When to Hold, When to Fold, and When to Reshuffle: The Art of Decisionmaking on a Multi-Member Court

When to Hold, When to Fold, and When to Reshuffle: The Art of Decisionmaking on a Multi-Member Court

Diane P. Wood

Judge, U.S. Court of Appeals for the Seventh Circuit; Senior Lecturer in Law, University of Chicago Law School. Judge Diane Wood of the United States Court of Appeals for the Seventh Circuit delivered the 2011-2012 Brennan Center Jorde Symposium Lecture at NYU Law on April 16

Introduction 

I. Traditions of Separate Writing

II. What Does a Separate Opinion Accomplish, and at What Cost?

A . Purpose of Opinion

1. Concurrences
2 . Dissents
B. Benefits of Separate Writing
C. Costs of Separate Writing

III. The Seventh Circuit

A . Principle-Based  dissents
B . Process-Based  dissents
C. Accuracy-Focused Dissents

Concluding Observations

INTRODUCTION

In History of the World.” Part I, King Louis XVI (played by Mel Brooks) reflects several times that “It’s good to be the king.”‘ What the king wants, he gets; he has only to say the word. I sometimes think that this is what it is like to be a district court judge, solo in the courtroom, mistress of all she surveys. True, the court of appeals is lurking around somewhere in the background, but let’s be realistic: only about 15 percent of cases are appealed,2 and the appeals court affirms the district court in most of them.3 The judges on the federal courts of appeals, in contrast, like their colleagues on virtually all other appellate courts in the United States, do almost all of their work through multi-member panels. By statute, those panels normally consist of three judges,4 unless a majority of the judges in regular active service vote to hear the case en banc. 5 A court of appeals judge thus cannot hope to get anything done without persuading at least one fellow judge to agree with her. At the en banc level, or at the Supreme Court, the problem is similar, but more complex. The Seventh Circuit, on which I sit, has eleven active judges when it is at full strength; the Ninth Circuit is authorized for twenty-nine active judges; and the federal Supreme Court of course has nine Justices. Given the challenging nature of many of the cases that come before these tribunals, it is no surprise that answers are not obvious and that people of good faith do not always agree on the rationale for a certain outcome, or even on the outcome itself.

I am well aware that I am not the first person to consider the ramifications of these uncontestable truths. My old boss, Justice Harry A. Blackmun, reflected on them when he administered the oath of office to me up in Spider Lake, Wisconsin, many years ago. Here is what he had to say:

Even though you will sit primarily with two other judges, as you sit in groups of three on the federal appellate bench, your vote will essentially be yours, and not theirs. There will be moments of a feeling of reward and satisfaction, and moments with a feeling of disappointment, and certainly moments of loneliness, despite the fact that you have a multiple judge court. Because that vote is yours, and only you can make it. Don’t let it discourage you.6

In essence, Justice Blackmun was describing the “hold ’em” strategy (with apologies to Kenny Rogers and his Gambler).7 Federal judges are surrounded with robust protections for independence-tenure during “good behaviour” and salary protection, to name the most prominent of them 8-and with that independence comes the obligation to vote in every case according to one’s best judgment. If that vote does not coincide with the votes of the other members of the panel or court, then the judge will sometimes believe that the only defensible option is to stand firm and, if necessary, to dissent.

As the Gambler suggests, however, “holding ’em” is not always the right strategy-not in poker, and not in courts. Sometimes, as the song suggests, the best thing to do is to acquiesce in the others’ views and save the disagreement for another day. That is what I’m calling the “fold ’em” option. Holding and folding do not, however, exhaust the options for a judge, even if they may for a cardplayer. An appellate judge often has an opportunity (to push the card metaphor about as far as it will go) to reshuffle the deck: the case that seemed to be about constitutional rights might be recast as one about how to present a complaint under Federal Rule of Civil Procedure 12(b)(6); the case that looked like an environmental showdown might instead be about which litigants have standing to sue under Article III of the Constitution; the case that initially presented a question about the rights of trespassers might be transformed into one about judicial federalism. The “shuffle” cases are the ones that may call for an opinion concurring in the judgment.

When to hold, when to fold, and when to shuffle are questions that many-including, notably, Justice Brennan-have addressed.9 Most of the attention has been devoted to the particular problem of dissent in the United States Supreme Court. Some of the observations made about the Court will apply with equal force to all courts of last resort as well as to intermediate appellate courts. But (as in the rest of life) one size does not really fit all-at least not very comfortably.

Because I sit on one of those intermediate appellate courts and have had an opportunity to observe its practices for more than sixteen years, my ultimate focus in this Essay will be on separate opinions at that level. From a certain perspective, the decisions of the courts of appeals are almost as final as those of the Supreme Court-in the Seventh Circuit, for instance, 3398 cases were terminated over the twelve-month period ending September 30, 2010;10 the Supreme Court reviewed five cases from the Seventh Circuit during October Term 2010.11 That sounds as if decisions from the courts of appeals enjoy considerable finality. But there is a world of difference between “almost final” and “absolutely final,” and that distinction has a profound effect on the considerations that impel judges on intermediate and final courts to write separately.

I.

TRADITIONS OF SEPARATE WRITING

Let me begin with a few words about the traditions of separate writing. Although the option to write separately has become firmly entrenched in American practice,12 it was not always thus, and it remains the case that a number of different models are actively used around the world. In most civil law countries, even the highest courts, acting through multi-member chambers, announce their judgments in a single, impersonal document. 13

This is thought to be consistent with the jurisprudential theory that lies behind the great code systems, under which the written law is supreme. For slightly different reasons, the European Court of Justice decided at its outset, in 1958, that it would eschew separate opinions. 14 Even with only six Member States (a far cry from today’s twenty-seven), there was a fear that the Court would splinter into six national fiefdoms if each country’s representative were to write separately. 15

By enforcing the single-opinion norm, the Court enabled itself to develop true European law without the baggage of the past dragging it down. 16 Finally, modem arbitral tribunals sometimes adopt the single-opinion approach, if they choose to explain themselves at all.’ 7 If dissent stays behind the door, then the parties may be more willing to accept the single-shot resolution of the dispute that arbitration provides.

The tradition across the Channel lies at the opposite end of the spectrum. For hundreds of years, the practice of the Appellate Committee of the House of Lords was that each judge would deliver his own speech, explaining the result he had reached. 18 (I say “he” deliberately, because it was not until October 24, 2003, that the first woman, Dame Brenda Hale, was made a Lord of Appeal.) 19 Since October 1, 2009, the Appellate Committee has ceased to exist; it was replaced with the new Supreme Court for the United Kingdom. 2° A look at the “decided cases” link on the Court’s website reveals that it has continued the practice of seriatim judgments that the Law Lords had used. 21 While the Justices do not go out of their way to reinvent the wheel, each one does explain his or her vote individually. 22

After a brief fling with seriatim opinions at the dawn of the republic, the U.S. Supreme Court abandoned that practice under the firm hand of Chief Justice John Marshall.2 3 As scholars have noted, the great Chief Justice “squelched dissent” to the best of his ability during his tenure.2 4 Over the two centuries that have followed, both the Supreme Court and the federal appellate courts have taken a middle path between the two extremes of seriatim opinions and impersonal public reports that reveal nothing about any internal dissension. The norm today is to produce an “opinion of the Court,” which signifies that a majority of the Justices or judges on the panel have approved the opinion, and that it will thus be precedential. Nevertheless, three other types of expression are also accepted: a dissent; a concurring opinion that joins the majority with supplemental observations; and an opinion concurring in the judgment that supports only the ultimate result, but not the majority’s reasoning. The highest state courts in the United States, as well as the Supreme Court of Canada, follow much the same practice.

While this may be what the U.S. courts are doing, it is another matter altogether to ask whether this is a desirable practice. Commentators have debated for years the question whether separate opinions are affirmatively good, a necessary evil, or something that should be avoided to the extent possible. Justice Brennan unapologetically thought that dissents “serve a very important purpose indeed., 2 6 The current Chief Justice, John G. Roberts, Jr., has expressed the opposite view, indicating that he regards dissent as a symptom of dysfunction 27 and that he has made it a personal priority “to discourage his colleagues from issuing separate opinions., 28

Nonetheless, if actions speak louder than words, there can be little doubt about the support for separate opinions at the Supreme Court. During October Term 2010, for example, the Supreme Court released eighty-two merits opinions, of which seventy-five represented plenary review, five were summary reversals, and two were nothing but one-line orders affirming by an equally divided Court. 29 The seventy-five decisions reached after full briefing and oral argument were accompanied by forty-nine concurring opinions and forty-seven dissenting opinions. 30 Only thirty-eight of the cases were unanimous.31 The rate of dissents in the courts of appeals is drastically lower: by one count, the percentage of dissents in the courts of appeals is just 2.6 percent (or a somewhat higher 7.8 percent of all published opinions), and the percentage of concurring opinions in published cases is a meager 0.6 percent.32 Before turning to the possible explanations for this difference, however, we ought to consider the pros and cons of separate opinions.

II.

WHAT DOES A SEPARATE OPINION ACCOMPLISH, AND AT WHAT COST?

When we turn to the question why a judge might choose to write separately, the distinctive positions of courts of last resort and intermediate courts become apparent. That is not to say that they have absolutely nothing in common. As noted earlier, the Supreme Court leaves untouched the vast majority of the decisions of the federal courts of appeals, as well as the highest state courts’ interpretations of federal law. Thus, a circuit judge contemplating a dissent is in some respects not too different from her colleague at the highest court. Some of the reasons that I have identified will thus apply to both levels; some will apply only to the intermediate court judge; and some will apply with particular force to the Supreme Court Justice.

A. Purpose of Opinion

The reasons why a judge writes separately vary, depending on whether the contemplated opinion will be a concurrence or a dissent. Judge Patricia Wald, who sat for many years on the U.S. Court of Appeals for the District of Columbia Circuit, has suggested that a judge who intends to dissent “is driven publicly to distance herself from her colleagues out of profound disagreement, frustration, even outrage.” 33 “Concurrences,” she wrote, “are a different matter…. Though certainly not as threatening as dissents, concurrences raise more collegial eyebrows, for in writing separately on a matter where the judge thinks the majority got the result right, she may be thought to be self-indulgent, single-minded, even childish in her insistence that everything be done her way.”34

With that said, Judge Wald offered five reasons why judges concur: first, to reassure the losers that the result is not as bad as it seems; second, to take issue with the majority’s rationale; third, to support the majority’s result with additional reasons; fourth, to warn that future cases might need to take a different path; and finally, to cabin the scope of the majority’s opinion. 35 All of those reasons, both for the concurrences and the dissents, strike me as plausible. And these are all motivations that would apply with equal force regardless of the court’s position within the judicial hierarchy. I would like to amplify on the reasons Judge Wald highlighted, first at a broad level of generality, and then with a focus on incentives that are unique to the intermediate court judge.

1. Concurrences

Looking first at concurrences, I see a number of reasons why a judge would take the trouble to write either a concurrence, or a concurrence in the judgment. First, the writer might want to set the factual record straight. As descendants of the legal realism movement, most judges and scholars know that it is normally impossible, in an appellate opinion, to set forth every single fact that has been developed in the case. Selectivity is inevitable, and in many cases it is entirely uncontroversial. Sometimes, however, the account of the facts proposed by the majority will appear incomplete or even misleading, and the writer might feel the need to reshuffle the deck. For example, the majority’s opinion might be written in a way that suggests that a search warrant was supported by probable cause, but the concurring judge, who would affirm on an alternate ground, may feel that the record cannot support such an interpretation. A more complete look at the facts may show this rationale to be problematic. A fuller picture of the facts will also provide fodder in future cases for distinguishing the decision at hand and for taking a more cautious, step-by-step approach.

Another goal of a concurrence might be to record disagreement with the legal rationale used to support the agreed result. It is often the case that the parties have given several reasons for the result they urge. And the difference among these theories is often important. For example, there are very important distinctions between affirming a criminal conviction for a harmless Evidence Rule 404(b) error, and a lack of error altogether. Furthermore, deciding whether a prison warden did not violate the Free Exercise Clause when she compelled inmates to enroll in Narcotics Anonymous and thus to affirm the existence of God as a condition of parole eligibility, or whether the rules were merely sufficiently unclear so as to grant qualified immunity, has serious and important consequences for future, similar cases. There is a similarly vast difference between denying a petition for habeas corpus as untimely and because the state court’s resolution of the constitutional issues was not unreasonable. Each of these examples illustrates how the concurring judge may feel the need to refocus the rationale behind the decision-to shuffle, in essence-in the interest of shaping the trajectory of the law.

Next, the concurring judge may be trying to establish the limits of the lead opinion, either for her colleagues, or for other courts, or for the bar, or for the public at large. Concurrences, like dissents, might be directed to any or all of these audiences, and at the Supreme Court level, it is safe to assume that the target is usually “all of the above.” A concurrence might emphasize a pivotal fact that either went unmentioned or was minimized in the lead opinion; it might place an otherwise sweeping ruling into a recognized legal structure (and thus incorporate the limits of that structure); or it might be intended to throw a spotlight on the opinion for the general public, in the hopes that others will point out the need to limit the court’s holding.

Even if the concurring judge concludes that the law compels the result the majority has adopted, she may consider that result to be an unfortunate one. She may thus write a separate opinion in the hopes of enhancing the chance of external correction through legislative or administrative action. This is a milder form of “holding” than dissent, but it nevertheless signals a serious disagreement. Such a concurrence might point out that a particular result is required only because of the way the court has been interpreting a particular statute or regulation. Courts occasionally say that the litigants should take their pleas next time to the branch of government that is empowered to act on the problem. If the majority is content with the outcome but the concurring judge is following it only because she believes herself to be constrained by the law, a concurrence may be the best option.

Finally, the concurring judge may simply wish to make a record for the future (or, as Justice Felix Frankfurter put it rather grandiosely, take the opportunity to “record prophecy and shape history” 36). The majority may have been content to write a narrow, workmanlike opinion, while the concurring judge may believe that a much broader principle is at stake. For example, the majority may reject a claim brought by a gay man on the ground that Title VII’s prohibition of sex discrimination in employment does not extend to sexual orientation, while a concurring judge may wish to acknowledge that this is the present state of the law but to argue that such a distinction is illogical. Or, borrowing from the facts of Kelo v. City of New London,3 7 the majority might take the position that a particular urban redevelopment project satisfies the “public use” requirement of the Takings Clause of the Fifth Amendment, but the concurring judge (as Justice Kennedy did in the actual case 38) may believe that the project in question passes muster only because it was one part of a comprehensive development plan in an economically distressed city and there was no evidence that it was a subterfuge designed solely to move the property from one set of private hands to another.

2. Dissents

As Judge Wald suggested, from a certain point of view dissents are easier to understand, whether one adopts a judicial process perspective, as she did,39 or the empirical approach advocated by scholars such as Lee Epstein, William Landes, and Richard Posner. 40 Looking at the question in more detail, we can see a number of goals that the dissenter (who by definition is “holding” rather than acquiescing or trying to salvage some agreement by shuffling) might hope to achieve.

First, the dissenter may wish to flag an error for either a higher court or for the public. In some instances, it might be enough to pinpoint the flaw (for example, the use of the wrong standard for dismissing a complaint, or the failure to appreciate how principles of trust law manifest themselves in ERISA cases, or the willingness to overrule a case on which the majority relied). In others, the dissenter is simply so convinced that the majority is wrong that her only goal is to be on record saying so.

The dissenter also may want to throw a spotlight on critical facts that the majority either has ignored or downplayed-facts that the dissenter believes compel the opposite result. Both of the dissents filed in DeShaney v. Winnebago County Department of Social Services recast the facts of Joshua DeShaney’s severe abuse and his interactions with the relevant public agencies and explain why, in the writers’ views, the law required a different response. In a different way, Justice Thomas’s dissent in Virginia v. Black,42 the cross-burning case, used a rich factual account of the background of that terrible practice to explain why he regarded it as violent conduct, not protected expression. Similarly, Justice Breyer’s dissent and associated appendix in Eldred v. Ashcroft.

Sometimes the dissenter will consider it of paramount importance to explain why agreement with the majority’s approach and result is impossible as a matter of principle. Justice Holmes’s famous dissenting opinion in Lochner v. New York 44 fits this bill. As he said, “I strongly believe that my agreement or disagreement has nothing to do with the right of a majority to embody their opinions in law.”45 More recently, we have Justice Stevens’s dissenting opinion in Seminole Tribe v. Florida, 46 in which he wrote that “the importance of the majority’s decision” about Congress’s inability to abrogate state sovereign immunity through legislation enacted pursuant to its commerce powers “cannot be overstated.” 47 Dissents of this kind, as Justice Brennan explained, can become what he called “the repeated dissent in which a justice refuses to yield to the views of the majority although persistently rebuffed by them. ‘ 48

The dissenter may write in an effort to prevent the majority from sweeping under the rug additional facts in the record or the implications of its ruling-in a sense, to hold its feet to the fire. This appears to be one of Justice Ginsburg’s goals in her dissenting opinion in Gonzales v. Carhart 49, the case that considered and upheld the constitutionality of the so-called Partial-Birth Abortion Ban Act. Concerns about the institutional consequences of the majority’s opinion were what Justice Kennedy emphasized in his dissenting opinion in Lee v. Kemna,50 a case that dealt with the somewhat dry but vitally important question whether a state ground is adequate to prevent federal habeas corpus review. Taking a pragmatic view, the majority had ruled that the petitioner had done all that he could and thus could proceed with his federal case despite his technical noncompliance with a state procedural rule. 51 Justice Kennedy objected that this overlooked the new unwarranted burden that this placed on state trial judges. 52

Another reason to write is to show why a seemingly broad majority opinion ought to be viewed in a much narrower light. (This comes close to a shuffling strategy.) Brief though it was, Justice White’s dissenting statement in Wallace v. Jaffree,53 the decision that invalidated Alabama’s statute calling for a silent prayer at the beginning of the school day, takes this form. He suggests that the majority would have permitted a slightly reformulated statute that provided only for a moment of silence but did not mention prayer.54 In later years, state legislatures were quick to pick up on this suggestion.

Lastly, the dissenter may consider it imperative to write for the future (whether she believes this to be a near-term or a longer-term prospect). This normally happens when the dissenter has a clear vision of the meaning of a particular constitutional provision, as Justice Brennan did for the First Amendment 55 or Justice Scalia does with respect to the question of abortion. 56

Justice Blackmun took this approach in his well-known dissent from the denial of certiorari in Callins v. Collins,57 in which he announced that “from this day forward, I no longer shall tinker with the machinery of death., 58 Using sweeping language, he explained that he had come to the conclusion that no combination of procedural or substantive rules could ever “save the death penalty from its inherent constitutional deficiencies.” 59 His appeal was expressly to the future.

Let me now turn to some special considerations that influence the intermediate appellate judge. Justices who publish separate opinions are always addressing the future-maybe the distant future, maybe the immediate future, but never the immediate outcome of the case before the Court. That battle has been waged and resolved. The war is far from over, however, for the judge on the court of appeals. The prospect of an immediate payoff from a dissent or concurrence, while still not great, is at least a positive number. But the court of appeals judge does not need to rest all of her hopes on that thin reed.

For one, the separate opinion may persuade the panel to grant rehearing and eventually reverse course. This is what happened in Board of Regents of the University of Wisconsin System v. Phoenix International Software, Inc.,61 where the initial panel opinion had held over my dissent that the University of Wisconsin enjoyed sovereign immunity against certain trademark counterclaims that Phoenix had filed. On panel rehearing, the court unanimously ruled that by choosing to file its challenge to a decision of the Trademark Trial and Appeal Board in the district court, the state had waived its immunity by litigation conduct.62

Alternatively, the separate writer may hope to set the stage for rehearing en bane according to the circuit’s procedures. These procedures vary among the circuits. In some, en bane consideration of the case is possible only after the panel has published its opinion. 63 In the Seventh Circuit, it is also possible for a panel (either unanimous or with a dissenter) to seek an en bane vote before publication, if the panel is proposing to overrule earlier circuit precedent, to create a conflict in the circuits, or to rule on an exceptionally important question. 64 No matter when the en bane vote is taken, the dissenter has a chance to persuade her colleagues to adopt her position.

If all recourse at the circuit level has been exhausted-that is, if both panel rehearing and rehearing en bane have been rejected-the dissenter knows that her views will still be available and possibly influential with two last constituencies: first, other courts of appeals, and second, the Supreme Court. Here we need to distinguish between a judge who was a member of the panel, and who presumably already has published her separate views, and a judge on the court who was not on the panel but who participated somehow in the en bane process. For someone who was a member of the panel, normally there is little else to say after rehearing en bane is denied. Other judges who have participated only in the en bane votes decide occasionally to add their voices to the mix by filing an opinion dissenting from the denial of rehearing en bane. 65

These judges operate under the disadvantage of not knowing the record as well as those who have been living with the case. With enough effort, they can certainly bridge that gap, since the briefs will be available to every member of the court and in this day of electronic records, much of what went on before will also be readily retrievable. Without going the extra mile, however, any such separate writer runs the risk of missing important features of the case.

B. Benefits of Separate Writing

It seems safe to assume that every judge who considers writing separately engages in a preliminary assessment of benefits and costs of doing so. Writing for the future looks in many ways like a selfless act, or at least one in which the return from the investment made in the writing is likely to be a long time in coming. But the success record of dissenting opinions at the Supreme Court level is good enough that people continue to write them. In so doing, they reap not only the satisfaction of a clear conscience, but also the hope that they too will be vindicated someday. And that vindication takes several forms: sometimes the Constitution is amended; sometimes legislation addresses the dissenter’s concern; and sometimes the Court itself changes course and endorses an earlier dissent. The prospect of any of these forms of success may be enough for the dissenter to decide to stand firm on her views.

Because the Constitution has been amended only twenty-seven times-or, if you prefer, only seventeen times not counting the Bill of Rights-it is obvious that Supreme Court dissents have not led to many constitutional amendments. But they have provided strong support for more than one might think. The clearest example remains the Eleventh Amendment, ratified on February 7, 1795, which as we all know was quickly introduced 66 and passed in the wake of the Supreme Court’s decision in Chisholm v. Georgia.67 That case held-over Justice Iredell’s dissent-that a citizen of South Carolina was entitled to sue the State of Georgia over that State’s protestation that it was a sovereign entity immune from suit. (Interestingly, the Supreme Court decided Chisholm during the brief period when it used seriatim opinions; 68 query whether the reaction to the decision would have been the same if the Court had presented a unified opinion with one dissent.) In terms, the Eleventh Amendment withdraws the language in Article III, Section 2 that extends the judicial power of the United States to suits against a state brought by citizens of another state or of a foreign country. 69 In substance, the Supreme Court has understood the Amendment to symbolize a much broader commitment to state sovereignty. 70 But the Eleventh Amendment does not stand alone. Eighteen years after the Supreme Court decided Pollock v. Farmers’ Loan & Trust Co.,71 which struck down a general income tax as an impermissible direct tax, the Sixteenth Amendment took effect and conferred the power on Congress to impose an income tax without regard to apportionment among the states or to the census. 72

The Twenty-Fourth Amendment, forbidding the payment of any poll tax as a condition of voting in a federal election, did not strictly overrule any decision, 73 since Breedlove v. Suttles74 (later overruled by Harper v. Virginia State Board of Elections on equal protection grounds75) upheld the poll tax only for state elections. In contrast, the Twenty-Sixth Amendment, which conferred the right to vote in both federal and state elections on eighteen-year-olds, 76 overruled Oregon v. Mitchell in part and thereby vindicated Justice William Douglas’s position (which had rested on both the Equal Protection Clause and the Privileges and Immunities Clause of the Fourteenth Amendment).77 Thus, once in a great while, dissenting positions have won the day through constitutional amendment.

Even though the Supreme Court hears a disproportionate number of constitutional cases, as measured by the overall volume of such cases in the federal system, constitutional matters do not consume the entire docket. Cases involving legislation are common, and this is an area in which the dissenter stands a comparatively better chance of prevailing within a reasonable time. Even if one’s colleagues remain steadfast, there is another crucial audience in statutory cases: the legislature. And it is easy to recall examples of legislative overruling of Court decisions. 78

When the Supreme Court ruled in 1986 that the First Amendment did not prohibit the Air Force from applying a general regulation mandating uniform dress to an orthodox Jewish officer (who happened also to be a rabbi)-in other words, forbidding the officer from wearing his yarmulke-it did so over dissents from Justices Blackmun, Brennan, and O’Connor.79 Two years later, Congress included a provision requiring accommodation of religious apparel in the 1988 Defense Authorization Bill.80 Someone obviously was paying attention. Similarly, in Public Employees Retirement System of Ohio v. Betts,s 81 the Supreme Court rejected a reading of the Age Discrimination in Employment Act (ADEA) that would have required age-based distinctions in employee benefit plans to be justified on a cost basis. It held instead that a plan could not be regarded as a subterfuge unless the plan was intended to serve a discriminatory purpose. Congress responded just one year later with the Older Workers Benefit Protection Act,82 which amended the ADEA to adopt the cost-justification rule. The same quick action/reaction pattern occurred more recently when the Supreme Court ruled 5-4 against Lilly Ledbetter’s claim of sex discrimination on the ground that she had waited too long to sue.83

Congress speedily passed the Lilly Ledbetter Fair Pay Act of 2009, 84 codifying the rule that a new act of discrimination occurs with each tainted paycheck.

Finally, with the Civil Rights Act of 1991,85 Congress responded to a number of Supreme Court decisions in a way that largely vindicated the dissenters. 86

Dissents sometimes make their way into the law through the simple expedient of the Court’s overruling the earlier decision and adopting the dissent’s view. Of the numerous examples one could cite, I will refer to only a few. Pride of place must go to the first Justice Harlan’s dissenting opinion in Plessy v. Ferguson,87 refusing to accept the “separate but equal” theory for de
jure race discrimination and paving the way for the Supreme Court’s flat rejection of that theory in Brown v. Board of Education.88 Next, Justice Brandeis’s dissent in Olmstead v. United States,89 rejecting the Court’s holding that surveillance without any trespass, physical penetration, or seizure of a material object fell outside the ambit of the Fourth Amendment’s restrictions on searches and seizures, became the law in Katz v. United States.90″ The Katz Court observed that “once it is recognized that the Fourth Amendment protects people-and not simply ‘areas’-against unreasonable searches and seizures, it becomes clear that the reach of that Amendment cannot turn upon the presence or absence of a physical intrusion into any given enclosure.” 91

Sometimes it does not take forty to sixty years for a dissent to take hold.

In Seminole Tribe v. Florida, 92 decided in 1996, the majority overruled the Court’s seven-year-old decision in Pennsylvania. Union Gas Co93., which had upheld Congress’s power to abrogate state sovereign immunity through the exercise of its authority under the Commerce Clause. It took a comparable time for the Court to conclude in Caterpillar,Inc. v. Williams94 that Justice Brennan had been correct after all in Federated Department Stores, Inc. v. Moitie95 about the removability to federal court of certain claims. And, in the finals for the prize for speedy reconsideration, we have the three-year turnaround in the flag salute cases from the World War II period-from Minersville School District v. Gobitis,96 where the Court ruled that school children objecting on religious grounds could be compelled to salute the American flag and recite the pledge of allegiance, to West Virginia State Board of Education v. Barnette,97 where the Court overturned Gobitis and held that such compulsions were unconstitutional.

C. Costs of Separate Writing

For all that one might say in defense of separate writing, there is no doubt that there are costs associated with the practice. These costs may lead the outlier judge on the panel to “fold. 98 The leading cost is an obvious one: it takes time and energy to write separately, and the payoffs from separate writing are few and far between. While Justice Brennan dissented repeatedly in a number of areas such as the death penalty,99 habeas corpus, 0 0 state sovereign immunity, 01 due process for alleged Communists, 102 the right to freedom of expression,’ 103 and the Fourth Amendment, 104 perhaps the workload at the Supreme Court is manageable enough that the time allocated to writing dissents does not eat into time that is needed for the preparation of majority opinions. At the court of appeals level, however, the decision to write separately must be weighed carefully. Circuit court judges have no authority to turn cases away, and time therefore must be rationed carefully. A dissent or concurrence has to pass an unstated bar before the judge will undertake to write it.

Other costs exist as well, although they will arise in varying degrees in different cases. Some dissents have the effect of painting the court as just one more political institution-but a scary one that is populated by unelected 105 officials with life tenure. The best example remains Bush v. Gore.’ Whatever one thinks of that case, now that more than a decade has passed, there can be no doubt that many still see it as a highly political decision. 106 Indeed, the various separate opinions, which charge that the Court’s decision to hear the case and issue a ruling have undermined “the Nation’s confidence in the judge as an impartial guardian of the rule of law,”‘ 107 have given ammunition to those who continue to decry the outcome as politically motivated. 108 Looking at this issue from the other end of the telescope, recall Chief Justice Earl Warren’s efforts to ensure that Brown v. Board of Education would be issued as a unanimous decision, to deflect any concern at all that it was driven by politics rather than law. 109 A related cost arises when separate opinions become too passionate in tone, or worse, degenerate into ad hominem attacks on other judges on the court. The public rightly may question whether something larger-the Constitution, a statute, or legitimate common law-compelled the result, when matters become so personal. Such separate opinions inflict a cost on the judiciary, in the form of a general loss of respect, that goes well beyond the immediate case.

Separate opinions can also muddy the waters in a way that underscores the lack of a clear rule in an area. Think of the many cases in the obscenity field during the 1960s that failed to command a majority for any position. 110 The same was true for years for Establishment Clause cases.111 ‘ More recently, even something as seemingly technical as the constitutional rule governing assertions of personal jurisdiction over an out-of-state company has proven devilishly hard to settle.’ 112 The lack of clarity naturally has the unfortunate effect of fomenting further litigation.

Two remaining costs can arise even if the court has managed to steer clear of the risks of politicization and indeterminacy. One is hard to observe from the outside, yet is quite important to the healthy functioning of a multi-member tribunal: separate opinions may create tension among the members of the court. Either side-the majority or the dissent-may feel personally attacked even if the other was striving to keep matters strictly professional. That risk may ebb if the dissenter becomes branded as a frequent complainer about one or more issues. But that raises a different problem: the dissenter will have lost credibility and may be disregarded altogether. Most judges will therefore think carefully before writing separately, even if they sincerely disagree with some or all of the proposed opinion.

III.

THE SEVENTH CIRCUIT

I now turn to an account of my own experience on the Seventh Circuit. Let me begin with a few statistics. Between June 30, 1995-the date of my commission-and October 2011, I sat on approximately 4500 panels.113 Those panels produced 2569 reported decisions. (Note that the ratio of reported cases to all cases is a bit more than 50 percent, which has been typical of the Seventh Circuit throughout the time I have been on the court.) Over that time, I wrote a grand total of eighty-five separate opinions; eighty-four of them were in reported decisions, and one (in 1996-perhaps before I had settled into a standard practice) was in an unreported case.’ 114 Forty-one of the separate opinions were dissents (forty in published opinions, plus the odd unreported dissent), twenty-six were concurrences, and eighteen were concurring in part, dissenting in part. In other words, I wrote separately in 3.3 percent of the reported cases in which I participated, and 1.9 percent of all cases. In another twenty-eight cases, I joined another judge’s separate opinion; this amounts to 1 percent of the reported cases and 0.6 percent of all cases.

Most cases that come before the courts of appeals are relatively straightforward. Any three judges would probably decide them the same way, and so it is not surprising if the three judges who were randomly chosen for the panel are unanimous, no matter which president nominated each one, what prior political affiliations the person had, or what other personal experiences each individual brings to the process. But with some regularity, an appeal comes along that presents more difficult questions-difficult because the law is unsettled, difficult because judicial philosophies differ, or difficult because both of the former two problems are overlaid onto a complex factual record.

In my case, this process begins immediately after our postargument conference on the case. I will know at that point whether I am likely to be in the majority or the minority, and I will also know whether the theory that I favor or my understanding of the facts is consistent with, or opposed to, the views of my colleagues. I take notes about those views as we are discussing the case, and I keep those notes for future reference. Assuming that I am not assigned to write the opinion, I normally wait until the proposed majority opinion is circulated before I do anything. This is a potential time saver, since I may be persuaded by the full exposition now in front of me, or on (rare) occasions the authoring judge will explain that the result for which he voted tentatively at the conference did not withstand closer scrutiny of the record or further research.

Once I have the proposed opinion in hand, I review it to see whether the disposition seems correct to me. If the general tenor of the opinion seems acceptable (even though it might not have been written the way I would have done it), but I have a few remaining concerns, I am likely to circulate a detailed note to the author explaining what changes I would like to see. I try to be as specific as possible in these notes, although I also take care to leave room for the author to respond as he or she sees fit. If I suggest proposed language, I will normally preface the suggestion with a phrase like “something along the following lines, or its equivalent, would address my concerns.”

If I am fairly sure that I will dissent, my practice varies. Sometimes I simply circulate a note saying that I regret that I cannot approve the proposed opinion, and that I am in the process of drafting a dissent. In other cases, I may send a comment to the panel saying that I have serious problems with the proposed opinion but that I am happy to review anything further that the author may wish to send around, if the author thinks that my issues can be addressed in a mutually satisfactory way. The Seventh Circuit has a rule, honored somewhat in the breach, under which each judge is to give priority to the other judge’s case once a draft has been circulated. This means that there is at least some social shaming that goes on if one waits too long to respond to an opinion with a draft dissent or concurrence. After the separate opinion is circulated, there are often a few more rounds in which each side polishes its statements; at the end of that process all opinions in the case are e-mailed to the printer.

Looking at my own dissents, I can identify three primary reasons for my decision to write separately. (Any one of these, or all three together, may play a role in a particular case.) In some instances, as Judge Wald indicated and I have already discussed, I dissent on the ground that a broad question of law or principle is at stake and I am not willing to compromise it-only the “hold” strategy will do. In another group of cases, I again believe that “holding” is necessary, although in this instance it is largely for institutional reasons. Sometimes I believe that the majority’s rule will fail to give necessary guidance to the bench and bar; sometimes I think that it will impose difficult procedural burdens on litigants or other affected parties. Sometimes I conclude that the majority is applying the wrong standard of review. My third reason to dissent stems from the fact that the courts of appeals are the last tribunal in our system with responsibility for error correction. Just because a case does not seem
destined to be the next Gideon v. Wainwright 1 15 does not excuse the court of appeals from ensuring the integrity of the facts and their application to legal principles. With that in mind, I will sometimes dissent or concur just for the sake of ensuring that the facts are properly understood.

A. Principle-Based Dissents

The set of dissents that are normally discussed in connection with the Supreme Court are those that raise a broad question of law or principle. These are the cases that are worthy of en banc review, or the attention of judges in different circuits, or Supreme Court attention. Approximately one-third of the separate opinions that I have filed (according to my own undoubtedly biased effort at classification) fall into this category. Some examples illustrate this point.

In Haas v. Chater, a mother brought an action seeking child insurance benefits under the Social Security Act for her son, Scottie, who had been born out of wedlock eight months after his father’s death.’ 16 The Social Security Act contains a number of provisions designed to permit proof of paternity when the parents were not married, but the only one available to Scottie was proof that he was entitled to inherit from the wage earner under the law of intestate succession of the wage earner’s state of domicile-in this case, Indiana. 117 As I implied, Scottie’s mother was just one month along in her pregnancy when his father was tragically killed in an automobile accident. One month after his birth, the mother filed a petition in Indiana state court to establish paternity, and the court granted that petition.” 8 The estate later sued the other party to the accident for wrongful death, and the court in that action determined that Scottie was entitled to the proceeds of that suit. 19 Nevertheless, the majority ruled that Scottie had failed to establish that he would have been entitled to inherit under Indiana’s intestacy rules because his parents never married and paternity was not established during his father’s lifetime or within five months after his father’s death. 20 The majority, expressing a concern about theoretical collusive paternity suits, rejected the argument that the five-month deadline was simply a statute of limitations that was waivable, even though it acknowledged that Indiana courts have sometimes so construed deadlines for bringing paternity actions. In ruling that way, the majority relied heavily on the Supreme Court’s decision in Lalli v. Lalli,121 which held constitutional a state statute that required a court order of filiation before the putative father’s death.

I found this to be an unwarranted extension of Lalli, which had not dealt with the problem of a one-month fetus’s duty to establish paternity before his father’s death. I also believed (then and now) that the majority’s decision conflicted with a line of Supreme Court decisions dealing with the status of illegitimate children, including Mills v. Habluetzel,122 Pickett v. Brown,12 3 and Clark v. Jeter.124 I thought it possible, if not likely, that Indiana might toll its five-month limitation period for a child in Scottie’s position, and I noted that Indiana (like many states) did not recognize an independent tort cause of action for the child until the moment of live birth. Even if the state would not have adjusted its rules in any of these ways, however, I thought that the majority had reached the wrong result and no compromise was possible. Such an intersection of Indiana’s law with the requirements of the Social Security Act under these unusual facts, I thought, would draw an unconstitutional distinction between legitimate and illegitimate children. After I wrote that dissent, the Seventh Circuit vacated the majority’s opinion and granted rehearing en banc. 2 5 The enbanc court divided evenly, and so the ultimate outcome of the case was to affirm the district court’s order denying benefits by an equally divided court; 126 the Supreme Court later denied certiorari.127

Holman v. Page is an example from a criminal procedure case in which I dissented from the denial of rehearing en bane.128 That case presented the question whether defective performance for purposes of a claim of ineffective assistance of counsel could ever be shown when the lawyer failed to file a motion to suppress evidence that had been gathered in violation of the Fourth Amendment. 129 The majority had reasoned that such claims were off the table,

30 Fourth Amendment claims may not be because Stone v. Powell1 holds that

raised in habeas corpus proceedings unless the state denied a full and fair opportunity to litigate them. 131 I believed that the majority’s position flatly conflicted with the Supreme Court’s decision in Kimmelman v. Morrison,12 which had held that defendants may properly base their Sixth Amendment ineffective assistance claims on an assertion that the attorney failed properly to litigate a Fourth Amendment issue. Although the Supreme Court denied certiorari in Holman,133 the Seventh Circuit later overruled Holman in Owens v. United States,134 not only because of a reconsideration of the holding of Kimmelman, but also because in the intervening eight years, not a single other court of appeals in the country had been persuaded to follow it.1 3 5

My next example comes from a case that was argued before the en bane court: United Phosphorus, Ltd. v. Angus Chemical Co.136 It raised a question near and dear to the hearts of procedure buffs: do the provisions of the Foreign Trade Antitrust Improvements Act of 1992 (FTAIA) describing the necessary effects on U.S. commerce address the subject-matter jurisdiction of the court, or do they merely layout elements of a proper Sherman Act claim? 137 By a 5-4 vote, the en banc court opted for the subject-matter jurisdiction interpretation. 138 I wrote a dissent arguing that the requirement to show certain effects on commerce is simply an element of the case.139 I offered a number of reasons for that conclusion, including the fact that the FTAIA itself says nothing about “jurisdiction,” the fact that the subject-matter jurisdiction approach is inconsistent with the history of international antitrust enforcement, and the fact that I read the Supreme Court’s decision in Steel Co. v. Citizensfor
a Better Environment14 0 to mandate an “elements” reading. 14 1 In August 2011, the Third Circuit issued an opinion in which it adopted my views in United 142 thereafter, the same issue came before a panel of the Phosphorus. Shortly Seventh Circuit, in the case of Minn-Chem, Inc. v. Agrium, Inc. 143 Although the panel found that the plaintiffs had failed to meet the requirements of the FTAIA, the en banc court decided to rehear the case. Taking note of a number of intervening Supreme Court decisions that had supported the “elements” characterization, the full court (in an opinion that I authored) unanimously decided to overrule United Phosphorus and adopt the position that I had urged there in dissent. 44 The en banc opinion also analyzed the requirements of the FTAIA and concluded that the plaintiffs were entitled to move ahead with their challenge to a global potash cartel.

No one writes a dissenting opinion under the illusion that vindication is inevitable. I have already given one example, Haas, in which my dissent had no effect for the case at hand (or possibly for anything else either). Another one was Crawford v. Marion County Election Board,145 a case dealing with the constitutionality of mandatory voter identification laws. The majority had upheld an Indiana law requiring voters to present state-issued photo IDs as a prerequisite for voting, over a strong dissent from my late colleague Judge Terence T. Evans.’ 46 A petition for rehearing en banc was filed and rejected by a 7-4 vote. 14 7 I wrote for the dissenting judges to explain why I thought the majority erred by failing to apply strict scrutiny to legislation that would place such a significant burden on the right to vote of specific and identifiable groups of voters. 148 The Supreme Court took the case, but in the end it found that the challengers had failed to present enough evidence to overcome the state’s concerns about possible fraud in elections.149

Sometimes, however, a dissenting position is vindicated. My last example of a dissent motivated by a conviction that the majority’s view of the law was incorrect is Bloch v. Frishholz,150 known around Chicago as the mezuzah case.

The case was brought by Jewish residents of a condominium in Chicago who complained that the condo association was enforcing a rule forbidding any objects to be left in the hallways in a way that amounted to intentional discrimination on the basis of religion.’ 5 1 Taking the position that this rule reached even small religious items like a mezuzah, affixed to a doorjamb at shoulder level, the condo representatives repeatedly took down the Bloch family’s mezuzah152 (even on one occasion while the family was at the funeral of a family member, and they discovered to their dismay that the mezuzah had been taken down when they returned home for shiva with the rabbi).’53 There were several important legal issues in the case: first, whether and to what extent the Fair Housing Act reaches discrimination that occurs after the moment of sale or rental;’ 54 and second, whether (by analogy to Establishment Clause cases) the undisputed facts showed that this was a neutral rule like the one the Supreme Court upheld in Employment Division v. Smith,155 or if instead there was at least a factual question whether the rule was adopted to target an unwanted group, like the law at issue in Church of Lukumi Babalu Aye, Inc. v. Hialeah.156 The majority took a very narrow view of the first issue, holding that the Act applied only if the postsale restrictions were tantamount to eviction; 157 it held on the second point that the hallway rule was neutral. I dissented; 159 the court decided to rehear the case en banc; and the en banc court unanimously ruled in favor of the plaintiffs on most of their claims. 160

B. Process-Based Dissents

Turning now to the dissents motivated by a concern that the law as stated by the majority was not clear, that the standard of review was incorrect, or that there was a problem with the record, I begin with a case dealing with a somewhat obscure but nevertheless important point about the Railway Labor Act. The case was Brotherhood of Maintenance of Way Employees v. Atchison, 161 which raised a question about the process Topeka & Santa Fe Railway Co.,that had to be followed to resolve a dispute between a union and the railway company. The underlying issue was about the extent to which the railroads had to compensate their maintenance workers for travel expenses.’ 62 If the governing collective bargaining agreement did not address this issue (in other words, if it was “new”), then the Railway Labor Act labeled it a “major” dispute. 63 Major disputes potentially take a long time to resolve: the parties must first bargain, then use mediation if necessary, and if all that fails, they may resort to raw economic force. 16 4 If the dispute requires only the enforcement of rights already addressed in the collective bargaining agreement, then it is deemed a “minor” dispute, and the parties must go to binding arbitration.165 The Act forbids strikes over minor disputes.’ 66 The railroads took the position that the dispute was minor, while the union urged that it was major. Without recourse to the bargaining history, the majority said that the railroads’ position that the dispute was minor was “frivolous or obviously insubstantial.”‘167 It chose, however, to look at extrinsic evidence despite this acknowledgement about the plain language of the agreement.68 My inability to go along was grounded in institutional concerns. I thought that the statute contemplated a straightforward and efficient process of sorting disputes into the “major” and “minor” categories; insisting that extrinsic evidence had to be consulted even if there was no threshold finding of ambiguity or vagueness in the agreement was contrary to that goal.

My next example comes from a decision that, in my view, muddied the standards for appealing orders granting or denying motions to compel arbitration: Napleton v. General Motors Corp.169 This was a commercial dispute over a lease; the lessor, Napleton, took the position that the lease had not been renewed properly and thus that the lessee, General Motors, should vacate the property, while GM thought that the renewal was in order and refused to leave.’ 70 In time, Napleton sued in state court, GM removed to federal court, and then GM argued that an arbitration clause governed the whole thing.’ 71The district court agreed with the last of these points and issued an order granting GM’s motion to dismiss without prejudice to allow for arbitration. 172 Napleton then appealed and was met with the argument that there was no jurisdiction over the appeal. 173 The majority agreed that jurisdiction was defective, not because of the lack of a final judgment, but because there was an additional question that had to be resolved: Was the motion for arbitration “embedded” in a larger, substantive suit, or was it an “independent” motion where the request for arbitration was the sole issue before the court? 174 Even after a judgment of dismissal, the majority concluded, the result had to be viewed as nonfinal for “embedded” decisions. 175 It then decided that GM’s motion was embedded in the broader commercial dispute and dismissed the appeal. 176

In order to explain my dissent from that holding, I need only quote from the opening paragraph that I wrote:

Procedural rules should not be matters of Byzantine complexity, known and accessible only to initiates who are privy to knowledge that is kept secret from the rest of the world. But that is the kind of rule I fear the majority has created in its decision to make our appellate jurisdiction turn on the distinction between “independent” and “embedded” proceedings, even when a party is seeking to appeal under 28 U.S.C. § 1291 from a final judgment dismissing an action. Because I believe that there is a crucial difference between an order staying an arbitration proceeding, and an order directing the entry of a final judgment dismissing an action so that arbitration may proceed, I respectfully dissent from the majority’s conclusion that we have no jurisdiction over Katherine Napleton’s appeal. 177

Two years later, in Green Tree Financial Corp.-Alabama v. Randolph,171 the Supreme Court took up this question and expressly disapproved the distinction between “independent” and “embedded” proceedings that a majority of the courts of appeals had been following. Green Tree confirmed a system that is predictable for all concerned: if the arbitral proceedings are stayed, then section 16 of the Federal Arbitration Act’ 79 governs appealability of orders granting or denying motions to compel arbitration, but if the action is resolved through a final judgment, the normal rules found in 28 U.S.C. § 1291 should apply.

In Perlman v. Swiss Bank Corp. Comprehensive Disability Protection Plan, I dissented because I believed that the majority had drawn too close an analogy between procedures applicable to Social Security Act appeals and the procedures that are appropriate for review of decisions of ERISA plan administrators to deny benefits, when the “arbitrary and capricious” standard applies. 180 Long-standing principles of law apply to ensure the procedural and substantive regularity of the Social Security decisions, but plan administrators enjoy a great deal of flexibility. The majority in Perlman ruled that the district court should not have examined the administrator’s decisionmaking process, and it should have confined itself to the record that had already been compiled.’ 8′ With review so narrowed, it concluded that the denial of benefits should not be disturbed. 182 Drawing the analogy to the law of trusts that the Supreme Court had endorsed in Firestone Tire & Rubber Co. v. Bruch,’ 83 I argued both that we could review the procedures that the plan had used and that it was within our authority to conclude that it had proceeded on the basis of an inadequate record. I would have sent the case back for additional fact-finding and reconsideration.

My last example of these process-driven dissents deals with the rules of procedural default that apply for collateral attacks on criminal convictions. Specifically, in United States v. Smith184 our court had to decide whether a criminal defendant had defaulted an argument flowing from the decision in Apprendi v. New Jersey,185 which held that facts that increase the penalty for a crime beyond a prescribed statutory maximum must be submitted to the jury and proved beyond a reasonable doubt, even in state criminal proceedings. Almost eight years after United States v. Booker,186 it is easy to forget how much of a shock Apprendi was to the legal system. The defendant in our case had failed to argue in his basic criminal proceeding that his sentence was impermissible for this reason, and the majority concluded that he had therefore procedurally defaulted the point.1 87 1 was disturbed by a rule that placed such a high burden on defense counsel. As I wrote in dissent from the denial of rehearing en banc, “[a]ccording to the panel, even if there is no way that a defendant, or more to the point, defense counsel, could have anticipated the later legal ruling, the defendant cannot raise the point in a later collateral attack. 188. I saw nothing in the Supreme Court’s decisions that compelled such a harsh rule. I concluded with the following words:

I fear that the panel’s rule will create an administrative nightmare not only for defense counsel trying to represent their clients responsibly, but also for the district courts and this court. After this, defense counsel will have no choice but to file one “kitchen sink” brief after another, raising even the most fanciful defenses that could be imagined based on long-term logical implications from existing precedents. The Supreme Court may never go down most of those paths, but that will not matter, because otherwise the defendant will find him or herself staring at a procedural default that cannot be overcome for good cause. I have the deepest concern about the consequences of this approach for the courts. 189

Somehow there must be a balance between the duty to raise an argument, even if it is novel, and the ability to rely on law that has become settled around the country, even if some residual chance exists that the Supreme Court may step in and upset those expectations. I believed that the court had drawn that line in the wrong place.

C. Accuracy-Focused Dissents

Finally, I have written dissents that strive only to set the factual record straight. Law obviously does not exist in a vacuum. It is entirely possible-indeed common-to agree entirely with the abstract propositions of law that the majority has set forth and yet to think that it has failed to apply them properly to the case before the court. It bears emphasis that these cases are usually not cert-worthy; the Supreme Court does not sit as a court of errors. This means that the courts of appeals have a special responsibility to ensure that factual errors are addressed, because they are the end of the line for that important aspect of the litigation process. Moreover, many of the values of separate opinions are served just as well for factual disagreements as they are for legal disagreements: confidence that each judge is working hard to do the job, assurance of transparency in the process, and honesty, to name a few.

I have chosen a single case to illustrate this point: Hammer v. Ashcrof. 190 The case concerned a policy that the Federal Bureau of Prisons (BOP) adopted about press access to inmates in the Special Confinement Unit (SCU) of the prison at Terre Haute, Indiana-the unit that houses federal prisoners under a sentence of death. 19′ At one time, reporters were able freely to interview inmates housed in the SCU. 192 But in March 2000, CBS News aired an interview with Timothy McVeigh, the Oklahoma City bomber, on its 60 Minutes show. 19 3 McVeigh used the interview as an opportunity to justify his actions and to extol terrorism. 194 Shortly after the broadcast, Senator Byron Dorgan wrote to the Director of the BOP complaining about it; Senator Dorgan felt that a prisoner like McVeigh should not be able to use a forum like 60 Minutes to advance his agenda of violence.195 Attorney General John Ashcroft and the Director responded with a new regulation for the SCU. 1 96 According to the majority opinion, the new rule “bans person-to-person meetings between reporters and inmates of the Special Confinement Unit, though it allows phone calls and correspondence.”‘ 97 Phone calls with reporters are subject to another limitation under which a reporter may not “obtain and use personal information from one inmate about another inmate who refuses to be interviewed.”‘ 98 Correspondence with the press, according to the majority, is unlimited and not subject to inspection or censorship. 199 On that understanding of the legal framework, the majority held that inmate Hammer’s challenge to the rules on First Amendment grounds had to fail.2 The Supreme Court’s decisions in Pell v. Procunier01 and Saxbe v. Washington Post Co., which held respectively that news reporters have no greater constitutional right of access to prison inmates than the general public and that the BOP was entitled to ban face-to-face interviews, doomed Hammer’s case. The majority concluded by noting that serious abuses within the prisons could still be revealed either through correspondence or the filing of lawsuits.20 3

I was among the dissenters to this opinion, largely because of my concerns about the record. As I said, “[c]entral to the majority’s analysis is the assumption that inmate correspondence is unlimited; an inmate’s letters to reporters are not subject to inspection or censorship.” 20 4 But a closer look at the regulations revealed that this was not necessarily so. Nothing in the Program Statement that forbids media representatives from obtaining and using personal information from one inmate about another limits that restriction to information gathered through telephone calls. 2 ° 5 Materials in the record actually supported the opposite conclusion-namely, that the restriction applies across-the-board. For example, a letter from the Warden rebuked a lawyer for publishing a news release that included statements that Hammer had made about other inmates.206

Hammer himself had also been reprimanded for providing information about others. In response to a request from Hammer for clarification about the policy, the Warden wrote, “You are hereby ordered not to provide any information concerning other inmates during news media interviews, social calls, or correspondence with the media.” 20 7 At oral argument, counsel for the government contradicted another of the majority’s assumptions: counsel stated that all mail sent by inmates in the SCU must be given unsealed to prison officials before it is sent out, so that it can be inspected. 208 These facts indicated to me that we were faced with a different, and far more difficult, case than the one portrayed in the majority’s opinion. With the safety valves the majority assumed were in place, perhaps the policies should have passed constitutional muster. Without them, the risks that serious abuses within the prisons might go undetected seemed to me a real possibility, and one that we needed to confront head-on.

CONCLUDING OBSERVATIONS

Let us return at this point to the fundamental question: What should a judge do when she disagrees with her colleagues on the bench? Should the potential dissenter always “fold”? Should the potential dissenter adamantly refuse to meet others halfway?

There is no singular answer to these questions. Rather, the answer depends on criteria like a judge’s aversion to the extra work that a separate opinion entails, her desire to get along with her colleagues, or the purity of her intentions.209 Chief Justice Roberts is not wrong, however, to worry about the effect on the legal system of an excessive amount of separate writing, especially when the court of last resort is unable to muster a single coherent, binding ruling. One wonders if the large number of separate opinions, especially at the Supreme Court, has come about in part because the courts, like so many other institutions in our society, have lost the knack of compromise. Might there be lessons from history to be learned, when one examines how the Court finally surmounted earlier deep splits, such as those in the fields of obscenity and the Establishment Clause?

Perhaps there should be a norm under which every judge or justice should stop and consider under what grounds a contemplated separate opinion can be justified. Just as the courts of appeals require advocates to include a section in their brief addressing the standard of review,210 maybe concurring and dissenting opinions ought to touch briefly on both the normative and instrumental purposes the writer hopes to achieve-an address to a higher court, a plea to the legislature or an agency, a call for a constitutional amendment or the overruling of a constitutional case, or even just a simple plea to get the facts and process right. These are all worthy goals. If separate writing were suppressed-unlikely though that is at this late date in our history-it seems likely that we would lose more than we would gain. This is particularly true for separate writing at the court of appeals level; it is entirely possible that there is too little dissenting and concurring by the circuit judges, while there may be a too much of a good thing in the courts of last resort.

If the balance is indeed out of whack, then we need to ask why. The culprit that comes most immediately to mind is the size and nature of the docket at each of the levels we have been considering. Although the number of petitions for certiorari at the Supreme Court hovers around 8000 per year, 211 the Justices hand down decisions on the merits in only about 1 percent of those cases. Spreading eighty-some cases among nine Justices yields a manageable workload, even factoring in the decision to write separately from time to time. 212 Each Justice would need to file a separate opinion two-thirds of the time to reach the mean number of majority opinions that each circuit judge writes. 213

It is worth noting that the courts of appeals for many years now have been slipping quietly toward a system that could be described as certiorari-lite, under which only the most challenging cases receive plenary treatment through oral argument. Circuits groaning under the weight of heavy caseloads and unfilled judgeships ration oral argument time tightly, and their publication policies are even more parsimonious. Even in the Seventh Circuit, which has the highest percentage of oral argument in the country, only 46.7 percent of the appeals were terminated after oral hearing. 14 At the low end, in the Fourth Circuit only 13.1 percent of appeals were concluded after oral argument. This may be undermining the nominally mandatory nature of the intermediate level’s jurisdiction. True, litigants still get an answer from the circuit courts, but deferential standards of review and abbreviated procedures have taken a toll on that part of the appellate process.

As a systemic matter, this is cause for concern, if one believes in the theory that the circuit judges are sending signals to the Supreme Court through the cases in which they write separately. It is just as difficult for the Justices to weed through 8000-odd filings in depth as it is for the courts of appeals to cope with large caseloads. Conflicts in the circuits, as Supreme Court Rule 10 recognizes, are one proxy for the kind of case that is worth the Court’s attention, but so too are well-reasoned separate opinions within one court. In the final analysis, therefore, we can return to the “Gambler,” or, if you prefer, to Justice Blackmun. While there are times when acquiescing to one’s colleagues is the correct thing to do, there are just as surely situations in which expressing a different view is called for. The trick is to know which to do at what times. Awareness of what we are actually doing, and to what end, will help judges use their prerogative to write separately in a more intelligent way. Even more importantly, it will help all who are affected by the decisions judges make to understand the dynamics of judicial decision-making and how our quasi-common-law system continues to evolve.

FOOTNOTES

1. HISTORY OF THE WORLD: PART I (Brooksfilms 1981).
2. See JAMES C. DUFF, ADMIN. OFFICE OF THE U.S. COURTS, JUDICIAL BUSINESS ANNUAL REPORT FOR 2010 (2010) [hereinafter AO 2010 REPORT]. The AO 2010 Report indicates that the district courts terminated almost 390,000 cases (combining civil and criminal, but excluding bankruptcy); the courts of appeals in the aggregate terminated a paltry 60,000-roughly 15 percent. Id at 14.
3. For a broader look at reversal rates, see Stephen J. Choi et al., What Do Federal District Judges Want? An Analysis of Publications, Citations, and Reversals (John M. Olin Law & Econs., Working Paper No. 508, 2010), available at http://www.law.uchicago.edu/files/file/508-eap-distjudge.pdf. The analysis reports an overall affirmance rate (including both appealed cases and unappealed cases) of 91 percent for all twelve circuits. Looking only at the dispositions of appealed cases, the AO 2010 Report, Table B-5, indicates that nationwide, 8.3 percent of cases that were terminated on the merits were reversed. Reversal rates range from 15.3 percent in the District of Columbia Circuit to 4.9 percent in the Eighth Circuit. AO 2010 REPORT, supra note 2, at 111 tbl.B-5.
4. See 28 U.S.C. § 46(b), (c) (2006).
5. Id § 46(c).

6. Quoted in Diane P. Wood, Justice Harry A. Blackmun and the Responsibility of Judging, 26 HASTINGS CONST. L.Q. 11, 18 (1998) (original is on videotape with the author).
7. “You got to know when to hold ’em, know when to fold ’em / Know when to walk away and know when to run.” KENNY ROGERS, The Gambler, on THE GAMBLER (United Artists 1978).
8. U.S. CONST. art. Ill, § 1.
9. For a representative sample, see Frank X. Altimari, Foreword The Practice of Dissenting in the Second Circuit, 59 BROOK. L. REv. 275 (1993); William J. Brennan, Jr., In Defense of Dissents, 37 HASTINGS L.J. 427 (1986); Ruth Bader Ginsburg, Remarks on Writing Separately, 65 WASH. L. REv. 133 (1990); M. Todd Henderson, From Seriatim to Consensus and Back Again: A Theory of Dissent, 2007 SUP. CT. REv. 283; Maurice Kelman, The Forked Path of Dissent, 1985 SUP. CT. REv. 227; Laura Krugman Ray, Justice Brennan and the Jurisprudence of Dissent, 61 TEMP. L. REv. 307 (1988); Rory K. Little, Reading Justice Brennan: Is There a “Right’to Dissent?, 50 HASTINGS L.J. 683 (1999); Antonin Scalia, The Dissenting Opinion, 1994 J. SUPREME CT. HIST. 33; Kevin M. Stack, The Practice of Dissent in the Supreme Court, 105 YALE L.J. 2235 (1996). A more recent branch of scholarship has taken an explicitly empirical approach to the issue. See, e.g., Andrew F. Daughety & Jennifer F. Reinganum, Speaking Up: A Model of Judicial Dissent and Discretionary Review, 14 SUP. CT. ECON. REv. 1 (2006); Lee Epstein et al., Why (and When) Judges Dissent: A Theoretical and EmpiricalApproach, 3 J. LEGAL ANALYSIS 101 (2011).
10. AO 2010 REPORT, supra note 2, at 83 tbl.B app.
11. Kedar Bhatia, Stat Pack for October Term 2010, SCOTUSBLOG 5 (June 28, 2011), http://sblog.s3.amazonaws.com/wp-content/uploads/2011/06/SBOTlO_stat packfmal.pdf.
12. Several other judges have noted how commonplace separate opinions have become. E.g., Michael Boudin, Friendly, J., Dissenting, 61 DUKE L.J. 881, 881 (2012) (‘American legal culture takes the dissent for granted …. “); Alex Kozinski & James Burnham, I Say Dissental, You Say Concurral, 121 YALE L.J. ONLINE 601, 602 (2012) (describing the “flourish[ing]” practice of
dissenting from orders denying rehearing en banc), http://yalelawjoumal.org/images/pdfs/1071.pdf.
13. See, e.g., Arthur J. Jacobson, Publishing Dissent, 62 WASH. & LEE L. REv. 1607, 1609
(2005). Some countries, like Germany, established constitutional courts after World War II. In someinstances those courts have decided to permit separate opinions, even if the ordinary courts continue to
avoid them. It would be interesting to study this phenomenon in greater depth and to see whether the
rationale coincides with the one often used to justify separate writings for constitutional decisions in
the U.S. Supreme Court.
14. See, e.g., Mark Tushnet, Incentives and the Supreme Court, 78 GEO. WASH. L. REV. 1300, 1304 (2010) (noting that the European Court of Justice “operates under the norm of consensus”).
15. Sanford Levinson, Speaking in the Name of the Law: Some Reflections on Professional Responsibility and Judicial Accountability, 1 U. ST. THOMAS L.J. 447, 457 (2003); see also Andrew Lynch, Dissent: The Rewards and Risks of Judicial Disagreement in the High Court of Australia, 27 MELB. U. L. REv. 724, 737 n.52 (2003).
16. The European Court of Human Rights, in contrast, has traditionally permitted both concurring and dissenting opinions. See Robin C.A. White & Iris Boussiakou, Separate Opinions in the European Court of Human Rights, 9 HUM. RTS. L. REV. 37 (2009).
17. See Barbara Black & Jill I. Gross, Making It Up as They Go Along: The Role of Law in Securities Arbitration, 23 CARDOzO L. REV. 991, 1000-01 (2002) (describing arbitrators’ reluctance to issue written opinions or indicate dissents and concurrences).
18. See Boudin, supra note 12, at 881.
19. See First Female Law Lord Appointed, BBC NEWS (Oct. 24, 2003, 6:51 AM),
http://news.bbc.co.uk/2/hi/uknews/3210003.stm.
20. See SUPREME COURT, http://www.supremecourt.gov.ukl (last visited July 28, 2012) (the Court’s website).
21. See Decided Cases, SUPREME COURT, http://www.supremecourt.gov.uk/decidedcases/index.html (last visited July 28, 2012).
22. There is one interesting exception to this practice. The Judicial Committee of the Privy Council, which still serves as the court of final appeal for the UK overseas territories and Crown dependencies, as well as for some Commonwealth countries that have not substituted their own institutions for it, issues only a single judgment, perhaps on the theory that it is advising the Crown and thus should speak with a single voice. See JUDICIAL COMM. PRIVY COUNCIL, http://www.jcpc.gov.uk/ (last visited July 28, 2012) (the official website of the Committee); see also Claire L’Heureux-Dub6, The Dissenting Opinion: Voice of the Future?, 38 OSGOODE HALL L.J. 495, 499 n.9 (2000) (citing Bora Laskin, The Supreme Court of Canada: A Final Court of and for Canadians, 29 CAN. B. REV.1038, 1073 (1951)).
23. See Stack, supra note 9, at 2238-39.
24. Henderson, supra note 9, at 283; see also Stack, supra note 9, at 2238-39; Scalia, supra note 9, at 34-35 (noting Thomas Jefferson’s disdain for the Marshall Court’s tendency to issue unanimous opinions).
25. A canvas of the state courts is beyond the scope of this Essay. For examples, however, see Robert G. Flanders, Jr., The Utility of Separate Judicial Opinions in Appellate Courts of Last Resort: Why Dissents Are Valuable, 4 ROGER WILLIAMS UNIV. L. REV. 401 (1999) (examining the practice of the Supreme Court of Rhode Island); R. Perry Sentell, Jr., Dissenting Opinions: In the Georgia Supreme Court, 36 GA. L. REV. 539 (2002). For a discussion of the practice in Canada, see L’Heureux-Dub6, supra note 22. In Australia, the practice appears to have remained closer to the British tradition. See Lynch, supra note 15.
26. Brennan, supra note 9, at 427.
27. Henderson, supra note 9, at 283 & nn.1-2.
28. Jeffrey Rosen, Roberts’s Rules, ATLANTIC, Jan.-Feb. 2007, at 105.

29. Bhatia, supra note 11.
30. Id
31. ld.
32. Epstein et al., supra note 9, at 106-07 (aggregate data from 1990 to 2007; published opinion data for the years 1989 to 1991).
33. Patricia M. Wald, The Rhetoric of Results and the Results of Rhetoric: Judicial Writings,
62 U. Ci. L. REV. 1371, 1412 (1995).

34. Id. at 1413.
35. Id. at 1413-14.

36. See Sentell, supra note 26, at 545 (quoting Felix Frankfurter, Mr. Justice Holmes and the Constitution: A Review of His Twenty-Five Years on the Supreme Court, 41 HARV. L. REv. 121, 162 (1927)).

37. See Kelo v. City of New London, 545 U.S. 469 (2005). The majority’s opinion in Kelo upheld a city’s decision to use its power of eminent domain for an urban redevelopment project. Justice Kennedy concurred, emphasizing both the standard of review and the facts of the case. Justice O’Connor, joined by three colleagues, dissented.
38. Id. at 493 (Kennedy, J., concurring).
39. See Wald, supra note 33.
40. See Epstein et al., supra note 9.
41. 489 U.S. 189 (1989); see id. at 203 (Brennan, J., dissenting); id. at 212 (Blackmun, J., dissenting).
42. 538 U.S. 343, 388 (2003) (Thomas, J., dissenting).

43. 537 U.S. 186, 242 (2003) (Breyer, J., dissenting).
44. 198 U.S. 45, 75 (1905) (Holmes, J., dissenting).
45. Id.
46. 517 U.S. 44, 76 (1996) (Stevens, J., dissenting).
47. Id. at 77.
48. Brennan, supra note 9, at 436.
49. 550 U.S. 124, 191 (2007) (Ginsburg, J., dissenting) (“[T]oday’s opinion …is hardly faithful to our earlier invocations of ‘the rule of law’ and the ‘principles of stare decisis.”‘).
50. 534 U.S. 362, 388 (2002) (Kennedy, J., dissenting).
51. Id.at387.
52. Id. at 395 (Kennedy, J., dissenting).
53. 472 U.S. 38,90 (1985) (White, J., dissenting).

54. Id. at 90-91.
55. See, e.g., Lehman v. City of Shaker Heights, 418 U.S. 298, 308 (1974) (Brennan, J., dissenting); Walker v. City of Birmingham, 388 U.S. 307, 338 (1967) (Brennan, J., dissenting).
56. See, e.g., Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 979 (1992) (Scalia, J., dissenting in part).
57. 510 U.S. 1141, 1143 (1994) (Blackmun, J., dissenting).
58. Id. at 1145.
59. Id.
60. I put to one side the hoped-for gains from a separate opinion that has been drafted and circulated in private to the panel, in the hopes of persuading the other members to change their views. I am not doing so because this is unimportant: to the contrary, it can be crucial. But it is invisible enough that there is little that I can say about it other than to commend to you the study of whatever judicial papers are available. Those papers, as well as other direct comments from judges, demonstrate that this kind of flip can and does happen, even if rarely. That chance may be enough to justify the drafting effort even for the judge on a court of last resort.

61. See 653 F.3d 448 (7th Cir. 2011), revg 630 F.3d 570 (7th Cir. 2010) (withdrawn from bound volume on grant of rehearing).

62. Id. at 471.
63. Compare 3D CIR. L.O.P. R. 5.5.4 (no prepublication circulation of unanimous opinions),
with 7TH CR. R. APP. P. 40(e) (allowing prepublication circulation of unanimous opinions).
64. See 7TH CiR. R. APP. P. 40(e).
65. See generally Kozinski, supra note 12.

66. Seminole Tribe v. Florida, 517 U.S. 44, 69 (1996) (quoting Monaco v. Mississippi, 292 U.S. 313,325 (1934)) (noting Chisholm was received with “shock of surprise”); RICHARD H. FALLON, JR. ET AL., HART AND WECHSLER’S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 871 (6th ed. 2009) (noting the decision “provoked a strongly adverse reaction”). But see Seminole Tribe, 517 U.S. at 106 n.5 (Souter, J., dissenting) (stating that the suggestion Chisholm was received with “shock of surprise” is an “erroneous assertion”); John J. Gibbons, The Eleventh Amendment and State Sovereign Immunity: A Reinterpretation, 83 COLUM. L. REV. 1889, 1927 (1983) (“[C]ontrary to the position of the profound shock school, the Second Congress did not regard Chisholm v. Georgia as a matter of great moment.”).Seminole Tribe v. Florida, 517 U.S. 44, 69 (1996) (quoting Monaco v. Mississippi, 292 U.S. 313,325 (1934)) (noting Chisholm was received with “shock of surprise”); RICHARD H. FALLON, JR. ET AL., HART AND WECHSLER’S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 871 (6th ed. 2009) (noting the decision “provoked a strongly adverse reaction”). But see Seminole Tribe, 517 U.S. at 106 n.5 (Souter, J., dissenting) (stating that the suggestion Chisholm was received with “shock of surprise” is an “erroneous assertion”); John J. Gibbons, The Eleventh Amendment and State Sovereign Immunity: A Reinterpretation, 83 COLUM. L. REV. 1889, 1927 (1983) (“[C]ontrary to the position of the profound shock school, the Second Congress did not regard Chisholm v. Georgia as a matter of great moment.”).

67. 2 U.S. (2 Dall.) 419 (1793).
68. See Stack, supra note 9, at 2238.
69. U.S. CONST. amend. XI; U.S. CONST. art. 111, § 2.
70. See the line of cases beginning with Seminole Tribe, 517 U.S. 44 (1996). E.g., Fla. Prepaid Postsecondary Educ. Expense Bd. v. College Sav. Bank, 527 U.S. 627 (1999), College Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666 (1999), Alden v. Maine, 527 U.S. 706 (1999).
71. 158 U.S. 601 (1895).

72. U.S. CONST. amend. XVI.
73. U.S. CONST. amend. XXIV.
74. 302 U.S. 277 (1937).
75. 383 U.S. 663, 665-67 (1966).
76. U.S. CONST. amend. XXVI.
77. 400 U.S. 112 (1970) (upholding federal law enfranchising eighteen-year-olds for federal elections, but striking it down over Justice Douglas’s dissent insofar as it addressed state and local elections).
78. See generally Michael E. Solinine & James L. Walker, The Next Word: Congressional Response to Supreme Court Statutory Decisions, 65 TEMP. L. REV. 425 (1992).
79. See Goldman v. Weinberger, 475 U.S. 503 (1986).
80. See National Defense Authorization Act of Fiscal Years 1988 and 1989, Pub. L. No. 100-
180, 101 Stat. 1019.
81. 492 U.S. 158, 175 (1989).

82. Pub. L. No. 101-433, 104 Stat. 978 (1990).
83. See Ledbetter v. Goodyear Tire & Rubber Co., 550 U.S. 618, 632 (2007).
84. Pub. L. No. 111-2, 123 Stat. 5.
85. Pub. L. No. 102-166, 105 Stat. 1071.
86. See Martin v. Wilks, 490 U.S. 755 (1989) (5-4 decision); Patterson v. McLean Credit Union, 491 U.S. 164 (1989) (5-4 decision); Price Waterhouse v. Hopkins, 490 U.S. 228 (1989) (no majority opinion; Congress adopts view of Justice Brennan, writing for the plurality, in mixed motive discrimination cases); Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989) (5-4 decision).
87. 163 U.S. 537, 559 (1896).
88. 347 U.S. 483,495 (1954).
89. 277 U.S. 438 (1928).
90. 389 U.S. 347 (1967).
91. Id at 353. Compare the several opinions in United States v. Jones, 132 S. Ct. 945 (2012), debating the relevance of literal trespass for Global Positioning System devices.
92. 517 U.S. 44(1996).
93. 491 U.S. 1 (1989).

94. 482 U.S. 386, 397 n.ll (1987).
95. 452 U.S. 394 (1981).
96. 310 U.S. 586(1940).
97. 319 U.S. 624 (1943).
98. Judge Gerard E. Lynch of the Second Circuit correctly points out that sometimes respect for a colleague’s strongly held view may also bring about acquiescence. Rather than a cost, this may be seen as an investment in collegiality. Gerard E. Lynch, Comment at the Columbia Law School Courts and Legal Process Workshop (April 23, 2012).
99. See Cabana v. Bullock, 474 U.S. 376, 393 (1986) (Brennan, J., dissenting); Baldwin v. Alabama, 472 U.S. 372, 392 (1985) (Brennan, J., dissenting); Antone v. Dugger, 465 U.S. 200, 207 (1984) (Brennan, J., dissenting); Pulley v. Harris, 465 U.S. 37, 59 (1984) (Brennan, J., dissenting); Sullivan v. Wainwright, 464 U.S. 109, 112 (1983) (Brennan, J., dissenting); Maggio v. Williams, 464 U.S. 46, 56 (1983) (Brennan, J., dissenting); Autry v. Estelle, 464 U.S. 1, 3 (1983) (Brennan, J., dissenting); Hopper v. Evans, 456 U.S. 605, 614 (1982) (Brennan, J., dissenting in part); Green v. Georgia, 442 U.S. 95, 97 (1979) (Brennan, J., dissenting); Dobbert v. Florida, 432 U.S. 282, 304 (1977) (Brennan, J., dissenting); Gardner v. Florida, 430 U.S. 349, 364 (1977) (Brennan, J.,
dissenting).
100. See Kuhlmann v. Wilson, 477 U.S. 436, 461 (1986) (Brennan, J., dissenting); Engle v. Isaac, 456 U.S. 107, 149-51 (1982) (Brennan, J., dissenting); Rose v. Lundy, 455 U.S. 509, 532-33 (1982) (Brennan, J., dissenting in part); Sumner v. Mata, 449 U.S. 539, 558 (1981) (Brennan, J., dissenting); Wainwright v. Sykes, 433 U.S. 72, 99 (1977) (Brennan, J., dissenting); Francis v. Henderson, 425 U.S. 536, 542 (1976) (Brennan, J., dissenting); Stone v. Powell, 428 U.S. 465, 502 (1976) (Brennan, J., dissenting).
101. See Cnty. of Oneida v. Oneida Indian Nation, 470 U.S. 226, 254 (1985) (Brennan, J.,
dissenting in part); Pennhurst State Sch. & Hosp. v. Halderman, 465 U.S. 89, 125 (1984) (Brennan, J.,
dissenting); Fla. Dep’t of State v. Treasure Salvors, Inc., 458 U.S. 670, 700 (1982) (Brennan, J.,
dissenting in part); Fla. Dep’t of Health & Rehab. Servs. v. Fla. Nursing Home Ass’n, 450 U.S. 147,
155 (1981) (Brennan, J., dissenting); Edelman v. Jordan, 415 U.S. 651, 687 (1974) (Brennan, J., dissenting).

102. E.g., In re Anastaplo, 366 U.S. 82, 116 (1961) (Brennan, J., dissenting); Konigsberg v. State Bar, 366 U.S. 36, 80 (1961) (Brennan, J., dissenting); Kimm v. Rosenberg, 363 U.S. 405, 411 (1960) (Brennan, J., dissenting); Nelson v. Cnty. of L.A., 362 U.S. 1, 10 (1960) (Brennan, J., dissenting); Uphaus v. Wyman, 360 U.S. 72, 82 (1959) (Brennan, J., dissenting); Beilan v. Bd. of Pub.
Educ., 357 U.S. 399, 417 (1958) (Brennan, J., dissenting); Lerner v. Casey, 357 U.S. 458, 417 (1958) (Brennan, J., dissenting) (majority opinion reported at 357 U.S. 468 because Justice Brennan’s dissent from Beilan applies to Lerner as well).
103. E.g., Zauderer v. Office of Disciplinary Counsel, 471 U.S. 626, 656 (1985) (Brennan, J., concurring in part, concurring in the judgment in part, dissenting in part); Regan v. Time, Inc., 468 U.S. 641, 659 (1984) (Brennan, J., dissenting in part); Clements v. Fashing, 457 U.S. 957, 976 (1982) (Brennan, J., dissenting); Broadrick v. Oklahoma, 413 U.S. 601, 621-22 (1973) (Brennan, J., dissenting).
104. In the area of expectation of privacy, see, for example, Illinois v. Andreas, 463 U.S. 765,
773 (1983) (Brennan, J., dissenting); Andreson v. Maryland, 427 U.S. 463, 484 (1976) (Brennan, J.,
dissenting); United States v. Miller, 425 U.S. 435, 447 (1976) (Brennan, J., dissenting); Cady v.
Dombrowski, 413 U.S. 433, 450 (1973) (Brennan, J., dissenting); Lopez v. United States, 373 U.S.
427, 446 (1963) (Brennan, J., dissenting). For Terry stops in particular, see, for example, New York v.
Class, 475 U.S. 106, 122 (1986) (Brennan, J., dissenting in part); United States v. de Hernandez, 473
U.S. 531, 545 (1985) (Brennan, J., dissenting); United States v. Sharpe, 470 U.S. 675, 702 (1985)
(Brennan, J., dissenting); Michigan v. Long, 463 U.S. 1032, 1054 (1983) (Brennan, J., dissenting);
United States v. Villamonte-Marquez, 462 U.S. 579, 593 (1983) (Brennan, J., dissenting); Michigan v.
DeFillippo, 443 U.S. 31, 41 (1979) (Brennan, J., dissenting); Adams v. Williams, 407 U.S. 143, 151
(1972) (Brennan, J., dissenting).
105. 531 U.S. 98 (2000).
106. E.g., Linda Greenhouse et al., My Florida Recount Memory, N.Y. TIMES, Nov. 21, 2010,at WK 10.
107. Bush v. Gore, 531 U.S. at 129 (Stevens, J., dissenting); see also id at 144 (Breyer, J.,dissenting) (arguing the Court should not have heard the case).
108. E.g., Linda Greenhouse et al., supra note 106.

109. See Dennis J. Hutchinson, Unanimity and Desegregation: Decisionmaking in the Supreme Court, 1948-1958, 68 GEO. L.J. 1, 55-56, 87 (1979).
110. E.g., Jacobellis v. Ohio, 378 U.S. 184 (1964); A Book Named “John Cleland’s Memoirs of a Woman of Pleasure” v. Massachusetts, 383 U.S. 413 (1966).
111. In Salazar v. Buono, 130 S. Ct. 1803 (2010), which considered the question whether the display of a Latin cross in a national preserve violated the Establishment Clause, six of the nine Justices wrote opinions. Six Justices also wrote in Wallace v. Jaffree, 472 U.S. 38 (1985), which considered an Alabama statute authorizing a one-minute period of silence. And Wolman v. Walter,433 U.S. 229 (1977), which involved the constitutionality of a number of provisions of Ohio law designed to provide aid to students of nonpublic schools, produced separate statements from every member of the Court except Justice Stewart.
112. See J. McIntyre Mach., Ltd. v. Nicastro, 131 S. Ct. 2780 (2011) (no majority opinion; three separate opinions); Asahi Metal Indus. Co. v. Superior Court, 480 U.S. 102 (1987) (same).

113. Westlaw’s “profiler” associates me with 4654 cases. Our own count shows 4555 cases. One has to take care with the numbers for the first nine years I was on the bench, because at the time I was fortunate enough to be sitting with Judge Harlington Wood, Jr., then a senior judge on the Seventh Circuit. In order to avoid confusion with one another, we both attempted to use our full names on panels. But there may have been some slippage at the outset, and thus some slight discrepancies in the numbers. Overall, however, these differences do not change the broader points I am making.
114. The unreported case was United States v. Webb, 74 F.3d 1242 (7th Cir. 1996).

115. 372 U.S. 335 (1963).

116. 79 F.3d 559, 560 (7th Cir.), vacated en bane, 89 F.3d 838 (7th Cir. 1996).
117. Id.
118. Idat561.
119. Id.
120. Id. at565.
121. 439 U.S. 259 (1978).
122. 456 U.S. 91 (1982).
123. 462 U.S. 1 (1983).
124. 486 U.S. 456 (1988).

125. Order Granting Rehearing En Banc and Vacating Decision, 79 F.3d 559 (7th Cir. 1996) (order of May 15, 1996).
126. Haas v. Chater, 89 F.3d 838 (7th Cir. 1996) (en banc).
127. Haas v. Chater, 519 U.S. 1108 (1997).
128. 102 F.3d 872 (7th Cir.), denying reh ‘g of merits opinion at 95 F.3d 481(7th Cir. 1996).
129. Holman, 95 F.3d at 491.
130. 428 U.S. 465 (1976).
131. Holman, 95 F.3d at 489.
132. 477 U.S. 365 (1986).
133. 520 U.S. 1254 (1997).
134. 387 F.3d 607 (7th Cir. 2004).
135. Id. at 611.
136. 322 F.3d 942 (7th Cir. 2003) (en banc).

137. Id. at 944.
138. Id. at 953.
139. Id. at 954 (Wood, J., dissenting).
140. 523 U.S. 83 (1998).
141. United Phosphorus, Ltd., 322 F.3d at 954 (Wood, J., dissenting).
142. Animal Science Prods., Inc. v. China Minmetals Corp., 654 F.3d 462, 469 n.8 (3d Cir. 2011).
143. 657 F.3d 650,657 (7th Cir. 2011).
144. Minn-Chem, Inc. v. Agrium, Inc., No. 10-1712, 2012 WL 2403531 (7th Cir. June 27, 2012) (en banc).
145. 484 F.3d 436 (7th Cir. 2007) (en banc), aff’d, 553 U.S. 181 (2008).
146. Crawford v. Marion Cnty. Election Bd., 472 F.3d 949 (7th Cir. 2007).
147. Crawford, 484 F.3d at 436.

148. Id. at 437 (Wood, J., dissenting).

149. Crawford,533 U.S. at 202.

150. 533 F.3d 562 (7th Cir. 2008), rev’d in part, 587 F.3d 771 (7th Cir. 2009) (en banc).

151. Id. at563.
152. Id.

153. Id. at 567.

154. Id. at 563-64.

155. 494 U.S. 872 (1990).

156. 508 U.S. 520(1993).

157. Bloch, 533 F.3d at 564.

158. Id. at 565.

159. Id at 566 (Wood, J., dissenting).

160. Bloch v. Frischholz, 587 F.3d 771 (7th Cir. 2009) (en banc).

161. 138 F.3d 635 (7th Cir. 1997).

162. Id at 638.

163. Id

164. Id; see also Consol. Rail Corp. v. Ry. Labor Execs.’ Ass’n, 491 U.S. 299, 302-03 (1989).

165. Bhd of Maint. of Way Emps., 138 F.3d at 638.

166. Id

167. Id at 640.

168. Id. at642-43.

169. 138 F.3d 1209(7th Cir. 1998).

170. Id.at 1210.
171. Id.at 1210-11.

172. Id. at 1211.
173. Id.
174. Id.

175. Napleton, 138 F.3d at 1214.

176. Id.

177. Id at 1214-15.

178. 531 U.S. 79, 80 (2000).

179. See 9 U.S.C. § 16 (2006).

180. 195 F.3d 975, 983-86 (7th Cir. 1999) (Wood, J., dissenting).

181. Id.at981.
182. Id at 982-83.

183. 489 U.S. 101 (1989).

184. See 241 F.3d 546 (7th Cir. 2001) (panel opinion), on remandfrom the Supreme Courtfor reconsideration in light of Apprendi v. New Jersey, 530 U.S. 466 (2000), 531 U.S. 941 (2000), vacating 215 F.3d 1331 (7th Cir. 2000).

185. Apprendi,530 U.S.466.

186. 543 U.S. 220 (2005).
187. Smith, 241 F.3d at 548.

188. United States v. Smith, 250 F.3d 1073, 1074 (7th Cir. 2001) (order denying en banc rehearing, with dissent by Judge Wood).

189. Id. at 1077.

190. 570 F.3d 798 (7th Cir. 2009) (en banc).

191. Id at 799.

192. Id
193. Id at 802.

194. Id
195. Id.
196. Hammer, 570 F.3d at 802.

197. Id at 799-800.

198. Id. at 800 (citing U.S. DEP’T OF JUST., FED. BUREAU OF PRiSONS, NO. 1480.05, PROGRAM STATEMENT § 7(d) (2000)).
199. Id

200. Id at 804-05.

201. 417 U.S. 817 (1974).

202. 417 U.S. 843 (1974).

203. Hammer, 570 F.3d at 805.

204. Id at 809 (Wood, J., dissenting) (internal quotation marks omitted).
205. Id
206. Id

207. Id.(quoted statement by Warden Lappin).
208. Id.
209. See Epstein et al., supra note 9.

210. See FED. R. App. P. 28(a)(9)(B), (b)(5) (the former governing the appellant’s brief and the latter governing the appellee’s brief).

211. According to the 2010 Year-End Report on the Federal Judiciary submitted by Chief Justice Roberts, the total number of cases filed in the Supreme Court during the 2008 Term was 7738; during the 2009 Term it was 8159. JOHN ROBERTS, 2010 YEAR-END REPORT ON THE FEDERAL JUDICIARY app. at 9 (2010), availableat http://www.supremecourt.gov/publicinfo/year-end/2010year-endreport.pdf.
212. See Epstein et al., supra note 9, at 104.

213. See id

214. AO 2010 REPORT, supra note 2, at 44 thl.S-1.