat to the
defendant’s solvency. In this situation, the defendant can more easily
tolerate the risk of trying cases than can any individual plaintiff.
Aggregation turns the tables on defendants. It puts them in the position of
being rendered insolvent by a single trial (a high variance event), thereby
saddling them with a large, undiversifiable risk. It also forces defendants to
go to trial knowing that plaintiffs’ attorneys have strong incentives to invest.
Many defendants are willing to pay large sums to avoid facing the prospect
of a real trial followed by liquidation.
10. Conservation of Defendants’ Assets
When a defendant’s assets (including insurance) are insufficient to cover all
plaintiffs’ claims, a group lawsuit can make plaintiffs better off by
conserving the defendant’s resources and preserving the going concern value
of the defendant’s operations. When claims are litigated in separate forums,
defendants must retain local counsel in each venue, pay experts to testify
concerning the same scientific issues at each trial, suffer the expense of
repetitive depositions of managerial personnel, and bear other duplicative
costs. These expenditures, which can be reduced when plaintiffs sue in a
single proceeding, consume resources, making them unavailable for transfer
The value of a defendant’s assets also can be diminished when a
defendant’s business is liquidated piecemeal rather than as a unit, so that the
value of the operation as a going concern is lost. Speaking less than strictly,
competing plaintiffs participate in an n-person Prisoner’s Dilemma. If each
strives independently for a share of the defendant’s resources, the
defendant’s operations will be liquidated piecemeal, going-concern value
will be lost, and plaintiffs as a group will net a smaller recovery than they
could have realized by keeping the defendant’s operations intact. Just as the
possibility of avoiding this race to judgment supports some scholarly
justifications of bankruptcy laws (Jackson, 1982), the same prospect may be
said to weigh in favor of group lawsuits in cases involving limited funds.
11. A Paretian Defense of Aggregation
Figure 1 displays the basic economic logic of group litigation. Where X and
Y are potential co-plaintiffs, point A represents their expected joint payoff if
each sues the common defendant in a separate lawsuit. The space defined by
triangle OGH contains all possible joint payoffs if X and Y combine forces
and sue as a group. The smaller triangle AIJ within OGH is the set of
possible payoffs from group litigation that are Pareto Superior to A. Curve
CD contains all possible divisions between X and Y of the expected joint
payoff from group litigation. EF is the set of possible divisions of the
expected joint payoff that are Pareto Superior to A. A simple normative
economic argument for group litigation is that, by joining forces and sharing
costs, co-plaintiffs hope to move themselves from A to a point on EF.
12. Why Require Participation? Transaction Costs and Free-Riding
Figure 1 raises an important question about non-consensual group lawsuits.
Why are they permitted? When plaintiffs stand to gain by joining forces,
they can be expected to form groups voluntarily. Presumptively, a plaintiff’s
refusal to join a litigation group would indicate that, in the plaintiff’s
opinion, any gains that may to flow from cooperation fail to offset the
expected costs and risks. It is therefore reasonable to ask why class suits and
consolidations should ever occur without the consent of absent plaintiffs.
One answer is that class actions and consolidations facilitate mutually
beneficial cooperation that is normally frustrated by the high cost of
transacting. Transaction costs are a serious problem when a wrongdoer’s
conduct inflicts small losses on the members of a large victim population.
For example, suppose that a credit card company overcharges millions of
cardholders a few dollars apiece by calculating interest incorrectly. For
cardholders as a group, the loss is large enough to be worth pursuing, but no
individual cardholder has an incentive to organize the group. The prospect
of recovering a few dollars would not justify the time or effort even if the
likelihood of winning were great.
The same problem arises in injunctive cases where litigation may prevent
small harms from occurring. Suppose that the members of a large population
are exposed to individually small health risks when a toxic substance is
disposed of improperly. A lawsuit brought to force the wrongdoer to clean
up the waste would protect the entire community and be cost-justified in
view of the risk. However, the expected gain to any individual will likely be
too small to motivate anyone to bear the expense of organizing a
Class action rules enable plaintiffs to avoid start-up costs that prevent
litigation groups from forming. A single named plaintiff can initiate a class
action on behalf of all persons with similar claims just by filing an
appropriate complaint. However, in keeping with the belief that
non-consensual lawsuits are permissible only when start-up costs are high,
the rules permit named plaintiffs to use this gambit only when it is
impracticable to join absent plaintiffs as parties voluntarily. This may be
because the group’s members are numerous and geographically dispersed,
because claims are too small to motivate anyone to bear the cost of locating
and bargaining with potential plaintiffs, or because group members
identities are unknown or change over time. By minimizing transaction
costs, class action rules enable named plaintiffs to wage advantageous group
lawsuits that would not otherwise be brought.
Free-riding can also impede the formation of voluntary groups when the
object of litigation is a non-exclusive good, that is, a good that will be
available to group members and non-members alike. Structural reform
lawsuits and other injunctive actions provide the best examples. Structural
reforms, including improvements in public housing conditions, increased
funding for public schools, and changes in hiring and promotional practices
by an employer, automatically benefit everyone who suffered under prior
arrangements. Consequently, potential plaintiffs may find it economically
rational to free-ride on the efforts of others in these cases. If the lawsuit
succeeds, the free-rider obtains the benefit without charge. If the lawsuit
fails, the free-rider foregoes the benefit but avoids litigation costs, including
the risk of being singled out for recrimination or retaliation by people who
oppose the lawsuit. Either way, the free-rider is better off not cooperating.
The exceptional case in which a free-rider’s contribution can make the
difference between the success and failure of a group lawsuit occurs too
rarely to worry about.
Free-riding can also be a rational strategy when plaintiffs seek damages,
even though a money judgment is an exclusive good. Suppose that X, who is
one of one thousand claimants, believes that it is important to preserve the
going-concern value of a defendant’s operations, and that X also believes
that piecemeal liquidation will not occur if any group containing at least 500
claimants forms. These beliefs may lead X to regard his own participation in
a litigation group as inessential and to remain apart from a group if there is
any advantage to doing so. If all claimants think like X, there may be no
litigation group despite the opportunity for collective gain.
Opportunities to gain free information may also cause free-riding.
Suppose that X and Y are both injured by the same defective product, that X
sues the manufacturer, and that X invites Y to join the suit. If Y is confident
that X will prosecute the lawsuit whether or not Y joins, Y may decide to let
X sue alone. X’s lawsuit then will be a valuable source of information for Y.
Whether X wins, recovers a substantial settlement, or loses, Y will be better
able to assess the value of Y’s claim. Y may also be able to take advantage of
favorable factual or legal rulings made in the course of X’s suit, as in times
past the class action device enabled Y to recover under the judgment
obtained by X. The possibility of capitalizing on X’s lawsuit creates
opportunities for Y to gain an advantage by waiting on the sidelines.
13. Non-Paretian Justifications for Non-Consensual Proceedings
One also can argue for non-consensual group lawsuits on a variety of
non-Paretian grounds. An obvious alternative is a Kaldor-Hicks approach on
which the gains of some, be they group members or others, more than offset
any losses some plaintiffs incur. The Kaldor-Hicks standard is often
described as a potential Pareto improvement (Cooter and Ulen, 1988).
An argument of this sort can be grounded in the desire to conserve the
resources of the legal system. The claim here is that one would squander
judges’ time by requiring them to process related claims individually. The
force of this argument is a function of the number of claims and the extent to
which they are related. The larger the number and the closer the factual and
legal connection, the more society can gain through aggregation (Bernstein,
1978). This insight has spurred many efforts to consolidate related tort
lawsuits, such as those arising out of occupational exposure to asbestos
(Hensler et al., 1985).
The judicial efficiency argument has merit when plaintiffs’ claims are
large enough to justify the expense of separate lawsuits. Then, economies of
scale can be gained by using a class action or a consolidation to streamline
the processing of claims. When plaintiffs’ claims are small, however, the
judicial efficiency argument may or may not be persuasive, depending on
how it is framed. The distinctive contemporary function of the class suit is to
facilitate litigation by claimants who hold small shares of a large
controversy, shares that are too small to be litigated individually (Kalven
and Rosenfield, 1941). By performing this service, class actions consume
judicial resources that would otherwise be spared, it being irrational for
individual plaintiffs to sue on their own. A single-minded concern save
judicial resources would therefore cause one to oppose small-claim class
actions. A slightly different and socially more appropriate concern - a desire
that judges’ time be used efficiently would lead to the opposite conclusion.
As explained above, certified class actions consume eleven times as many
hours as conventional lawsuits, but they increase the rate at which judges
can resolve claims hundreds or thousands of times, more than offsetting
their marginally greater cost (Willging, Hooper and Niemic, 1996).
One must also consider other social costs that would be incurred if
wrongs that visit small losses on large numbers of individuals were not
policed. Losses imposed on third parties are externalities that producers will
ignore unless forced to account for them. By bringing widely dispersed
losses home to producers, the class action discourages them from acting in
ways that are socially inefficient. It forces them to consider public as well as
private production costs (Rosenberg, 1984, 1987; Wright, 1969).
The cost-internalization function of the class action is independent of its
compensatory function and may be served when the latter cannot.
Difficulties relating to the identification of class members, the measurement
of their individual damages, or the cost of fine-tuning payments, may make
it impossible or impracticable to restore claimants to the positions they
would have come to but for the wrong. Even so, it may often be possible to
deter wrongdoers by measuring aggregate social costs and imposing them on
malfeasant defendants. It being as important to avoid losses as it is to
compensate victims, an inability to compensate should not by itself prevent a
class action from proceeding. Alternatives such as fluid class recoveries and
escheat of unclaimed funds should instead be explored.
D. A Digression on the Law of Restitution
14. Why Pay Class Counsel?
The preceding sections examined the propriety of drawing plaintiffs into
class actions without their consent. This section considers a related issue:
why require absent plaintiffs to pay attorneys they never agreed to hire? The
answer may seem simple. Unless attorneys are paid to represent absent
plaintiffs, they will not do so and there will be no class actions. However, in
a legal system committed to freedom of contract, more must be said than
this. The usual rule is that people need not pay for services they do not ask to
receive. Restitution is rarely required, especially when a provider acts
pursuant to a contract with another, as an attorney representing a named
plaintiff does. B, who enjoys looking at A’s beautiful yard, has no duty to
compensate A’s gardener. The gardener’s only right is to receive the
contracted for amount from A.
It has been seriously maintained that the practice of allowing class
counsel to recover fees from common funds over and above the amounts they
agree to accept from named plaintiffs violates the law of restitution
(Dawson, 1975). The argument has not persuaded judges to abandon the
practice, which is as entrenched as ever. The desire to cure unjust
enrichment has led judges to permit attorneys to claim compensation in
limited circumstances. Their efforts must confer benefits on non-clients. It
must be infeasible or impracticable for them to bargain with non-clients in
advance. And it must be difficult for them to prevent non-clients from
deriving benefit from their work. These requirements often are met when an
attorney brings a class action to a successful close (Silver, 1991a; see also
The restitutionary approach caps class counsel’s fee at the market rate or
the size of the recovered fund, whichever is less. It allows judges to use any
payment formula employed in the private market, including percentages,
hourly rates, and hybrids of the two. Percentage compensation has long been
the norm in common fund class actions. Federal judges began a shift to the
hourly-rate based lodestar method in the 1970s, but the experience proved
disappointing and most federal courts have shifted back to the percentage
approach. The emerging practice is to base fee awards on an established
benchmark, usually 25 percent of the recovered fund, and to use the lodestar
to make small deviations from this figure in exceptional cases.
E. Disadvantages of Aggregation
15. Paretian Misfires
Group lawsuits often have the potential to make litigants worse off than they
would expect to be suing on their own. In Figure 1, A represents the
expected joint payoff if X and Y wage separate lawsuits. A thus serves as a
boundary for Paretian claims about the affects of group lawsuits, but it is not
a practical boundary of any kind. It does not restrict the actual allocation of
relief. Figure 1 shows this in two ways. First, OGH, the set of possible
outcomes of group litigation, includes points that are Pareto Inferior to A.
The darkened area to the southwest of A is the set of possible Pareto Inferior
payoffs. That such points exist when plaintiffs’ claims have positive
expected values in conventional litigation is not controversial. Second, OGH
also includes many points that are Pareto Noncomparable to A. When a
group secures relief equal to or greater than A, the fruits can be divided
between X and Y in a manner that leaves one of them worse off than he or
she would have been at A. Thus, X is worse off at C than at A, and Y is
worse off at D than at A, even though the expected recovery from group
litigation is obtained at both C and D. The rectangles AKGI and ALHJ are
the sets of possible Pareto Noncomparable results.
The Pareto noncomparable and Pareto inferior outcomes are more than
academically interesting. There is good reason to fear they often occur.
Empirical evidence shows that class actions often settle for amounts that
seem small in relation to the losses claimants incur (Carleton, Weisbach
and Weiss, 1996; Martin et al., 1996). Anecdotal evidence further suggests
that judges frequently approve distribution plans that divide settlement funds
other than in proportion to the size and strength of class members’ claims
(Coffee, 1998, 1987a; Koniak, 1995).
16. Agency Failures and Sub-Optimal Recoveries
Whether the goal is compensation or deterrence, the importance of
regulating class actions in ways that encourage attorneys to maximize the
value of plaintiffs’ claims is clear. If class suits settle too cheaply or
plaintiffs lose at trial more often than they should, wrongdoers will escape
some of the social costs of their behaviors. Undercompensation and
underdeterrence will be the results.
A central thesis in the literature is that class actions often settle for too
little (Coffee, 1986; Hay, 1997a). There also is evidence that defendants win
class-wide trials more often than one would expect. Both results are thought
to be due to agency failures that plague class actions. These failures have
several sources, including defective incentives, inadequate monitoring,
economic asymmetries that favor defendants, lack of market discipline,
certification for settlement of classes that could not be certified for trial, and
conflicts involving settlements of inventories of signed clients.
17. Defective Incentives and Opportunism
The impact of fee arrangements on lawyers’ incentives has been widely
studied (Clermont and Currivan, 1978; Johnson, 1980-81), and there is
broad agreement that many shortcomings of class action practice can be
traced to judges’ use of inappropriate fee arrangements. The lodestar fee
formula, which bases compensation on time expended and hourly rates, is a
primary culprit. By rewarding effort instead of results, it discourages lawyers
from maximizing the value of plaintiffs’ claims and encourages sweetheart
deals that make class counsel and defendants better off at absent plaintiffs’
expense (Coffee, 1986, 1995a; Silver, 1992).
An ideal fee arrangement would be self-policing. It would harmonize the
interests of absent plaintiffs and lawyers, making it in lawyers’ interest to
prosecute class actions zealously, as duty requires. The lodestar method fails
to align interest with duty. It relies on monitoring to encourage superior
performance instead of incentives.
A self-interested class member would want to maximize his recovery net
of expenses. He would be perfectly willing to pay a larger fee when this
would yield a larger expected net recovery. Otherwise, he would prefer to
keep for himself money that could be transferred to his attorneys.
The lodestar method does not reflect this preference. It transfers money
from a class member to an attorney whenever the attorney works more,
whether or not the exchange of time for money leaves the class member
better off. Moreover, under the lodestar method, even a class member who
profits from a forced transaction may have a valid complaint. Work by the
hour can be done poorly or well, and class members may find it
advantageous to pay far more for the latter than the former. Suppose a class
could offer a lawyer $1000 for an hour of top-quality work that would
increase its expected recovery by $5000 or $200 for an hour of mediocre
work that would yield an expected return of $201. Both transactions would
leave class members better off, but they would rationally prefer a surplus of
$4000 to a surplus of $1. Unfortunately, the lodestar prevents class members
from choosing between these exchanges. By holding compensation per hour
constant, it motivates a lawyer to perform at the lowest quality level
sufficient to generate the stipulated fee.
The conventional wisdom is that class members often lose by exchanging
money for time when the lodestar method is employed. When a judge
decides which hours will be compensated and the hourly rate exceeds a
lawyer’s opportunity cost, a lawyer’s incentive is to bill as many hours as a
judge is likely to allow. For example, suppose that a lawyer with an
opportunity cost of $150 per hour expects a judge to authorize payment at
$200 per hour for up to 1000 hours of work. It would be irrational for the
lawyer to charge for less than the full 1000 hours. Any time diverted from
the class action to other matters would cost the lawyer $50 an hour. The
incentive to bill the entire 1000 hours will exist even if the last 100 hours
contribute nothing of value to the class.
In the example just described, the lawyer could do even better by
committing fraud. If the lawyer spent 900 hours on the class action, billed
the class for 1000, and spent the 100 hours saved on other cases, the lawyer
would gain an additional $15,000 by working on the other cases. Thus does
the lodestar method encourage lawyers to falsify billing records and create a
related need for monitoring.
Unfortunately, if also understandably, in most class actions no one is
likely to monitor class counsel closely or well. Judges are reluctant to make
careful evaluations of fee records, expenses and lawyers’ activities,
especially in the absence of serious complaints from absent plaintiffs. The
task is subjective, arbitrary and time-consuming. Absent plaintiffs also lack
incentives to bear the cost monitoring entails, and they confront the logic of
collective action as well.
Scholars have proposed many improvements on the lodestar method, the
simplest of which is increased use of contingent percentage compensation
arrangements and decreased use of the lodestar. Unfortunately, certain ways
of applying the percentage-based approach share the lodestar’s defects
because they too sever the connection between reward and return. When
class actions settle, attorneys are always paid in cash but absent plaintiffs
sometimes receive discount coupons or other in-kind relief (Miller and
Singer, 1998). The economic value of this relief is difficult to gauge,
especially when coupons are not actively traded. To help judges assign the
relief a value and justify their requested fees, plaintiffs’ attorneys offer
experts’ assessments. Often, these are based on assumptions about utilization
rates that are wildly optimistic. Coupon settlements typically benefit class
members far less than experts predict (Borenstein, 1996; In Camera,
May-June 1993, July-August 1993).
By weakening or severing the connection between fees and results, the
lodestar method and coupon settlements create opportunities for ‘sweetheart
deals’ that are good for defendants and class counsel but bad for absent
plaintiffs. Suppose that class counsel expects to win a judgment for $10
million at trial, $2.5 million of which will be paid to counsel as fees. Now
suppose the defendant, who privately agrees with the $10 million estimate,
offers coupons worth $5 million in settlement of class members’ claims and
an additional $3 million in fees. By making the offer, the defendant can save
$2 million. By accepting the offer, class counsel can gain $500,000. The loss
comes at the expense of class members whose recovery net of fees fall from
$7.5 million to $5 million. Not knowing that the negotiators think the case
is worth $2 million more, the judge is likely to go along with the deal,
especially when expert testimony inflating the value of the coupons is
An alternative to the coupon deal is an early settlement at a low price.
Suppose a defendant who expects to lose $20 million at trial offers $10
million in settlement, including $3 million in fees, shortly after the
complaint is filed. The defendant also threatens to litigate aggressively if the
offer is declined. Class counsel will find the offer attractive. Why risk $3
million in fees today in order to win a somewhat larger amount, say, $5
million, two years from now after an expensive fight? If the lawyer has not
logged hours sufficient to justify a $3 million fee, announcement of the
settlement can be delayed, allowing time for more discovery.
A third type of sweetheart deal expands the class definition or broadens
the release to make a settlement more valuable for a defendant. When
settling, a defendant naturally wants to free itself from liability to the fullest
possible extent. To accomplish this, a defendant may ask class counsel to
bring additional claimants into a class or to release claims a class was not
asserting. For example, a defendant may want to include personal injury
claimants in the settlement of a property damage class action, or a defendant
may want claims released over which other courts have exclusive
jurisdiction. In return, the defendant will offer a small fee increase that,
from class counsel’s perspective, will be an unearned bonus (Coffee, 1995a).
A fourth type of sweetheart deal uses a reversion to make a defendant’s
obligation appear larger than it is. A defendant promises to make a large
fund available, say, $20 million, subject to the condition that funds not
claimed by class members revert to the defendant. The attorneys’ fee is then
based on the $20 million. Because claim rates in class actions tend to be low,
the defendant is likely to keep a good deal of its money and the attorneys’
fee per dollar of relief actually received by class members is likely to be
Percentages that vary inversely with settlement size also discourage
lawyers from maximizing claim values. Suppose that a lawyer would receive
30 percent of the first $10 million won by a class, 25 percent of the second
$10 million, and 20 percent of the third $10 million. Now suppose the
defendant offers $20 million in settlement but that class counsel thinks the
case is worth $30 million. To reject the offer, class counsel must be willing
to risk $5.5 million in fees to win an additional $2 million. On the
reasonable assumptions that the higher dollars are riskier than lower dollars,
the gamble will attract only risk-preferring attorneys.
As a group, plaintiffs’ attorneys are risk-averse. Consequently, they must
be paid more to recover risky dollars than dollars that are easy to come by.
They must also be paid more in large cases that require great financial
commitments and that entail risks that are hard for plaintiffs’ attorneys to
diversify. A good fee arrangement would therefore vary directly with the
amount of the recovery rather than inversely to it, increasing the fee
percentage at the margin as the recovery grows (Stock and Wise, 1993).
Such an arrangement would also partially offset the declining marginal
value money has for attorneys, a problem of great importance in cases where
hundreds of millions or billions of dollars in damages are at stake.
A final but fundamental incentive-related shortcoming derives from the
possibility of parallel litigation, which prevents class counsel from obtaining
a property right in a certified class action. To appreciate this problem, it
must be understood that the first class action to settle moots all other similar
class actions and that only the attorney handling the settled case is likely to
earn a fee.
Suppose that two class actions alleging the same violations and brought
on behalf of the same individuals are filed in separate courts against the
same defendant. Now suppose the defendant makes two low-ball settlement
offers, one in each case, subject to the condition that when either offer is
accepted the other lapses. Unless the plaintiffs’ attorneys handling the cases
collude, a race to settle is likely to occur because only the attorney who
accepts first will be paid.
To make matters worse, collusion is nearly impossible. If both attorneys
reject the low-ball offer, the defendant can simply find a third lawyer, offer
him the deal, and agree to settle a class action filed by that lawyer. For
collusion to work, it must involve all plaintiffs’ lawyers, an impossibly large
It may seem unlikely that a defendant would search for an attorney to
represent a class. Yet the phenomenon has happened often enough to have a
name: the reverse auction (Coffee, 1995a). The name reflects the
defendant’s desire to find the attorney who will sell it immunity from
litigation at the lowest price. In a reverse auction, the low bidder gets the
18. Inadequate Monitoring
No incentive arrangement can wholly eliminate opportunities for agents to
benefit at principals’ expense. It is therefore important for principals to
monitor agents. In class actions, absent plaintiffs are unlikely to monitor
class counsel as carefully as they should. Monitoring by absent plaintiffs
may also lead to undesirable results when it occurs. For these reasons, judges
have a duty to examine how class actions are handled. Unfortunately, even
they miss many instances of opportunism.
There are only four candidates for the job of monitor in a class action:
the defendant, the named and absent plaintiffs, the presiding judge, and
third parties, such as uninvolved attorneys, who may keep track of the case.
The defendant can be ruled out immediately. A party that stands to gain
from class counsel’s opportunistic acts cannot be trusted to police them.
Nor are named plaintiffs necessarily good monitors. Having chosen or
been chosen by the attorneys who lead the class, named plaintiffs may too
often give them the benefit of the doubt. Named plaintiffs also, and properly,
benefit disproportionately from class settlements. They often receive
incentive payments that are intended to compensate them for special risks
and costs they incur (Krislov, 1990; Solovy, Kaster and Jiganti, 1990). The
availability of these payments may color their judgment.
Although harmed by opportunism, absent plaintiffs also cannot be
expected to monitor class counsel efficiently. Many or most absent plaintiffs
hold stakes that are too small to warrant the expense. It would be irrational
to spend $100 monitoring a lawsuit in which one has a $100 interest, but to
oversee class counsel effectively an investment of thousands could be
required. When expected recoveries are small to begin with, as in class
actions they usually are, only class members who ignore economic factors
will monitor class counsel actively.
Class members with sizeable claims may also be discouraged. First, they
often learn that class actions are pending late in the day when settlement
notices are delivered. They then face the daunting and expensive task of
reconstructing events and finding signs of collusion, an undertaking in
which they can expect no help from class counsel or the settling defendant.
Second, an absent plaintiff who uncovers wrongdoing cannot privatize the
entire return. If the size of the settlement should ultimately increase, all class
members will share the gain. The private cost of monitoring may exceed the
expected private gain even when the expected class-wide return would be far
greater. Third, the inability to privatize returns on monitoring creates a
free-rider problem within a class. When any absent plaintiff bears the cost of
monitoring, all profit. Why, then, should any particular absent plaintiff bear
the expense instead of free-riding on others who choose to do so?
A further difficulty with relying on absent plaintiffs is that they too can
be bought off. A class member who takes a special interest in a case can be
made a named plaintiff by class counsel and offered an incentive bonus. Or a
defendant can offer to allocate settlement benefits in a way that favors an
involved class member. Because the interests of particular plaintiffs always
diverge from those of others, an absent plaintiff who serves as a monitor
may himself have to be watched.
In theory, judges could be excellent monitors. They understand legal and
factual issues better than claimants are likely to, and they are increasingly
likely to have experience managing complex litigation. In practice, the
quality of judicial monitoring is uneven because judges lack appropriate
incentives and have limited fact-gathering capabilities. Many judges are
overworked, underpaid, and subjected to strong institutional pressures to
move cases along. It is unreasonable to expect them to divert time from live
matters to class actions that are all but settled. Judges also have to rely on
others for most of their information. Yet their usual sources class counsel
and defendants - cannot be relied upon to pinpoint evidence that would
scuttle a deal. To make matters worse, in many cases the data judges need to
answer the crucial question - Is the money enough? - do not exist. The
history of previously settled cases may be thin or no one may have collected
it. Often, the most comparable cases are other settled class actions that also
may be tainted.
Many judges are poorly placed to serve as monitors because they are
architects of the agreements they are asked to review. Often, judges structure
rulings to encourage settlement negotiations, they (and the magistrates or
masters they appoint) become deeply involved in the negotiation process,
and they use powerful signals about future rulings to soften plaintiffs’ and
defendants’ bargaining positions (Resnik, 1982; Schuck, 1986, 1987). This
behavior may be motivated by a jumble of forces, including solicitude for
plaintiffs facing extraordinary trial delays, pressure to clear dockets,
personality quirks, and a desire for the prestige enjoyed by judges who settle
big cases. Because trial judges who orchestrate settlements can be expected
to approve them, the primary monitors of such settlements must appellate
Judicial scrutiny of fee requests is also uneven. Monitoring is crucial
under the lodestar method, which encourages lawyers to inflate hours and
claim excessive hourly rates. It is also tedious and time consuming (Silver,
1992; Tomkins and Willging, 1986). Judges must cast themselves as
auditors and examine lawyers’ time sheets and expenses line by line. Many
judges short-circuit this process. They either approve fee applications as
submitted or they make rough-cut reductions. Some judges also intentionally
deviate from market standards when setting hourly rates, believing that
market rates are too high. Judicial monitoring is not an especially good
means of policing fee-related misconduct (Coffee, 1986; Silver, 1992).
Finally, consider other third parties. These may be attorneys whose
clients are absent class members who object to a settlement. Or they may be
public interest lawyers or public officials, such as attorneys general or
consumer protection agencies. There are many cases in which third parties
have become beneficially involved, but as a general matter they are in short
supply. Class members usually lack incentives to get them engaged.
Moreover, private lawyers can usually protect clients with large claims by
opting them out. This being the cheaper course, it is the one usually taken.
The participation of third parties also can raise several difficulties. First,
public officials may intervene to advance political agendas or interests that
have little to do with claimants’ welfare. For example, they may oppose
certain kinds of lawsuits and, assuming the guise of consumer advocates,
they may challenge fee awards to discourage lawyers from bringing more
cases. Second, class counsel may co-opt third parties, especially lawyers
representing claimaints, by making them co-counsel and offering them fees.
Third, third parties may intervene solely to be bought off. There are many
lawyers, colloquially referred to as ‘bottom feeders’, who use objections to
extort payments. They hold up settlements unless and until class counsel
offers to share fees with them, effectively paying them to withdraw. Public
officials can also extort payments. They may want settling parties to place
funds under their control or to assist them with other endeavors by donating
services or cash. One can never be certain that third parties’ motives are
Because none of the available candidates has incentives to monitor
appropriately, it should not be surprising that objections to proposed class
settlement are both rare and rarely successful. Judges approve settlements
without modification in 90 percent of the cases where objections are filed.
This is true whether an objection concerns the adequacy of a settlement or a
proposed award of attorneys’ fees (Willging, Hooper and Niemic, 1996b).
What the statistics cannot show may be equally important. Many objections
that succeed in altering or forestalling settlements may be strategic or
19. Unresolved Economic Asymmetries
Fee awards in class actions average just above one-third of the amount
recovered. Rarely do they exceed 50 percent (Martin et al., 1996; Willging,
Hooper and Niemic, 1996b). Many judges also reduce the size of a fee as a
percentage as the size of the recovery rises (Lynk, 1990, 1994). These
practices tend to tip the balance of economic power against claimant groups
and in favor of defendants.
As a practical matter, class members can spend only about 33 cents out
of every dollar in controversy to prosecute a lawsuit. This is the amount they
can offer their lawyers as fees. By contrast, defendants can spend 100 cents
on the dollar or more. Judges do not control defendants’ spending. Because
the amount one spends affects one’s chances of winning, free-spending
defendants will predictably gain an edge over class members. This
advantage should manifest itself at trial, where plaintiffs should lose
relatively often, and in settlement, where bargains reflect one’s prospects at
Differences in ability to diversify risks and manage litigation costs
should strengthen the spending advantage defendants enjoy. Plaintiffs’
attorneys practice mainly in small firms of 10 or fewer lawyers. Plaintiffs’
firms with 25 or more lawyers are exceedingly rare (Stock and Wise, 1993).
Firms of this size have limited capital to invest and great difficulty
diversifying large risks. For them, class actions, which require cash outlays
of $500,000 on average, demand lawyer time worth far more, and take
longer than conventional lawsuits to settle, can be bet-the-firm cases. To
protect themselves, plaintiffs’ attorneys spend conservatively and otherwise
minimize their financial commitments to class actions. They are low-cost
operators because they have to be (Coffee, 1986).
Defendants find class actions easier to manage. Only large businesses,
governmental entities, and other well-heeled institutions are likely to be
targets for class actions because only they have the resources or insurance
needed to pay large judgments. But the size and wealth that make them
attractive targets also enable them to tolerate and diversify litigation risks
more easily than plaintiffs’ attorneys can.
Being relatively wealthy, defendants usually pay lawyers to defend class
actions wholly or partly on the basis of hourly rates. They also reimburse
their lawyers for expenses. With their clients absorbing the costs and risks of
litigation, defense lawyers have every incentive to turn class actions into
protracted lawsuits and frequently do. Every motion a defense lawyer files,
every discovery objection and request, every interlocutory appeal, and every
other dilatory tactic forces plaintiffs counsel to consume scarce resources.
The incentive to force plaintiffs’ counsel to bankrupt themselves may partly
explain why class actions consume so much more of judges’ time than other lawsuits.
Defendants also can take advantage of political connections that are not
available to plaintiffs. For example, insurance companies faced with class
actions have sought protection from insurance regulators and legislators. In
effect, they have sought to cut short litigation by acting in other forums
where they enjoy significant advantages. When success outside the
courtroom can be a ‘silver bullet’ that puts a conclusive end to litigation,
plaintiffs’ attorneys have to respond. But doing so requires more political
savvy, greater financial resources and better connections than many of them
There are some economic asymmetries that favor plaintiffs. For example,
plaintiffs who survive early dispositive motions are entitled to discovery on
the merits, and defendants usually possess far more discoverable information
than group members do. By requesting this information via discovery
requests that are themselves inexpensive to produce, plaintiffs’ attorneys can
force defendants to bear the considerable cost of locating, reviewing, and
producing it. Particularly costly to defendants is lost managerial time that
interrupts its business.
Defendants can do little to expose group members to offsetting costs,
which means that once discovery gets underway, plaintiffs possess unique
leverage. Recognizing this, defendants work hard to prevent discovery from
starting and, when their efforts fail, they often settle to avoid discovery costs,
whatever the merits of plaintiffs’ claims may be. This threat advantage is a
temporary one - it decreases in size as discovery is taken and future costs are
converted to sunk costs. Its size also bears no necessary relation to the value
of class members’ claims. Finally, defendants can offset plaintiffs’ leverage
somewhat by producing documents and witnesses in volume. A plaintiffs’
attorney who is swamped with information may lack the resources needed to
sort the wheat from the chaff.
20. Lack of Market Discipline
Many of the forces that encourage agents to represent principals well in
other contexts are absent from class actions and consolidations. For
example, claimants have little or no power to hire or fire lead attorneys or to
decide how much they will be paid. Claimants may not control the
settlement decision. They cannot sell their claims when they are
disappointed with counsel’s performance, as owners who are dissatisfied
with managers can sell their shares. Nor in the litigation realm is there an
analogue to the takeover market. Attorneys who fail to maximize the value
of assets cannot easily be dethroned.
Because so many market checks are missing, the quality controls that do
exist must operate especially well. Primary among these are the choice of
counsel itself and counsel’s compensation arrangement. As explained, the
latter often misfires because judges use the lodestar method or otherwise fail
to compensate counsel appropriately.
It is difficult to overstate the importance of the attorney chosen to lead a
litigation group. This person must marshal the resources needed to prosecute
a case, see that they are used efficiently, and decide important matters of
strategy. When a consensual litigation group is built via a referral market,
there is reason to expect a qualified person to be given these tasks. Referring
lawyers, who serve as brokers for their clients, have an interest in selecting
competent lead lawyers who will maximize referral fees by maximizing
claimants’ recoveries. Referring lawyers also are potential monitors and
sources of future business. Lead lawyers who shirk or settle cheaply can
expect criticism from referring lawyers. Lead lawyers who develop bad
reputations can expect to be denied future work (Resnik, Curtis and Hensler,
1996; Spurr, 1988).
When a judge selects a lead attorney, the danger is greater that an
unsuitable individual will be chosen. A judge has no stake in the amount
class members recover. Consequently, a judge has no particular interest in
selecting a lawyer who can be expected to maximize the value of their
claims. Absent a contest for control, a judge has no reason to do other than
appoint the attorney who files the class action complaint.
Sometimes, it is safe to assume that the attorney who files a complaint is
the wrong lawyer for a class. This is true, for example, when a defendant
recruits lead counsel. Defendants facing large numbers of individual
claimants sometimes see the class action as a means of resolving all their
legal difficulties in a single forum at a bargain price. With this goal in mind,
they approach one plaintiffs’ firm after another until they find a lawyer who
is willing to negotiate a cheap package deal. This is another example of the
The law requires judges to cure the problem of inadequate lead counsel.
More accurately, it permits them to allow class actions to proceed only when
absent plaintiffs are adequately represented. This includes representation by
adequate counsel. It is clear that judges can appoint substitute counsel when
a defect of counsel is all that stands in the way of class certification.
Judges are reluctant to assert this power, for several reasons. First, they
rarely see lawyers they recognize as bad apples. Lawyers tend to file lawsuits
in courts where they are liked and respected. They also enlist respected local
lawyers as co-counsel when they are forced to litigate in unfamiliar courts. A
lawyer who is a judge’s favorite or has a judge’s favorite in tow stands a
high likelihood of winning appointment.
Second, many judges like lawyers who are deferential and respectful and
who help them by settling cases. It is unreasonable to expect these judges to
prefer aggressive lawyers to passive ones. The preference for cooperative
attorneys may explain why judges so often give the role of lead counsel to
plaintiffs’ lawyers who negotiate settlements before seeking class
certification. One study reported that ‘a proposed settlement was submitted
to the court before or simultaneously with the first motion to certify’ in 49
percent of class actions certified for settlement (Willging, Hooper and
Niemic, 1996b). It also may explain why judges frequently re-appoint
certain prominent lawyers who are notorious for settling quickly (Tidmarsh,
Third, judges are rightly concerned about denying lawyers who uncover
wrongdoing the reward of serving as class counsel. Unless lawyers are
compensated for time spent inventing liability theories and investigating
misconduct, they will be discouraged from doing so. The privilege of being
appointed class counsel and the contingent right to compensation that comes
with the position maintain the incentive to police wrongdoing.
Some judges have attempted to stimulate competition for the role of lead
counsel and to reduce attorneys’ fees by auctioning the right to represent a
class (Macey and Miller, 1991, 1993). The results have been mixed. A judge
cannot sell a class action to the highest bidder, a transaction that might be
predicted to identify the attorney who can derive the greatest value from a
set of claims. They can auction only the right to control the class and, in the
event of success, to be paid. The temptation is to sell the lead counsel
position to the lawyer willing to handle a case for the smallest fee.
This is a serious mistake. The cheapest lawyer is likely to be the one
whose opportunity cost is lowest because he cannot attract other business.
Such a lawyer would be a poor candidate to represent a class.
An attorney who represents an individual client with a large stake, such
as an institutional investor, would be better suited for the task. Sophisticated
clients with large claims are likely to pick good lawyers. Because they can be
expected to perform quality control for themselves, they can provide the
same service for a class. A lawyer who represents a large number of clients
with small stakes could also be a good choice. The lawyer’s success in the
referral market would show that other attorneys think he or she has the
ability to run a group suit. Great financial wealth, a history of committing
significant resources to lawsuits, and a record of success in high-stakes
litigation are also important. Judges should consider indicia of quality in
addition to price.
Two other quality controls that remain in most class actions are the right
to object to proposed settlements and the right to opt out. These rights are
not especially meaningful.
Absent plaintiffs are likely to object only when the cost of objecting is
less than the expected gain. The cost of making anything more than a
perfunctory objection can be quite large. An absent class member must
retain counsel to canvass the record, attempt discovery of class counsel, and
argue against approval of a settlement in court. By contrast, the expected
payoff from objecting usually is small. It is the difference between the
amount being offered and the amount one might recover under a more
favorable settlement or trial judgment. Because objectors rarely succeed in
changing settlements, the gain is likely to be worth the cost only for
plaintiffs whose claims are very large. Even they may be discouraged by the
logic of collective action, which creates pressure to free-ride.
In most situations, procedural rules entitle absent claimants to exclude
themselves from class actions by filing opt out forms. This opt out right
bears the same relation to objecting that exit bears to voice (Hirschmann,
1970). In theory, claimants could use their opt out right to guarantee
themselves at least the expected value of their claims in conventional
litigation (A in Figure 1). Some claimants have significantly increased their
recoveries by opting out of class actions and filing conventional lawsuits
The right to opt out may be more efficacious than the right to object, but
it generally means that a claimant must litigate alone rather than as part of a
second class action that is more efficiently run. For plaintiffs with small
claims, this is impracticable, and for plaintiffs with large claims it may be
disadvantageous. The empirical findings that class members rarely opt out
and that those who do so overwhelmingly have large claims and are
represented by counsel are not surprising (Willging, Hooper and Niemic,
1996b). Finally, by opting out absent plaintiffs protect only themselves.
Other class members remain exposed to exploitation.
21. Certification of ‘Settlement-Only’ Classes
Class actions often are certified and settled at the same time. In
approximately one-third of all cases where certification is granted, it is for
settlement only (Willging, Hooper and Niemic, 1996b).
The set of class actions that are certified for settlement contains two
subsets: class actions that would be certified for trial because they meet all
the criteria for certification, and class actions that would not be certified for
trial because they do not meet the criteria. For shorthand, the latter will be
referred to as ‘settlement-only classes’. Sell-outs are especially likely in
To see why, one must understand that plaintiffs’ attorneys need
defendants’ cooperation to get settlement-only classes certified. These cases
do not meet the requirements for certification. Consequently, certification
would be denied in these cases if defendants were to object. To obtain
defendants’ acquiescence, plaintiffs’ attorneys must make concessions.
The magnitude of the concession depends on the size of the fee interest
putative class counsel has in the individual cases that will proceed in the
absence of a class action. Assume the following: 100 individual plaintiffs
will file cases; the attorney seeking to act as class counsel will represent 10
of these; each plaintiff’s case has a settlement value of $500,000; and the
attorney will contract with each plaintiff for a 25 percent contingent fee. In
this example, the attorney expects to earn $125,000 in fees per case times 10
cases, a total of $1.25 million. The attorney will make nothing off the
remaining 90 cases because other lawyers will handle them. Clearly, a
class-wide settlement of all 100 cases yielding a fee greater than $1.25
million would be better for the attorney. For example, a fee award of $2.5
million in a class action would be an increase of 100 percent.
Therein lies the rub. At a rate of 25 percent, a class-wide recovery in the
amount of $10 million would generate a $2.5 million fee. But a $10 million
settlement would grossly undervalue the 100 claims, which would generate
$50 million in winnings if tried separately. Settlement-only classes create
opportunities for defendants to pay class counsel huge premiums while
saving themselves enormous sums.
Defendants’ opportunities to save money and to purchase class counsel’s
cooperation are especially great in so-called ‘futures’ class actions. These
lawsuits cover persons whose injuries are latent. For example, persons
exposed to a disease-causing substance who are asymptomatic could fall
within a futures class. Although many of these persons can be predicted to
sue after their injuries ripen, few are likely to retain counsel during the
latency period. A futures class action offers a defendant an opportunity to
settle all possible claims cheaply before any symptoms appear. By offering a
bribe to an attorney who is willing to serve as class counsel, a defendant can
settle immature claims for pennies on the dollar.
22. Settlements of Inventories of Signed Clients
Attorneys who handle class actions often have inventories of signed clients.
In principle, this is good. An attorney who manages to build a client base in
a litigation area where other lawyers are competing for business has
convinced the market that he is a competent lawyer. The inference of quality
is especially strong when an attorney has acquired clients by referral from
Concerns arise when an attorney settles a client inventory apart from and
on terms that differ from those he negotiates for a class. This is
accomplished by excluding the attorney’s signed clients from the class
action. For example, the class definition may cover only persons who file
complaints after a certain date. If class counsel’s signed clients filed before
that date, a class-wide settlement would not cover them.
When a class action and a block of signed clients settle separately, there
is reason to fear that one group may benefit at the other’s expense. Usually,
the worry is that the signed clients will exploit the unsigned class members.
A wealth transfer from the latter to the former will benefit counsel if fees in
the signed clients’ cases are higher than those typically awarded in class
actions. A transfer may also smooth the process of settling the client
inventory by creating a surplus that can be used to buy off disgruntled clients
or to reward lawyers or organizations, such as labor unions, who provided
referrals (Coffee, 1995a; Koniak, 1995; Nagareda, 1996b).
23. Maximization-Related Conflicts
A subject neglected in the scholarly literature on class actions is the
inevitability of conflicts in the process of preparing cases for trial. Although
class counsel can often advance all absent plaintiffs’ interests concurrently
by supplying them with joint goods, even related claims invariably have
idiosyncratic aspects that require individual attention and targeted
expenditures. By devoting time and resources to these issues, class counsel
neglects the interests of class members who claims lack these characteristics
and who would prefer that counsel’s time and resources be spent in ways
more likely to be of benefit to them (Rhode, 1982a; Silver and Baker, 1998).
Resource allocations on common issues can also require inter-claimant
trade-offs, and usually do. Class members with small compensatory claims,
such as employees who were recently hired at discriminatorily low wages,
may be best off if litigation resources are devoted mainly to the pursuit of
punitive damages, pay increases and promotions. This use of resources may
not appeal to class members with large compensatory claims, such as
long-time employees. They may benefit most from efforts to model the
impact of wage discrimination on their lifetime income and retirement
plans. Other class members, such as future job applicants, may care only
about new job openings.
By rejecting a settlement offer, class counsel also trades off the interests
of some class members against those of others. Every litigation group
contains members with different attitudes toward risk, with different views
on the time-value of money, and with different desires for relief. By pursuing
a high-risk/high-reward approach to litigation or a low-risk/low-reward
approach, class counsel necessarily decides which class members’
preferences will be favored. Class counsel does the same thing by deciding to
settle for a larger amount tomorrow instead of a smaller amount today and
by pressing for dollars rather than in-kind relief or symbolic vindication.
Every decision to use time or money for one purpose rather than another or
to adopt one litigation strategy instead of another necessarily resolves
conflicts among class members’ interests.
Maximization-related conflicts inhere in all class actions because there is
nothing like the stock market that claimants can use to ameliorate them.
Shareholders can freely and cheaply move between different companies and
different investments so that their portfolios accommodate their risk
preferences, attitudes toward the time value of money, and so on. Claimants
cannot move from one class action to another to achieve the same result.
F. Inappropriate Settlement Allocations
A recovery can be allocated equitably or inequitably. From a Paretian
perspective, allocations on CE and FD in Figure 1 would be inequitable
despite being efficient. They would leave some class members worse off than
they would have expected to be had they sued on their own.
A nascent but developing literature examines the equity of settlement
allocations in class actions and other group lawsuits (Coffee 1998; Hay,
1997b; Hazard, 1995; Koniak, 1995; Menkel-Meadow, 1995; Silver and
Baker, 1998). Even scholars who care mainly about deterrence agree that it
is important to compensate plaintiffs fairly when that can be done at
reasonable expense (Rosenberg, 1987, 1996).
There is a tendency in the literature to run allocation issues and
efficiency issues together. When certain claimants are shortchanged, it may
be because others receive too much, because a settlement fund is too small,
or because others receive too much and a settlement is was too small. The
unique contribution of equity is to set standards for dividing funds, not for
overall fund size.
24. Allocation-Related Conflicts Inhere In All Class Actions
When it comes to dividing settlement funds, class members’ interests always
conflict. In Figure 1, X and Y share an interest in moving away from A in
the direction of GH, but they have opposing preferences concerning where
on GH they wind up. X prefers divisions to the northwest; Y prefers
divisions to the southeast. A group lawsuit is a mixed-motive game in which
the interest in cooperating for mutual gain is in tension with the interest in
getting as much as possible for oneself. The tension exists whether a group
seeks injunctive or monetary relief.
The number of ways of allocating recoveries and the conflicts they
generate make clear the need for standards. If a class containing only
members X and Y were to recover $1000, the entire range of options from
($1000,$0) through ($0,$1000) would be available. If member Z were added
to the group, the number of possible allocations would expand exponentially.
To avoid arbitrariness, judges and lawyers require normative allocation
Unfortunately, there is no widely accepted, general account of the equity
of class action settlements. There are instead many assessments of particular
allocations that commentators either found deficient or wished to defend.
Many of these assessments rest on deeper principles of equity or distributive
justice that are neither made explicit nor carefully examined. For example, a
scholar may criticize an allocation plan for ignoring differences in the size
and strength of claims without explaining why these differences should
Also missing is a cost-benefit analysis of equity. It is more expensive to
pay claimants amounts that roughly reflect the size and strength of their
claims than it is to engage in damages averaging and pay them equal
amounts, and it is more expensive still to distribute payments that reflect
fine differences between claimants (Coffee, 1987a, 1998; Silver and Baker,
1998). One therefore wants to know how much cost should be borne to
obtain a marginal increase in equity. There is general agreement that the
importance of equity increases with claim size. There is also a growing
consensus that because class settlements yield less than $400 per claimant
on average, often the game is not worth the candle. Between the extremes of
large claims and small recoveries, however, there is little agreement as to
how much equity is enough. This dissensus has important practical
25. Causes of Arbitrary Allocations
Class counsel and defendants design allocation plans. In effect, they serve as
agenda setters, deciding which of many possible allocations are presented to
courts and claimants for consideration. Like other agenda setters, they have
considerable power to control outcomes.
On the surface, this power may seem wasted on class counsel. Because
allocation formulas do not affect the size of fee awards, class counsel might
not seem to care how relief is divided. Reality is more complicated than this.
Class counsel will receive no fee until a settlement is judicially approved.
Approval depends, among other things, upon the judge’s views and the
reaction of the class. Class counsel therefore has an interest in selecting
allocations that satisfy these constituencies. Class settlements also frequently
contain ‘walk away’ provisions that condition settlement on an agreed level
of claimant participation. When allocating common funds, class counsel
may keep the required level of participation in mind.
As explained, only class members with large claims are likely to object
or opt out. This is why ‘walk away’ provisions usually target large-claim
plaintiffs. A rational strategy for class counsel may therefore be to
over-compensate large-claim plaintiffs and to shortchange everyone else.
This approach enables class counsel to mollify large-claim plaintiffs without
risking a revolt by small-claim plaintiffs, who rarely complain.
The desire to avoid objections can also lead to class counsel to
overcompensate class members who belong to organizations. For example,
an asbestos class action may include workers who are union members and
workers who are not. Because a union can be an effective advocate of its
members’ interests, it may contest a settlement that treats them poorly. Class
counsel may therefore be inclined to involve the union in the design and
administration of the settlement and to overcompensate its members.
A separate dynamic can cause class members with large claims to be
underpaid. The typical class contains a small number of persons with
sizeable claims and a far larger number whose claims are comparatively
small or weak. The latter claims may add little value to the group lawsuit,
but they increase greatly the number of mouths to be fed. Unfortunately for
class members with large claims, judges are reluctant to deny any subgroup
of absent plaintiffs a share of settlement benefits, especially a subgroup that
is large. The price of feeding everyone is that plaintiffs with large claims
receive less than they should.
The weak may also beggar the strong in litigation groups simply because
the weak have less at stake. In bargaining situations, the player who stands
to lose the most from the failure to reach an agreement can be expected to
give up the most to salvage a deal. Plaintiffs with large claims have more at
stake than plaintiffs with small claims, often far more. Small-claim plaintiffs
can therefore take advantage of them. Risk aversion will predictably cause
plaintiffs with large claims to sacrifice even more because the risks attached
to large claims are harder to diversify and because large claims constitute a
more sizeable portion of plaintiffs’ wealth.
Defendants also may have interests in skewing allocation formulas. For
example, they may want to direct as much money as possible to claimants
who are current customers, suppliers, employees, or shareholders and as
little as possible to others (Coffee, 1998). A plaintiffs’ attorney who is
indifferent to the manner of allocation will have no reason to resist a
defendant’s efforts to direct funds to its favorites.
26. Possible Allocation Standards
The possibility of arbitrary allocation plans makes plain the need for
concrete normative guidelines. Many scholars seem to embrace a Paretian
standard of equity according to which class members should receive at least
the expected value of their claims in conventional litigation. This probably
includes most scholars who criticize allocations that ignore factors bearing
on the size and strength of claims or that take account of irrelevant factors
(Coffee, 1995a; Cramton, 1995; Koniak, 1995; Menkel-Meadow, 1995).
They seem to assume that size and strength have the greatest impact on
claim values in conventional litigation, and they therefore conclude that
when size and strength are ignored, claimants are shortchanged.
The merit of the Paretian standard is that it respects the intuition that
rational persons would not join litigation groups hoping to make themselves
worse off than they are on their own. They would do so expecting collective
action to enhance the value of their claims.
There are many difficulties with the Paretian standard, however. One is
that the value of claims in conventional litigation is not self-evident. It must
be assigned, and the data needed to make the assessment may be
unavailable. A second is that the proper baseline to begin with may not be
the expected value of claims in conventional litigation. Instead, one could
start with a claim’s highest expected value in any possible coalition of
claimants. Needless to say, this value is not self-evident either. A third is
that a commitment to Paretianism may not lead to the conclusion that
allocations should reflect only the size and strength of claims. Other factors
could also affect the value claims have in conventional litigation, including
factors that may have nothing to do with the merits such as sex, age, race, or
jury appeal. It is an empirical question which factors matter in particular
cases. A fourth difficulty is that a Paretian standard does not tell one how to
divide any surplus that exists above the chosen baseline. This shortcoming is
especially problematic when class actions include some or solely claims that
are too small to litigate individually, as most do.
A fifth problem is that claimants who participate in consensual group
lawsuits rarely take steps in advance to identify or protect anything that
might serve as a baseline for Pareto comparisons. Most such plaintiffs do
nothing to provide for the division of settlement funds. They trust their fates
to the discretion of their attorneys, reserving only the right to veto
settlements as a means of protecting themselves. The veto right is often more
illusory than real, however, because plaintiffs rarely have independent
counsel to advise them how to use it.
Scholars who embrace Paretian standards must also consider how
shortfalls should be dealt with. Settlements yield actual values, and actual
values may be smaller than expected values even when class actions are run
efficiently. When a group-wide settlement is too small to pay everyone the
amount required by the Paretian baseline, some or all claimants must be
shortchanged. The difficulty is in deciding which claimants and how much.
Again, the number of possible allocations is exceedingly great.
27. Reform Proposals
There is broad agreement that judges should change certain aspects of the
way they manage class actions. For example, whether an auction, a
contingent percentage or even the lodestar method is used to set fees, most
scholars encourage judges to set the terms of compensation up front. Today,
judges often delay this matter until class actions settle. This practice requires
class counsel to devote significant economic resources to a risky venture
with no formal understanding as to the expected return. It also complicates
settlement negotiations by leaving an additional and sizeable matter for the
negotiators to resolve. The practice also burdens class counsel with a conflict
of interests by creating the possibility that an impasse over fees may block an
attractive settlement on the merits.
Investors in risky undertakings rarely tolerate this degree of uncertainty
or conflict. They provide in advance for the division of returns, even when
there may be no returns to divide. By doing so, they encourage their agents
to work hard for their benefit. Judges presiding over class actions should
mimic the market in this regard (Federal Judicial Center, 1995).
The practice of setting fees at the end also skews fee awards downward.
In hindsight, the risk of loss will naturally seem smaller than it was. It is
also more difficult to award large fees when the precise amount of the fee is
known. This problem disappears when fees are set before any money is
offered in settlement.
Another way to improve the class action to use percentage compensation
instead of the lodestar method whenever possible and to increase
percentages at the margin as the amount recovered rises. By offsetting risk
aversion and the declining marginal value of money, this will encourage
class counsel to strive for larger recoveries instead of being satisfied with
Other reform proposals include giving class members the right to opt out
of all settlements (Rutherglen, 1996), forced fee sharing between class
counsel and lawyers representing clients who opt out (Coffee, 1987b),
improved notice (Woolley, 1997), allowing absent plaintiffs to sue class
counsel for malpractice and collusion (Koniak and Cohen, 1996),
encouraging judges to decide merits issues before motions for class
certification (Berry, 1980; Priest, 1997), increased use of injunctions to
prevent parallel class litigation (Miller, 1996), restricting opt-out rights
(Perino, 1997), and alternative fee arrangements that link counsel’s
compensation to fees paid in individual representations (Hay, 1997b).
28. The Potential for Good
The class action is a procedural device with extraordinary potential for good.
It can force product manufacturers and service providers to take account of
externalities that they would otherwise ignore, encouraging them to perform
more efficiently and compensating victims who would otherwise lack
remedies. It can make the procedural system operate more efficiently by
enabling judges to process related claims en masse at less cost per claim. It
can save claimants’ and defendants’ money by enabling both to take
maximum advantage of economies of scale. It can align the stakes of
defendants and claimants so that litigation investments are more nearly
equal and settlement bargains better reflect the merits of claims.
29. The Potential for Bad
The class action also entails important risks and dangers. Conflicts among
claimants inhere in all class actions and must be resolved as cases are
prepared for trial and settled. Lack of market oversight and improper
judicial regulation create opportunities for class counsel and defendants to
collude. This can frustrate the cost-internalization function of the class
action by enabling wrongdoers to escape the social costs of their conduct. It
can frustrate the compensatory function as well by shortchanging claimants.
The latter can occur independently of the former as a result of improper
30. The Need for Dispassionate Analysis
At the microeconomic level, class actions are fascinating and truly intricate.
Agency problems abound, and strategies for dealing with them are often
subject to counter-strategies that are difficult to plan for in advance.
Compounding the problem is the fact that judges do not always understand
or even care about the impact their decisions have on the microstructure of
the class action. Many decisions, such as those that require individual notice
in small-claim class actions and those that announce the homogeneity of
injunctive class actions, are patently flawed. Politics also takes a toll,
especially when it comes to lawyers’ fees, a subject of great partisan debate
in the United States. If the atmosphere surrounding the class action ever
cools, it may be possible to reform the device and avoid many excesses.