The Case. In Amchem the trial judge, applying the class action rules, certified a class made up of all workers who had worked at certain times for certain specified employers, thereby bringing into the case workers who had not yet filed suit against the defendants. The judge then approved a settlement that required the employers and their insurers to pay many millions of dollars into a fund that would remain in existence for many years. The fund set minimum criteria for compensation. The settlement provided that the administrator would pay injured workers who met those criteria several thousand to several hundred thousand dollars each, depending on how much harm the worker had suffered.13
The trial judge found that a class action was appropriate because common questions of law or fact predominated over separate questions. He decided that the class action method was superior because otherwise too many workers would have to wait too long to receive too little compensation. And he concluded that given the alternatives, the settlement was a good one—even for those exposed workers who had not yet shown any sign of disease.14
The Supreme Court eventually heard the case and disagreed. The Court conceded that the entire set of cases presented some “common questions of law or fact.” Each case involved exposure to asbestos. Each plaintiff would benefit from a quicker settlement involving fewer costs of administration. But the individual cases also differed in many important respects—for example, in the kind of exposure, length of time exposed, and kind of disease that the exposure produced. The Supreme Court thought the district court should have created subclasses of plaintiffs. Each subclass would represent a subgroup with members who had more in common. Different lawyers representing different subclasses would help the judge better ensure fair representation of the different groups, particularly those made up of individuals whose disease had not yet appeared. Given the possibility of subclasses, the Court held, the common interests among the plaintiffs in the single big class did not predominate, and the single big class therefore did not “fairly and adequately protect the interests of the class.”15
All the members of the Court recognized that the district court was more familiar with the case than were they. And they all were willing to give weight to the district court’s analysis and conclusions. Nonetheless, the majority thought that the district court had gone beyond a fair application of the class action rules, whereas in my (dissenting) view the law gave the district court adequate authority to decide as it did. The district court had considered the matter at length; it had weighed (1) the common fact of asbestos exposure and the plaintiffs’ common interest in receiving compensation quickly without paying large legal fees (2) against the differences; and it had made more than three hundred factual findings that explained why the former predominated. The district court had concluded that the settlement was fair to all, in particular because the fund would contain enough money to pay those who were not yet ill and not yet represented.16
The difference between the majority and the dissent is one of degree, not of legal principle. Yet practical considerations support more, rather than less, deference, and they thereby underline the need to take full account of specialization. The district court found that, had the Amchem settlement taken effect, it would have paid “an estimated $1.3 billion and compensate[d] perhaps 100,000 class members in the first 10 years.” The alternatives to settlement included long delays, high administrative costs, potential bankruptcies, and smaller eventual payment to victims, or perhaps no payment. Indeed, empirical studies show that the difficulty of handling the hundreds of thousands of asbestos cases that crowd court dockets means that, without settlement, administrative costs exceed amounts paid victims by a ratio of nearly two to one. Delays were often so long that in one 3,000-member asbestos class action, 448 class members died during the course of litigation.17
Moreover, a basic mission of the court system as a whole is to bring about the fair and speedy resolution of disputes. The circumstances of asbestos litigation offer a vivid illustration of the Court’s need to grant trial courts considerable leeway in their effort to bring about a fair resolution of the underlying disputes. That leeway finds its justification in the principle of specialization. Adherence to that principle helps courts work more effectively, which in turn helps maintain public acceptance.
Chapter Twelve
Past Court Decisions:
Stability
WHEN THE SUPREME Court considers issues that it has previously decided, how much weight should it give to precedent? When should the Court overrule that precedent? The relevant legal doctrine, stare decisis, emphasizes the need to “maintain what has been decided.” Judges do not find it difficult to follow that doctrine when they believe an earlier decision is right. But suppose a judge believes an earlier decision is wrong. And suppose the judge belongs to a Supreme Court with the power to overrule an earlier decision. What then? Under such circumstances, the judge must make a pragmatic decision, weighing the harms and benefits of stability against change. Furthermore, the judge must emphasize stability. Stability makes the judicial system and the law itself workable. Without stability the Court’s decisions seem ad hoc and unpredictable—not part of a system at all. This is contrary to the Constitution’s objectives and tends to undermine public acceptance of the Court’s decisions.
BROWN—WHEN STABILITY MUST GIVE WAY
SOMETIMES A COURT must overrule an earlier decision. The passage of time may make clear that the legal rule set forth in the case was wrongly decided to begin with. It may also show that the earlier case is harmful or that it has become outdated as, in light of changing circumstances, the law in related areas has passed it by. Consider Brown v. Board of Education. An earlier case, Plessy v. Ferguson, decided by a Court fifty-eight years before, examined whether a state could require a railroad’s black customers to sit in a separate railroad car, segregated from its white customers. The Court answered that question yes, and in doing so, it established the legal doctrine of “separate but equal.” The case held that the Fourteenth Amendment’s equal protection clause permitted a state to segregate the races by law as long as it provided members of each race with equal facilities.1
In Brown the Court overruled Plessy and substituted for “separate but equal” the doctrine that segregated facilities are inherently unequal. The Court had to weigh the benefits of stability against those of change. On the one hand, the South had relied on the Plessy decision. Indeed, southern states had built not only schools but an entire society on racially segregated foundations. Their citizens had woven racial segregation into the fabric of their daily lives.2
On the other hand, the Court, the legal community, and much of American society had begun to see the Plessy decision as legally wrong and the segregated society it helped build as morally wrong. It is difficult if not impossible to reconcile racial segregation with the language and purpose of a Fourteenth Amendment that forbids “any State” to “deprive any person of … the equal protection of the laws.” Plessy’s rule was no longer in step with constitutional case law that had required the South to integrate its schools of law and of education. Nor was it in step with a society that, in its armed forces and elsewhere, had begun to embrace integration.3
Most important, it was clear by 1954 that Plessy’s rule had worked incalculable harm. That rule could not achieve its own stated goal. Instead, schools, parks, and public (and private) facilities were separate but hopelessly unequal. If Plessy had hoped that “separate but equal” would create a way station on the path to equality, that way station had become a terminal destination. It was impossible to see how a racially segregated nation could become a nation that equally respected all its citizens.
Thus, in Brown a unanimous Court overturned an earlier decision that the justices considered legally wrong, out of step with society and the law, and unusually harmful. Subsequently, the Court modified or overturned law set forth in a host of other cases, destroying rules that permitted racial segregation, and modifying the law of remedies, in order
to make its Brown decision effective.4
Directly overruling an earlier decision, as the Court did in Brown, is exceptional. Ordinarily, stare decisis is the rule. Lower-court judges, lawyers, clients, and ordinary Americans all need stable law so that judges can decide their cases, lawyers can advise their clients, clients can make decisions, and ordinary Americans can buy homes, enter into contracts, and go about their daily lives without fear that changes in the law will turn their lives topsy-turvy.5
A printed, circulated Supreme Court decision helps the judge, lawyer, client, and ordinary American know what the law is. By overturning a case, the Court can create uncertainty and undermine the reliance that bench, bar, and public have invested in the earlier decision. Moreover, the more the Court overrules earlier cases, the more it will gain a reputation for being willing to do so. And that reputation itself creates uncertainty. Is the legal material circulated authoritative? Will it remain so? Will legal changes undermine business, family, or social decisions? Will a new case that resolves uncertainty long remain the law, or will a new Court overturn it, thereby denying the public the advantages of the newer, “better” second decision for which the Court had hoped? At the same time, a Court that overturns too many earlier decisions encourages the public to believe that personalities or politics, not law, determine the outcome of Court cases. And that belief undermines the public’s confidence in the Court.
GENERAL PRINCIPLES
IF THE COURT should normally apply the principle of stare decisis but can sometimes overturn an earlier case, how does it know when to do which? The Court has referred to several factors that help answer the question. First, the Court has said that the principle of stare decisis applies more rigorously when a statute, rather than a constitutional provision, is at issue. That is because Congress can easily change a statutory decision, but neither Congress nor anyone else can easily amend the Constitution. Normally, the only practical way to change a constitutional decision is for the Court to reconsider it.6
Second, the public’s reliance on a decision argues strongly (but, as Brown shows, not determinatively) against overruling an earlier case. The public may well rely, for example, on a decision that affects property or contracts. Individuals and firms may have invested time, effort, and money based on that decision. The more the Court undermines this kind of reliance, the riskier investment becomes. The more the Court engages in a practice that appears to ignore that reliance, the more the practice threatens economic prosperity.7
Third, the more recently the earlier case was decided, the less forcefully the stare decisis anti-overruling principle should be applied. When only a short time has elapsed, we may not yet know that a decision will have harmful effects; it is also unlikely that either the bar or the public will yet have relied significantly upon the earlier case.8
Fourth, the Court can, and often should, overrule an earlier decision that has created a set of unworkable legal rules. Such a decision may have proved confusing or created legal conflict or otherwise caused serious harm. Confusion may mean no one reasonably relied on the case. In any event, overruling is more likely, on balance, to prove beneficial.9
Fifth, if case B has overruled case A, it is more reasonable for the Court to overrule B, thereby restoring A. That is because case B has already upset expectations and a restoration may not, on balance, cause further difficulty.10
Sixth, the Court should exercise particular caution before overturning a case that has become well embedded in national culture. In the 1960s, for example, the Court decided Miranda v. Arizona, which held that the police must warn a suspect of his constitutional rights to remain silent and to have a lawyer. Over the next few decades most Americans, through television or otherwise, became aware of this basic legal rule—that the police must warn suspects before questioning them. Then, in 2000, when the Court considered whether to overrule Miranda, it took account of the fact that the general public understood Miranda and had come to expect the police to follow its holding. And for that reason even members of the Court who thought Miranda wrongly decided have refused to overrule it.11 On the other hand, the law has long strictly regulated (and often forbade) expenditures made by corporations and trade unions to help elect candidates for public office. Yet, in January 2010 the Court (by a vote of 5 to 4) held that a congressional statute of this kind violated the First Amendment’s free speech guarantees. In doing so, the Court overturned two recent cases, and in the view of the dissenters (of which I was one), the Court disregarded a traditional legal view that stretched back as far as 1907.12
RESISTING THE TEMPTATION TO OVERRULE
THE ABOVE SIX principles emphasize harm, confusion, change, and lack of reasonable reliance as justifications for overruling an earlier case. But (as the examples just given, of Miranda-warnings and campaign expenditures suggest) they often do not dictate any particular result, they can conflict, they are highly general, and their application calls for the use of pragmatic judgment. That is, these principles can do no more than guide judges who might well disagree about how they apply in any particular case. Nonetheless, several special features of the Supreme Court offer a temptation to overrule earlier cases. In order to resist that temptation, when we apply our principles, we should ordinarily place a thumb on the scale in the direction of stability.
First, an earlier decision that the Court’s members now think wrong is not likely to change without their intervention. That fact is obvious when the earlier decision interprets the Constitution, for it is difficult to amend the Constitution. Doing so normally requires a favorable vote of two-thirds of each house of Congress and ratification by three-fourths of the states. And the nation has amended the Constitution only twenty-seven times. The difficulty of obtaining change is less obvious when the Court interprets a statute, but it is nonetheless true. In theory, Congress can write a new statute. In practice, Congress will do so only if it can find the legislative time and the necessary political will. It will probably find that political will only when change is politically popular or when well-organized groups make that change a major issue. Thus judges sometimes think that they themselves should overturn a case that they believe is wrong or when change will probably never take place.13
Second, the Court does not have many opportunities to overturn earlier cases. The Court fully hears only about eighty cases per year, very few of which require or permit the Court to reconsider a previously decided case the Court believes is wrong. Those that do involve such a case may fail to meet other criteria, such as a division of opinion among the lower courts, that determine whether the Court should hear a case. Thus justices who have before them a case that provides an opportunity to overturn an earlier decision that a number of them think is wrong know that the opportunity will not likely arise again. It is “now or never.”
Third, because life tenure for the justices means a Court membership that changes only slowly over time, it also means that different members appointed after long intervals by different presidents may well have different philosophical views. A president cannot control the votes of the Court during his limited time in office or even of the judges that he has appointed. President Theodore Roosevelt appointed the great justice Oliver Wendell Holmes, who almost immediately took an antitrust-law position totally contrary to Roosevelt’s views on the subject. And Roosevelt reportedly responded, “I could carve out of a banana a judge with more backbone.”14
Presidents and their judicial appointees are more likely, however, to share a basic philosophical approach to the country and to the law, and how they relate to each other. Thus without any Court judge compromising his or her total independence, different judges will have different philosophical approaches about how best to apply highly general constitutional terms such as “liberty,” “equal protection,” and “due process of law.” And judges on one Court may have different basic views on such subjects from their predecessors. It is consequently not surprising if a later Court considering an earlier case believes that the
earlier Court decision was absolutely wrong.
Members of the Court might reasonably think (1) “that earlier case is completely wrong, even untrue to the Constitution,” (2) “if we do not change it, no one will,” and (3) “it’s now or never.” I would add one further consideration, namely, that a case does not always unambiguously stand for a clear proposition of law. Consider a mother who tells a child not to bring up medical matters when Grandma comes because it will make Grandma think of her own recent illness. The mother has taught the child a rule of behavior, but she cannot describe its precise boundaries. Consider a case that holds that when a policeman properly arrests a driver, he can search the car without getting a warrant. What shapes the boundaries? Police safety, the need for a clear, simple rule, both? How does the case apply to a search after the arrested driver is handcuffed and put in the back of a squad car? My point is that stare decisis is not mechanical and fidelity to an earlier Court decision is often a matter of degree or interpretation.15
As a result, a judge will find it psychologically difficult to remain faithful to a decision with which he or she disagrees. A judge will almost inevitably resist adherence, for example, to an earlier interpretation that gives the Constitution a meaning that the judge is convinced the Constitution does not have. At the same time, overruling inevitably reflects a pragmatic decision-making process in which different judges give different weights to the individual factors that the Court has found relevant. Thus in one recent term, the members of the Court debated (dividing 5 to 4) whether they should overrule or depart from prior decisions in numerous areas of law, such as abortion, campaign finance, and antitrust. In all of these decisions I was in the minority in part because I thought the Court had overruled prior decisions and was wrong to do so.16
Making Our Democracy Work Page 18