29 VTLR 555 29 Vt. L. Rev. 555 (Cite as: 29 Vt. L. Rev. 555) Vermont Law Review Winter, 2005 Article *555 JUST WORDS? THE EFFECTS OF NO-CITATION RULES IN FEDERAL COURTS OF APPEALS Jessie Allen [FNa1] Copyright Copr. 2005 by Vermont Law School; Jessie Allen TABLE OF CONTENTS Introduction .............................................................. 556 I. The Use of No-Citation Rules ........................................... 561 A. How No-Citation Rules Developed ...................................... 561 B. No-Citation Rules Today .............................................. 568 II. No-Citation Rules and the Due Process Right to Be Heard ............... 574 A. The Basic Argument: Consistency Is Deeply Associated with Fairness and Correctness ......................................................... 575 B. Applying Due Process Doctrines ....................................... 583 III. The Impact of No-Citation Rules on Courts' Precedential Practice ..... 591 A. The Connection between Availability of Information about Prior Cases and Courts' Adherence to Precedent ...................................... 592 B. How No-Citation Rules Reshape Precedential Practices ................. 594 C. Allowing Citation of Nonbinding Opinions for a More Flexible Precedential Practice ................................................... 601 IV. Precedent, No-Citation Rules, and Courts' Power to Affect Society ..... 603 A. Precedent and Judges' Transformative Verbal Power .................... 604 B. No-Citation Rules and Judicial Action ................................ 608 *556 INTRODUCTION Morality or custom may be embedded in human behavior, but law--virtually by definition--comes into being through language. Peter M. Tiersma [FN1] The object of our study, then, is prediction, the prediction of the incidence of the public force through the instrumentality of the courts. The means of the study are a body of reports, of treatises, and of statutes, in this country and in England, extending back for six hundred years, and now increasing annually by hundreds. In these sibylline leaves are gathered the scattered prophecies of the past upon the cases in which the axe will fall. These are what properly have been called the oracles of the law. Oliver Wendell Holmes [FN2] Judicial power is word power. Judges decide the cases before them with words that change our lives. This article is about what happens when judges then refuse to hear what they have said. Four of the largest and most influential federal courts of appeals forbid discussion of many of their decisions by litigants in subsequent cases. [FN3] The effects on practice can be bizarre: A lawyer arguing in the Ninth Circuit may discuss cases decided in Australia 100 years ago, but may not mention apparently relevant decisions made last year by the very judges who will decide her case. Those unmentionable cases are not just a few peripheral decisions. In 2003, the Ninth Circuit decided more than 5,000 appeals. [FN4] Citation to some 4,400 of them is prohibited. [FN5] At this rate approximately 84% of the Ninth *557 Circuit's recent output is off limits to advocates arguing before that court and to trial judges in the circuit. Other courts with citation bans are not far behind. In 2003, the Second Circuit prohibited the citation of 75% of its decisions, and the Seventh Circuit forbade citation of 57% of its output. [FN6] In response to complaints from the appellate bar, a new rule of federal appellate procedure was proposed that would do away with the citation bans. [FN7] The no-citation circuits strongly resisted it, however, and the proposed rule has at least temporarily stalled. [FN8] My guess is that most Americans would be amazed to learn that such a rule is necessary, let alone that it is being contested. Doesn't it go without saying that judges look back at their previous rulings when they decide new cases? What is going on here? Though citation bans seem strangely at odds with standard notions of how American courts operate, there are plausible arguments to be made in their favor. Most importantly, they respond to heavy case loads that make it impossible for appellate judges to issue careful and detailed written *558 opinions in every case they decide. [FN9] Judges use summary procedures for deciding routine cases, often providing explanations of those decisions via lightly edited memos written by clerks and staff attorneys. [FN10] With no-citation rules in place, judges can be confident that the hastily reviewed legal reasoning in these summary decisions will not come back to haunt them. Judges say that the time saved by not stopping to carefully craft explanations of routine rulings allows them to concentrate on cases that present more complex, less familiar, and more widely significant legal questions. [FN11] Courts with no-citation rules distinguish sharply between the results and the reasoning of their summary decisions. They stand by the results, but they say the explanations of those results are questionable. [FN12] Some federal courts of appeals assert that summary decisions are not binding in future cases. [FN13] The no-citation courts go one step further and prohibit any *559 mention of such "nonprecedential" cases before courts in their jurisdiction. [FN14] These are routine cases whose outcomes are squarely controlled by existing precedents, the courts say, so adding them to the precedential pool would only be confusing. [FN15] Barring their citation makes it unnecessary for judges to consider these summary decisions, or to explain subsequent departures from their reasoning. This, judges say, relieves pressure to spend additional time clarifying the decisions in these routine, redundant cases, allowing judges to focus their attention on the more difficult and important precedential cases and ensuring that half-baked legal explanations do not work their way into caselaw. [FN16] There are reasons to question courts' justifications for citation bans. For one thing, anecdotal evidence suggests that courts do not use uncitable opinions only in easy cases with outcomes that are legally obvious. [FN17] Uncitable summary decisions are by no means limited to affirmances, and sometimes even carry dissents. [FN18] That means that in some of these cases, federal judges disagreed with the results as well as the reasoning. If the legally correct outcome is so clear, why the dispute? It is hard to see how all those disputed uncitable cases fail to meet any of the criteria for a *560 precedent, i.e., that it "[e]stablishes, alters, modifies or clarifies a rule of law." [FN19] Be that as it may, this article concentrates on the effects of no-citation rules that arise even if those rules are being used just as courts say they are. I consider the meaning of those effects both for litigants and for the role of courts in society. The article begins by reviewing the development of citation bans and examining in greater detail the reasons courts give for maintaining them. The following sections examine the citation bans' effects from several different planes of analysis. Part II of the article looks at the problems citation bans create for litigants as a potential violation of procedural due process. Courts with no-citation rules in effect refuse to hear litigants complain that a challenged judgment is inconsistent with the court's decisions in routine cases or that the legal rule the litigant is urging the court to follow has been previously applied, perhaps numerous times. In a legal culture that strongly associates consistency with fairness and correctness in decisionmaking, this refusal is not trivial. In Part III of the article, I consider some institutional effects of no-citation rules from a policy perspective. In my view, no-citation rules contribute to three interrelated developments in courts' precedential practices. The first of these is the institutional counterpart of litigants' inability to argue that judges should follow their decisions in most routine cases: the rules turn judges' attention from deciding all like cases alike to maintaining coherent written caselaw. Courts' primary function becomes a treatise-like exposition of precedential rules rather than decisionmaking that also produces precedent. Citation bans also split judges' lawmaking--or law-interpreting--function from the actual process of deciding cases, and they largely undo the reciprocal relationship between creating and applying precedent. None of these changes is necessarily bad. They are, however, fundamental shifts in judicial practice that have gone largely unacknowledged by courts that use no-citation rules. In particular, notice that these developments move adjudication away from a common-law model and toward a civil-code system that maintains a separate set of legal rules that are applied in cases whose outcomes do not affect the content of the rules themselves. Finally, in Part IV of this article, I offer some observations about how following precedent may enhance courts' power to transform society and how no-citation rules could undermine that power. This is a phenomenological investigation that draws on examples from personal *561 experience and anthropology. The basic notion is that by adopting a precedential practice, courts commit themselves to a kind of verbal irrevocability. Though irrevocability is a common feature of everyday embodied life--no use crying over spilled milk--it is less generally attributed to language. Perhaps through this association with physical action, judges' precedential words acquire a transformative effectiveness also more typically associated with physical intervention. If judges' words have rhetorical power to shape our social world--beyond threatening to punish disobedience with governmental force--precedent may be a crucial component of that power. But no-citation rules work against the association of judges' decisions with action. The citation bans make it possible for judges to control which of their decisions will have long-term legal consequences, a control that is absent from the chancy world of embodied activity. Thus, no-citation rules tend to move judges' pronouncements back out of the realm of consequential action and into the category of ordinary communication. By trying to increase their power to determine future judicial behavior, judges on no-citation courts might risk losing some of their power to influence behavior in the real world. No-citation rules may increase the coherence of courts' written precedents, but decrease those precedents' transformative effects as we begin to see that judges' decisions are, after all, "just words." I. THE USE OF NO-CITATION RULES A. How No-Citation Rules Developed There is general consensus that the practice of issuing unpublished opinions whose citation is forbidden or discouraged was adopted in the 1970s because of what judges saw as an unmanageable growth in reported caselaw. [FN20] The use of such opinions then increased dramatically in response to exponential growth in the federal appellate caseload. [FN21] Without *562 challenging the primary causal role played by swelling case reporters and expanding appellate dockets in the courts' adoption of non-publication policies and citation bans, I want to look briefly at two larger trends that were part of the world in which these practices began and flourished. The social and historical context in which federal courts' summary procedures developed is rarely discussed, but it seems worth considering as a factor in their adoption and their persistence. After all, judges had been decrying the growth in caselaw at least since the eighteenth century, [FN22] and the simple expansion of appellate dockets does not explain why courts responded with summary procedures rather than calls for more judges. 1. Original Rationales The practice of designating some court decisions "nonprecedential" and withholding their publication is generally traced to the 1964 Federal Judicial Conference, which noted "the rapidly growing number of published opinions" and resolved to "authorize the publication of only those opinions which are of general precedential value." [FN23] In 1971, a study group authorized by Congress to recommend changes in federal judicial administration reported that there was "widespread consensus that too many opinions are being printed or published." [FN24] In the years that followed, the study group recommended that the Judicial Conference ask each federal circuit court to come up with a plan to reduce its publication of opinions and to restrict citation of unpublished opinions. [FN25] By 1974, all the circuits had some kind of rule limiting publication. [FN26] Originally, no-citation rules insured that non-publication did not create advantages for institutional litigants who might have special access to unpublished decisions. The fear was that only insiders would know about decisions in unpublished cases. The courts thus prohibited everyone from citing unpublished decisions in order to avoid privileging any party who might have special access to sources not generally available. [FN27] Most *563 commentators also agree that the spread of non-publication practices to a majority of the courts' decisions was a response to the rapidly expanding volume of appellate dockets. [FN28] Judges simply did not have time to craft full-fledged written opinions in all their cases, or, eventually, in most of their cases, and increasingly turned to the new summary procedures. It is striking, however, that court procedure was so readily altered in such a basic way with so little consideration of alternative ways to meet the new demands courts faced. In particular, there seems to have been no serious effort to respond to the increased caseload by significantly increasing the number of federal appellate judges. [FN29] 2. No-Citation Rules and "Too Much Law" If judges' time was the main concern, why was there not a push for more judges, rather than allowing summary practices to change longstanding judicial process? There is no clear answer to this question. Today, adding the number of new judges that would be required to fill the gap is neither politically nor institutionally feasible. [FN30] No one believes that Congress is likely to suddenly increase the federal appellate bench by "a factor of five," and even if that were possible, such an expansion would swell existing circuits to the kind of numbers that are currently debilitating the Ninth Circuit's operation or would suddenly add several new courts and a vastly increased chance of non-uniform national law. [FN31] But at the beginning of the big rise in caseload, federal appeals courts might have pushed for expansion. Some commentators have suggested that the federal courts resisted expanding because judges wanted to preserve the prestige *564 and institutional power that comes from exclusivity. [FN32] Perhaps more to the point, increased numbers of judges would not have helped with the problem that the nonpublication policies were ostensibly designed to solve--the proliferation of published precedents. Moreover, the nonpublication and no-citation rules were already in place when the steepest rises in caseload were recognized; at that point summary procedures may have seemed like a less dramatic institutional solution than significantly expanding the appellate bench. Certainly summary procedures could be accomplished far more easily, without involving significant congressionally authorized budget expansions. Whatever motivated the initial adoption of summary procedures, I want to consider two other trends that have tended to normalize federal courts' new summary procedures, including no-citation rules, despite their divergence from traditional appellate procedures. Around the same time that federal appellate courts were adopting these practices, a new critique of the legal system was developing in the academy and in the popular press. Beginning in the 1970s, scholars and pundits began to warn that there was something wrong with law in America--basically, there was too much of it. [FN33] According to critics, the United States was in the midst of a "litigation explosion" [FN34] set off by a "growing contentiousness" [FN35] among Americans, who for some reason were suddenly eager to haul neighbors, bosses, and government officials into court. This urge to litigate was portrayed as sociopathic--a kind of legal hypersensitivity driving people to court to complain about injuries that in previous eras would have been stoically ignored or dealt with in private. "[W]hen people feel wronged by another *565 person or institution, the immediate reaction is not to turn the other cheek, but to serve process and bring suit," scolded one federal appellate judge. [FN36] In addition, both the "too much law" [FN37] critique and the rising rates of appeals came at the end of a long-term change in the subject matter of litigation in this country, especially federal litigation. [FN38] In federal court, civil rights filings were rising on a raft of new legislation: the Civil Rights Act of 1964; [FN39] the Voting Rights Act of 1965; [FN40] the Age Discrimination in Employment Act of 1967; [FN41] and the Civil Rights Act of 1968. [FN42] Filings were also rising in response to the Supreme Court's reinvigoration of longstanding civil rights statutes. [FN43] The Court's interpretations of the Fourteenth Amendment, its decision in Brown v. Board of Education, [FN44] and its other civil rights cases invited more constitutional civil rights litigation. The Criminal Justice Act of 1964 provided free lawyers for criminal appeals in federal cases, and after that, unsurprisingly, rates of appeal from federal convictions rose rapidly. [FN45] Congress also passed laws in the 1960s that authorized people to appeal decisions of federal agencies to federal courts. [FN46] Thus, besides *566 hearing many more cases than they had previously, judges on federal courts of appeals were being asked to decide different kinds of cases. Moreover, changes in federal court dockets were only a small part of a much broader shift in courts across the country "from business and property cases to tort, criminal law, and public law." [FN47] The changing subject matter of litigation had implications for the existing social hierarchy. As Marc Galanter puts it, "There were more opportunities for successful assertion of rights by outsiders, dependents, and subordinates against society's managers and authorities." [FN48] So, while nonpublication and no-citation rules may have spread in response to rising caseloads, it was not just the volume of appeals that had changed. The content of those cases was different, too. [FN49] Meanwhile, there was a growing popular tendency to portray Americans as obsessed with bringing lawsuits for ephemeral injuries or political purposes that had no place in a court of law. [FN50] With this broader context in mind, it may be less surprising that the federal appellate courts allowed novel summary procedures to take over to such an extent. After all, the problem, at least as some people saw it, wasn't just too much of the same old thing. The "flood," or "explosion," [FN51] or "tidal wave" [FN52] of claims threatening to overwhelm the courts was full of novel rights and actions brought by litigants rarely seen before in federal courts of appeals. [FN53] The academic and popular press continued to warn that *567 Americans' new-found penchant for suing one another was socially and economically destructive. [FN54] From that perspective, treating all the new cases with traditional process might have seemed irresponsible. If the new onslaught of claims was driven by an unhealthy excess appetite for litigation, then it would be reasonable to dispose of them through new summary procedures that reduced the amount of legal process they received. I am not arguing that the decision to allow unpublished, uncitable opinions was a direct response to these social trends. The idea that Americans were litigating too much and in novel, sometimes inappropriate, ways may never have entered the minds of the judges who signed off on the 1964 Judicial Conference's call for nonpublication policies or the judges who crafted the individual courts' no-citation rules. But as part of the background against which the citation bans developed, the critique of "too much law" tended to normalize them. It is unclear to what extent nonpublication and no-citation practices were originally conceived for use on certain types of litigation. The courts that retain no-citation rules apparently do use them somewhat more frequently for certain types of claims. What really stands out in today's picture, however, is the extent to which unpublished, uncitable decisions now pervade every substantive area of the courts' dockets. If these summary procedures were devised with certain kinds of cases in mind, they are now ubiquitous. In today's no-citation courts, the majority of all types of cases are being adjudicated with decisions labeled "nonprecedential" whose citation is forbidden. [FN55] *568 B. No-Citation Rules Today 1. Current Use Today, every federal circuit uses summary decisionmaking procedures to decide most cases. [FN56] The majority of federal appellate courts now allow citation of their summary decisions, although most assert either that such decisions are not "precedent" [FN57] or not "binding precedent" [FN58] or that their citation is "disfavored." [FN59] On the other hand, the Second, Seventh, Ninth, and Federal Circuits-four of the largest and most influential courts of appeals-still ban citation, and the Eleventh Circuit still keeps its summary opinions unpublished, suggesting that the court might wish to adopt a no-citation rule when publication becomes mandatory in 2005 under the E-Government Act. [FN60] A new Rule of Federal Appellate Procedure proposed by the Judicial Council's Advisory Committee on Appellate Rules would have mandated that all federal appellate courts' opinions be available for citation. [FN61] The proposed rule was tabled by the Federal Judicial Council's Standing Committee, however, and thus has not reached the full Council for adoption. When it was published for public comment, the rule drew some 500 responses, the vast majority of which opposed its adoption. [FN62] If anything, the proposed uniform rule seems to have hardened resistance to allowing citation in courts that now impose citation bans, as judges on these courts have adamantly defended the bans as fully constitutional and institutionally optimal policy responses to overloaded appellate dockets. [FN63] *569 2. Courts' Reasons for Imposing Citation Bans Nowadays, the practice of issuing uncitable decisions is defended primarily as an aid to judicial efficiency and precedential clarity. [FN64] The idea is that judges gain time to work on the small number of designated precedents by issuing summary opinions, which are not binding on subsequent decisions, in most of their cases. And by limiting precedential caselaw, courts say, they keep the law coherent. "Restrictive rules regarding unpublished dispositions 'keeps the books from being cluttered with dicta that could result in confusion for lawyers and tribunals addressing similar issues."' [FN65] Conversely, allowing open citation of summary decisions is said to create confusion and opportunities for twisting the law: "Courts contribute to the growing imprecision, uncertainty and unpredictability of the law by issuing repetitive opinions on subjects that have been thoroughly irrigated." [FN66] Occasionally some of the rhetoric of natural catastrophe from the "too much law" critique reappears in defense of citation bans: the Federal Circuit explains that courts should strive "not to add to the explosion of legal opinions." [FN67] A Ninth Circuit judge warns that the new rule requiring open citation "will punch a hole in one of the few dikes judges have to protect themselves from inundation by the ever rising tide of litigation that besets the court system." [FN68] And sometimes courts suggest that limiting access to judicial decisions is a matter of duty, rather than expediency. "Maintaining a coherent, consistent and intelligible body of caselaw is not served by writing more opinions," admonished the Ninth Circuit in Hart, "it is served by taking the time to make the precedential opinions we do write as lucid and consistent as humanly possible." [FN69] As for efficiency, judges on courts that bar citation say that if their summary decisions could be cited, judges would feel bound to spend more *570 time on them, either out of a concern that they would be mistaken by district judges for the law of the circuit [FN70] or because appellate judges in subsequent cases would have "a moral duty to explain, distinguish, reaffirm, overrule, etc. any unpublished order brought to [their] attention by counsel." [FN71] No-citation rules guarantee that the hastily reviewed language of summary opinions will not come back to haunt judges in future cases. [FN72] "If unpublished dispositions could be cited as precedent, conscientious judges would have to pay much closer attention to their precise wording," explains Judge Kozinski of the Ninth Circuit. [FN73] When pressed, judges on no-citation courts offer several different reasons why summary decisions would be inadequate in their current form if they could be cited. One argument is that allowing citation is tantamount to making summary decisions binding precedent. [FN74] In this view, if cases can be cited, it is illogical to treat them as nonprecedential. Judge Kozinski explains that "by citing what a court has done on a previous occasion, a party is saying: This is what that court did in very similar circumstances, and therefore, under the doctrine of stare decisis, this court ought to do the same." [FN75] That is certainly one reason for citing the court's previous decisions. Another possibility is that the litigant wants to demonstrate the court's application of contested precedents in recent routine cases. Arguing that the court should follow a five-year-old precedent is far easier if one can show that the court has invoked that case in numerous recent decisions. Conversely, one can more easily distinguish an arguably applicable precedent if it can be shown that the court has routinely read that case narrowly in subsequent decisions. But whether the focus is consistency with the reasoning in a particular summary decision or with recent patterns of precedential application in routine cases, it does not follow that citing those cases automatically makes them binding on subsequent panels. To put it another way, the baseline norm of consistent treatment is related to, and overlapping with, the doctrine of binding judicial precedent, *571 but not equivalent to it, at least not as that doctrine is generally understood to apply in the federal courts of appeals today. A litigant might quite coherently ask to be treated consistently with the court's previous summary decisions without asserting that the panel of judges adjudicating her claim is bound to follow those decisions, like it or not. Certainly citing prior decisions of the court deciding one's case has a different argumentative purpose than citing decisions by other courts, or articles by law professors. Telling a decision maker that she, or her colleagues, did what you want her to do once--or many times--puts some burden on the decision maker to at least acknowledge a departure from that previous position. This is precisely the burden the no-citation circuits reject. "Citability would upgrade case-specific orders that this circuit has intentionally confined to the law of that particular case to de facto precedents that we must address," say some Seventh Circuit judges. [FN76] Undeniably, citation does confer some obligation to address, or at least to silently consider, the referenced decisions. As the Ninth Circuit explained in Hart, regarding cited opinions from other jurisdictions, a court "would consider it bad form to ignore contrary authority by failing even to acknowledge its existence." [FN77] It seems a court would have at least a comparable obligation to consider any of its own decisions cited in the parties' briefs. But that is far from saying that cited summary decisions must then have the effect of opinions federal appellate courts consider binding precedent, which a panel is bound to follow, or distinguish, unless they are overturned by the full court sitting en banc. Whatever one's views about the practical strength of that precedential rule, it is clear that simply allowing citation need not trigger that kind of institutional obligation. After all, as the Ninth Circuit described, the obligation to consider cited cases from other courts does not trigger an obligation to follow those outside authorities. One federal appellate court (which allows citation of its summary orders) has drawn this distinction, conceding that "any decision is by definition a precedent, and ... we cannot deny litigants and the bar the right to urge upon us what we have previously done," yet maintaining that subsequent panels are not strictly bound to follow cited summary decisions. [FN78] Without that strict binding effect, judges confronted with summary decisions would not necessarily feel pressure to give lengthy *572 explanations (or any explanations at all) for taking a different approach. In fact, the explanation now given for no-citation rules might often suffice as a reason to diverge from a given summary decision: this opinion was produced in a summary process that did not include careful articulation of the reasoning at work. There seems to be an abiding concern in the no-citation courts, however, about dealing directly with the limitations of summary decisions. In opposition to the proposed uniform federal rule allowing citation, a group of Seventh Circuit judges protested, "if a lawyer states in its brief that in our unpublished opinion in A v. B we said X and in C v. D we said Y and in this case the other side wants us to say Z, we can hardly reply that when we don't publish we say what we please and take no responsibility." [FN79] But that circuit's absolute ban on citation must mean something fairly close to that. Putting it somewhat more neutrally, the rule seems to say that because summary decisions are not articulated the court will not be responsible for dealing with them in subsequent cases. What is not clear is why it should be institutionally acceptable to take that stance implicitly through the no-citation rule, but not to adopt it openly in response to a litigant's citation of a previous summary decision that the panel wants to depart from or disregard. Nor is it clear, as a matter of fact, that knowing they can be confronted with summary decisions will drive most judges to devote significantly more time to them than they now do. As Judge Easterbrook, a Seventh Circuit judge who supports open citation, has pointed out, "It has never been true that judges write these orders for the parties and counsel alone, and thus are certain to include more (or less) when strangers can use them; the audience always has included the Supreme Court, which can and does review unpublished decisions." [FN80] Moreover, these opinions are already publicly accessible in searchable form. [FN81] Something between one-third and one-half of the appellate attorneys surveyed in the no-citation circuits said they read the courts' unpublished decisions when they come up in their research. [FN82] *573 Given how widely available and widely read summary opinions are already, allowing their citation may add little to existing pressures to spend time perfecting them. Some judges assert, however, that if citation is allowed, rather than spending more time on summary decisions, they will stop issuing them altogether and decide cases through one-word dispositions: "Affirmed." [FN83] This may be a harder point to answer. One can say, however, that there has been no evidence so far that this is happening in federal circuits that allow citation. Moreover, the one-word approach is clearly untenable in a significant number of summary decisions in which the word necessarily would be not "affirmed" but "vacated" or "reversed." Without some explanation of the appellate court's reasoning, the district judge whose opinion has just been reversed will not know how to proceed on remand. More fundamentally, courts' power and legitimacy is bound up with judges' willingness and ability to present reasoned explanations for their rulings. Federal judges do not represent the popular will the way legislators do. [FN84] Thus their lawmaking--or law interpreting--authority needs a basis in reasoning and/or in precedent. Regarding the nature of judicial power, the Supreme Court has observed that "a decision without principled justification would be no judicial act at all." [FN85] A court that declines to verbalize the grounds for most of its decisions is a court that risks losing the respect, and perhaps even the compliance, of the people it purports to govern. Finally, it is not entirely clear that judicial time and attention in no-citation circuits is now optimally divided between citable and uncitable cases. The wide disparity between the time and effort judges currently spend on precedential and summary decisions is justified by the belief that summary decisions are dictated by existing law. But in at least one no-citation circuit, the number of dissenting opinions in uncitable cases calls that judgment into question. In the first two months of 2004, the Ninth Circuit issued fifteen dissents in uncitable summary dispositions. [FN86] Testifying before a legislative hearing on no-citation rules, Judge Kozinski asserted that "[u]npublished dispositions are cases that are squarely controlled by existing precedent, squarely controlled by existing Ninth *574 Circuit opinions, that and nothing more." [FN87] But the number of dissents in those cases suggests otherwise. Perhaps some of the time judges now use perfecting the language of the few designated precedential cases might be better spent working out the rulings in disputed summary cases. II. NO-CITATION RULES AND THE DUE PROCESS RIGHT TO BE HEARD The "right to be heard" lies at the heart of due process, as it has been conceived by American courts. [FN88] Institutionally, due process ensures fairness. [FN89] Though many imaginable procedures would provide some assurance that government officials are acting fairly, in American caselaw the concept of due process has centered on giving those affected "an opportunity to be heard respecting the justice of the judgment sought." [FN90] The notion is that before government acts to deprive someone of liberty or property, "a person has an opportunity to speak up in his own defense, and ... the State must listen to what he has to say." [FN91] As Judge Henry Friendly explained, a right to be heard includes "an opportunity to present reasons why the proposed action should not be taken." [FN92] The state can act legitimately to disadvantage someone only after hearing what the disadvantaged person has to say about it. [FN93] *575 Courts with no-citation rules refuse to hear litigants argue, "you should do for me what you previously did in this summary decision--or in this series of summary decisions." [FN94] Does that refusal violate a litigant's right to be heard? Whether or not a due process challenge to no-citation rules would succeed under current doctrinal standards, a due process analysis shows that the intuitive sense of unfairness and arbitrariness people often express when they first learn about citation bans can be articulated in coherent constitutional terms. A. The Basic Argument: Consistency Is Deeply Associated with Fairness and Correctness. 1. Fairness The idea that fairness entails consistency runs through both formal moral philosophical theories and homespun values. As Frederick Schauer has observed, "Whether expressed as Kantian universalizability, as the decisions that people would make if cloaked in a Rawlsian veil of ignorance about their own circumstances, or simply as The Golden Rule, the principle emerges that decisions that are not consistent are, for that reason, unfair, unjust, or simply wrong." [FN95] Among other things, the idea of consistency has a central place in the constellation of concepts that we call "rule of law." [FN96] The basic idea is that consistency indicates a neutral application of principles, regardless who is affected. Thus, part of how we evaluate the fairness or arbitrariness of government decisions is by judging their consistency with other government actions in similar situations. In modern adjudication, the norm of consistency is, of course, most fully embodied in the doctrine of precedent. But it does not cease to operate at a more basic level simply because a court announces that most of its decisions are "non-precedential." The pursuit of consistent outcomes has long been recognized as a guiding principle of judges' work. "It will not do to decide the same *576 question one way between one set of litigants and the opposite way between another," said Justice Cardozo. [FN97] The core notion is not dependent on courts' institutional commitment to a specific doctrine of precedent. As Judge Henry Friendly put it, the duty to "act alike in all cases of like nature" is "the most basic principle of jurisprudence." [FN98] And that basic principle has been repeatedly articulated in Anglo-American caselaw back through the centuries. These are aspirational principles. Courts are not held to a standard that requires them to produce entirely consistent outcomes, but they are expected to strive for consistency to the extent possible. [FN99] Consistency is a perennial judicial goal in part because of its association with correctness and predictability, but the core value it expresses is evenhandeness--justice that does not vary depending on who it affects. 2. Consistency in Administrative Law Cases reviewing agency action are a likely source of courts' views on the role of consistency in adjudication, because this area of the law is all about evaluating government decisionmaking. Administrative caselaw recognizes the importance of consistency as a hallmark of fairness, an indication that a government decision is correct, and a necessary component of governmental predictability. There are two basic ways that the law of administrative procedure and review makes consistency a factor in assessing the validity of government action. First, through the doctrine of Skidmore deference, [FN100] a reviewing court considers consistency with previous agency actions in determining how skeptically to treat a challenged agency decision. Second, the cases treat consistency as an index *577 of a challenged agency decision's substantive correctness, and of the basic rationality for which all government decisions are responsible. [FN101] a. Skidmore Deference The respect a court will accord an administrative agency's judgment depends in part on the judgment's consistency with the agency's prior determinations. This principle was most famously articulated in Skidmore v. Swift & Co., a 1944 case in which the Supreme Court said that the weight a court gives to an agency's judgment in a particular case depends on "the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking in power to control." [FN102] Though Skidmore has been limited, the role of consistency as a factor in the court's review of agency judgments has survived the watershed decision in Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., [FN103] and continues to be active today. [FN104] Indeed, in Chevron itself the Court applied this principle, finding it necessary to rationalize the varying interpretations that the agency gave to the legislation in question by explaining that they did not indicate an inconsistent view of the act, but rather that the agency had "consistently interpreted it flexibly." [FN105] Cases decided after Chevron confirm the continued vitality of consistency as a factor in courts' deference to administrative decisionmaking. [FN106] In a case involving banking regulations, the Supreme Court explained that while a self-contradictory decision was not necessarily "fatal," a "[s]udden and unexplained change ... may be 'arbitrary, *578 capricious [or] an abuse of discretion."' [FN107] Note that the Court has recognized the importance of consistency not just in the rulemaking context, but where decisions being evaluated are the judgments of a tribunal adjudicating individualized claims. In Cardoza-Fonseca, for example, the Court refused to give heightened deference to the INS's interpretation of a standard for denying asylum to political refugees in part because of the inconsistency in the agency's position over the years. [FN108] The Court explained: "An agency interpretation of a relevant provision which conflicts with the agency's earlier interpretation is 'entitled to considerably less deference' than a consistently held agency view." [FN109] The pattern of inconsistent decisionmaking included judgments the Bureau of Immigration Appeals made in deciding individual applicants' appeals of denials of asylum. [FN110] b. Consistency as a Substantive Test of Correctness Beyond questions of deference, the cases on administrative review sometimes look to consistency as part of a substantive evaluation of the correctness of an agency decision. Inconsistency raises concerns about reliance and arbitrariness. A well-recognized administrative law principle requires remand of any agency decision that diverges without explanation from a settled course of agency action. The emphasis here is on providing reasons for the new course. So it is not the case that agencies must always consider, let alone defer to, their previous decisions. Nevertheless, the decisions on this principle articulate a strong relationship between consistency and correctness. "A settled course of behavior embodies the agency's informed judgment that, by pursuing that course, it will carry out the policies committed to it by Congress." [FN111] An agency's decisions, then, *579 create "at least a presumption that those policies will be carried out best if the settled rule is adhered to." [FN112] In addition, if agencies ignore contradictory past decisions when they decide similar issues in new contexts, those previous decisions cease to "check arbitrary agency action." [FN113] Here we return to the association of inconsistency with arbitrariness. Though judges may not be bound to follow a previous ruling, they are not free to cover their ears when someone tries to tell them that their recent summary rulings demonstrate a consistent pattern of decisionmaking that runs counter to a ruling being appealed. The basic responsibility of judges to decide rationally the case before them forbids such willful ignorance. As the Ninth Circuit acknowledged, even cases from other courts must be "acknowledged and considered" in order to comply with the deciding judges' "common law responsibilities." [FN114] Courts may not simply "ignore contrary authority by failing even to acknowledge its existence." [FN115] Particularly when a court has made a number of recent summary rulings in cases similar to the one at hand, the refusal even to consider how those rulings inform the decision before them smacks of arbitrariness, and "[t]he touchstone of due process is protection of the individual against arbitrary action of government." [FN116] Neither administrative agencies nor courts are required to always produce consistent decisions--indeed, other values may sometimes require inconsistency in order to achieve a correct result. But the administrative caselaw makes government agencies responsible not to arbitrarily disregard previous patterns of decisionmaking when deciding subsequent cases. Likewise, courts cannot arbitrarily prevent advocates from arguing that subsequent cases should be decided consistent with those prior decisions. Of course, the courts' response is that they are not arbitrarily disregarding their previous summary decisions. Those summary decisions should be ignored, no-citation courts say, because in some important ways they are not full-fledged judicial decisions *580 and because they are so deeply flawed that they offer no valid information about the judicial decisions that they represent. If it were true that summary opinions are not really official adjudicative decisions of the courts, that would be a good reason to disregard them. Again, the agency cases offer parallels, sometimes finding an alleged inconsistency is insignificant exactly because the previous rulings were not full official acts by which the agency's policy should be judged. "We doubt whether either of these statements was sufficient in and of itself to establish a binding agency policy," said the Court in Smiley, reviewing apparently inconsistent statements of the Office of the Comptroller. [FN117] One of the two statements at issue, for instance, was a "letter from the Comptroller" to an executive committee, which the court found "too informal" to create or reflect agency policy. [FN118] Might not the courts' unpublished opinions likewise be useless as benchmarks for the fairness and correctness of subsequent judicial decisions? The rejection of the letter in Smiley is particularly suggestive, because one justification for prohibiting citation of summary decisions is that they were never intended to be public statements of the courts' views on legal issues. The Ninth Circuit has explained that a summary decision is "more or less, a letter from the court to parties familiar with the facts, announcing the result and the essential rationale of the court's decisions." [FN119] The problem is that unlike the letter in Smiley, the Ninth Circuit's "nonprecedential" summary dispositions are otherwise treated as public court judgments whose results command obedience because the court warrants they are legally correct: "That a case is decided without a precedential opinion does not mean it is not fully considered, or that the disposition does not reflect a reasoned analysis of the issues presented." [FN120] Thus the judgments in such cases must be the sorts of government decisions against which other judgments can be tested. [FN121] *581 The no-citation courts are thrown back, then, on their second rationale for preventing discussion of the majority of their recent decisions. Summary decisions, they say, are so poorly articulated that they do not actually communicate their own results in a meaningful or useful way. [FN122] On this view, because the results of cases adjudicated with summary opinions are only available through the opinions' deeply flawed descriptions and explanations, barring their citation is justified. Ninth Circuit uncitable opinions "have zero precedential value," argues Judge Kozinski, "no inference may be drawn from the fact that the court appears to have acted in a certain way in a prior, seemingly similar case." [FN123] They may even be positively misleading because they reiterate precedential legal principles in slightly different terms that create apparent nuances or ambiguities where none were intended. [FN124] The concern is both that judges in subsequent cases will be deceived into changing rulings by false presentations of what the court has done in the past and that cited language will become part of the caselaw, muddying up the clarity of judicial legal rules. [FN125] If summary decisions are actually so error ridden or fictional that they provide no useful information about the courts' decisions, one wonders what good they do for the parties to whom they are ostensibly addressed. And the courts' decision to publish them is incomprehensible. Presumably courts publish these "letters" to the parties in order to alleviate concerns about private law and to provide a way for interested members of the public to review courts' actions in some area of law. They provide a source of institutional accountability. But if they have enough informational value to help provide answers about the courts' activities from an accountability or general public information standpoint, why isn't that information potentially useful to judges adjudicating subsequent cases? *582 The concern about deception seems misplaced. Who, after all, is better equipped to recognize the limitations of summary opinions than judges on the court that produced them or district judges whose decisions that court reviews. Judges can take or leave these imperfect expressions of routine rulings for what they are worth. Worries about precedential clarity are reasonable to an extent, but, in another sense this kind of ambiguity is the inevitable result of a precedential system. Indeed, one person's ambiguity is another person's crystallization. Paradoxically, repeated application of precedential rules in numerous similar factual scenarios both confuses and refines the meaning of the rules. That is the nature of precedent. The relationship between precedential articulation and application is discussed further in Part III below. It is important to remember that the potential value of these decisions is not only, perhaps not mostly, the reasoning of any one summary decision. Of at least equal interest are the patterns of application of the courts' precedents in recent routine cases. Additionally, courts need to consider the possibility that excising summary decisions from the mix of citable caselaw may distort the precedential record by omission. This distortion need not occur because judges are manipulating the system to hide certain types of decisions; it can come about by accident. By definition, summary decisions are routine cases whose outcome is easily determined by existing precedent. [FN126] Thus, summary decisions include significantly higher rates of affirmances than the courts' precedential caselaw (because the three appellate judges and the district judge who initially decided the case are likely to agree on the proper result). [FN127] Conversely, reversals of lower court decisions are overrepresented in citable precedent. That skewed result can have substantive ramifications. Consider, for instance, challenges to criminal convictions or civil jury verdicts based on trial judges' alleged failure to exclude improper evidence. Routine decisions to affirm convictions because trial judges' evidentiary rulings were either correct or harmless will be overrepresented among uncitable summary decisions and underrepresented in citable precedent. *583 Meanwhile the occasional decision to overturn a jury verdict based on wrongly admitted prejudicial evidence will be more likely to appear as a citable precedent. That means that a no-citation court's available caselaw will exaggerate appellate panels' willingness to overturn jury verdicts for evidentiary trial errors. In effect, it will represent the court's precedential rules on evidentiary exclusion and the standard of appellate review as more stringent than they actually are. Unlike the false importance of slight variations in the articulation of familiar legal principles, this is a kind of distortion that might be very hard for district judges in the circuit to recognize. And it could certainly play to the disadvantage of parties who want to argue in support of verdicts being appealed, but find that many of the appeals court's most recent decisions upholding verdicts are uncitable. B. Applying Due Process Doctrines The Supreme Court has taken several different doctrinal approaches to evaluating procedural due process claims. It is far from clear under the caselaw which of the Court's due process doctrines would apply to a challenge to the no-citation rules. In claims of inadequate process in both criminal and civil trials, the Court has focused on whether the process at issue diverges from "some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental." [FN128] Under this line of cases, the focus would be on the no-citation rules' divergence from traditional common-law procedure. In another line of cases addressing limits on what litigants can say in court, however, the Court has asked whether those limits are "arbitrary or disproportionate to the purposes they are designed to serve." [FN129] Finally, some due process challenges to procedure in civil litigation use a balancing test first articulated in an administrative challenge, Mathews v. Eldridge. [FN130] The Mathews test takes a utilitarian approach, emphasizing accurate results as the goal of due process. Each of these three tests illuminates different aspects of the citation bans' effects on court process. [FN131] *584 1. Fundamental Fairness and Common-Law Tradition In some challenges to adjudicative process, the Supreme Court has looked to whether procedures diverge from common-law tradition, and whether they violate a "recognized principle of 'fundamental fairness."' [FN132] This approach is highly protective of procedures that can be identified as traditional: "If a thing has been practised for two hundred years by common consent, it will need a strong case for the Fourteenth Amendment to affect it." [FN133] Though conformity with common-law tradition does not entirely end the due process inquiry it provides a heavy presumption of rectitude. [FN134] On the other hand, divergence from common law may trigger a presumption of unconstitutionality. Striking down Oregon's procedures for judicial review of punitive damages awards, the Court explained that the state's "abrogation of a well-established common-law protection against arbitrary deprivation of property raises a presumption that its procedures violate the Due Process Clause." [FN135] Thus, under the fundamental fairness approach, most of the action will be in determining whether citation bans conform with traditional common-law process. This is not as straightforward as it at first may appear. Although formal citation bans are clearly a modern creation, arguably their effect of making only certain selected decisions available for citation restores a situation that existed for centuries before the adoption of universal official case reporting. [FN136] Thus, deciding whether citation bans diverge from tradition depends on whether one focuses on the judicial prohibition against citing what is available or on the practical availability of reported opinions for discussion. One commentator stresses that early English lawyers--and early American lawyers, too--were allowed to cite to all previous judicial decisions. [FN137] To the contrary, in Hart v. Massanari, the Ninth Circuit *585 pointed out that up until the late eighteenth century, few judicial decisions were reported and available for citation. [FN138] Arguably no-citation rules recreate that traditional situation, or at least do not diverge from it dramatically enough to offend some fundamental traditional principle of justice. The traditional reality, however, did not include the notion that judges could forbid the discussion of any available decision. [FN139] Hart contends that "common law has long recognized that certain types of cases do not deserve to be authorities," [FN140] but what the historical analysis in that case actually seems to show is that common-law practice accepted that chance and the preferences of practitioners would partly determine which judicial decisions were reported. That is not the same thing as allowing courts to prohibit litigants from discussing previous court cases and pleading for consistent treatment. [FN141] Traditionally, the freedom to cite previous cases was limited only by the practical availability of case reports and the breadth of the lawyer's knowledge and memory. [FN142] Several federal appellate judges, without going into history, have expressed the view that prohibiting citation may violate "fundamental fairness." A 1986 three-judge dissent to the adoption of the Tenth Circuit's no-citation rule explained that "as a matter of essential justice and fundamental fairness" litigants should be able to point to the court's prior decisions as a basis for a judgment. [FN143] Denying that right, the dissenting judges concluded, "may well have overtones of a constitutional infringement because of the arbitrariness, irrationality, and unequal treatment of the rule." [FN144] Today's limits result from judges' deliberate *586 refusal to hear about previous decisions they have chosen to make unmentionable, as opposed to a general lack of reliable case reports. For those who are focused on the practical availability of legal authorities, that is likely not a significant difference. But for those who are offended by a government official's refusal to confront her own past decisions, no-citation rules are both fundamentally different, and fundamentally unfair. 2. Arbitrary or Disproportionate Restrictions Another place to look for due process doctrines relevant to the no-citation rules is cases challenging other limits on what a litigant may present in court. [FN145] In the context of a criminal trial, the Supreme Court has held that "restrictions of a defendant's right to testify may not be arbitrary or disproportionate to the purposes they are designed to serve." [FN146] To think about the application of that doctrinal test to citation bans, consider the Court's opinion in Rock v. Arkansas, a due process challenge to the ban on hypnotically refreshed testimony. [FN147] Concluding that the testimonial ban was unconstitutional, the Court emphasized the ban's inflexibility. [FN148] The rule did "not allow a trial court to consider whether post-hypnosis testimony may be admissible in a particular case; it is a per se rule," the Court observed, predicated on the notion "that such testimony is always *587 unreliable." [FN149] Ultimately, the Court held that the state's "legitimate interest in barring unreliable evidence does not extend to per se exclusions that may be reliable in an individual case." [FN150] The Court's focus on the "wholesale inadmissability" created by the per se exclusion in Rock is suggestive in the context of the no-citation rules. After all, citation bans are a wholesale approach to evaluations an appellate court traditionally makes about the relevance, soundness, and usefulness of the reasoning in prior decisions. Just like the per se exclusion of all post-hypnotic recollections, the ban on citing any opinion labeled "nonprecedential" takes from the court the reasoned individual evaluation of most of its recent decisions. Courts do employ other per se limits on what litigants may argue, and the sources they may cite, apparently without offending due process. For instance, appellate advocates may not cite to evidence that is not in the trial court record. [FN151] But as the comparison with Rock shows, there are plausible arguments to be made that no-citation rules impose arbitrary or disproportionate limits on litigants' due process right to present their case. The rules ban every "nonprecedential" opinion, regardless how closely it duplicates the facts of the case at issue, how carefully its analysis is articulated, and how little other authority exists on the relevant issues. Moreover, citation bans can exclude information about the court's application of its precedents that has little to do with the reliability of individual case descriptions and is not available from other sources, for instance the court's degree of skepticism about evidentiary decisions or the frequency with which it uses arguably applicable precedents to impose liability in certain contexts. Finally, the ban on citing summary decisions excludes material whose quality courts are uniquely capable of judging on an individual basis. Evaluating the significance, reliability, correctness and applicability of a judicial decision made previously by their own court--or the court that reviews their decisions-- is what judges do day in and day out. The evidentiary ban struck down in Rock involved testimony for jurors, who might have deferred to the testifying experts. The no-citation rules substitute a per se ban for an individual calculation that is the absolute dead center of a judge's expertise. *588 3. Balancing Benefits and Burdens under Mathews v. Eldridge A three-part balancing test first articulated in an administrative context is worth considering here, because the Supreme Court has sometimes used it to evaluate process in civil litigation. Under Mathews v. Eldridge, a court balances: 1) "the private interest that will be affected by the official action;" 2) "the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards;" and 3) "the Government's interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail." [FN152] In a challenge to the no-citation rules, the first factor in the Mathews test--the private interest--will be variable, but generally significant. It will depend on what is at stake in the underlying lawsuit. In criminal or habeas proceedings a litigant may stand to lose his liberty, or even his life. [FN153] Evaluating the second Mathews factor--the risk of erroneous deprivation under the no-citation rules--raises the key question of how to define error in this context. There are two questions, really. First, is part of the definition of a "correct" outcome that the case is decided in accord with other like cases? Second, even if a correct outcome is judged only in terms of the just application of broad legal principles to the facts at hand, are judges more likely to reach the correct result if they have access to all their court's decisions in factually similar cases, even hastily written, cursory decisions not intended to create binding precedent? Taking the second question first, even if a correct result is defined solely as the application to the case at hand of the court's designated precedents, suppressing information about the court's decisionmaking in similar nonprecedential cases may increase erroneous outcomes. Particularly if a court has recently decided a series of cases with similar facts and a consistent approach and outcome, the presumption should be that those cases were correctly decided. [FN154] Knowledge of their results *589 would presumably help the court to make the correct decision in a subsequent case with very similar facts. Where there is a single, poorly articulated summary decision whose facts are difficult to decipher, attempts to conform subsequent decisions to the analysis in that case would likely confuse the issue, and might even increase the likelihood of error. But the fact that in some situations summary decisions might be unhelpful does not mean they never are. And, again, there is every reason to think judges are well equipped to tell the difference. Stepping back for a moment, this part of the Mathews test clarifies a contradiction in the no-citation courts' position. By declaring themselves free to diverge from their "nonprecedential" rulings in subsequent cases, courts have at least implicitly rejected case-by-case consistency as a primary test of the correctness of any given judicial ruling. Preventing judges from hearing about the results of "nonprecedential" cases makes it more likely that subsequent cases with similar fact patterns will be decided inconsistently with the uncitable ones. Even assuming all cases are decided consistent with the legal principles embodied in existing precedent, keeping judges in the dark about summary decisions applying those principles to various fact patterns will lead to greater inconsistencies among those applications. But at the same time the courts apparently remain concerned with maintaining consistent results in their summary decisions, otherwise they would not be worried about being confronted with contradictory reasoning if citation bans are lifted. It is worth turning again to the administrative law arena, this time for a contrasting approach. In some forms of administrative adjudication, intercase consistency is simply not a goal. Consider this statement from a tax court judge, hearing an appeal of an IRS ruling on a taxpayer's deductions: "It has long been the position of this Court that our responsibility is to apply the law to the facts of the case before us ...; how the Commissioner may have treated other taxpayers has generally been considered irrelevant in making that determination." [FN155] Where the aim is the fair administration of benefits in a massive bureaucratic system, consistency among like cases may not be an aspiration. Under the model of "bureaucratic justice" the goal is, rather, to avoid gross errors in individual results and to evenly distribute marginal errors, that is, to wind up with about as many erroneous grants of benefits as mistaken denials. [FN156] Thus, it is possible to have a *590 rational model of fair judicial decisionmaking that would view inter-case consistency as largely irrelevant to the correctness of individual rulings. But as Professor Jerry Mashaw has pointed out, although consistency may not be a goal of a bureaucratic system of mass justice, it is an important norm in traditional, individualized adjudication [FN157]--quintessentially practiced in Article III courts. One of the implications of courts' reliance on uncitable, nonbinding opinions may be that Article III judges should view themselves more as arbiters of bureaucratic rationality than as dispensers of individual justice. At least for the time being, however, most federal judges surely would insist that the rulings in their uncitable cases are correct not only in a bureaucratic sense of erring as much in the direction of mistaken benefit as deprivation. Courts still maintain that their summary rulings reach the correct result in the sense of meting out individual justice. Indeed, to embrace the bureaucratic rationality standard for uncitable "nonprecedential" cases while maintaining a standard of "individual justice" in cases that go by designated precedential opinions would raise other due process and equal protection problems. The Mathews analysis reveals the extent to which no-citation courts want to have their cake and eat it too. These courts want both to be free to ignore most of their recent decisions when deciding a new case with similar facts and to present that new ruling as consistent with like cases previously decided. Those goals appear to be in conflict. Continued reference in all like cases to a few designated precedential decisions in that area cannot produce the kind of case-by-case consistency achieved when each new ruling must be harmonized with all previous cases. When new decisions ignore many other routine cases with slightly different facts, it stands to reason that factual divergences will wind up being treated differently, resulting in inconsistency. The third Mathews factor also takes us back to the rationales courts give for their no-citation rules. Certainly the time and docket pressures federal appellate courts labor under are real, and most members of no-citation courts apparently believe citation bans are necessary to their ability to do their job properly. On the other hand, a number of other busy circuit courts get along without such bans, and courts' predictions of the results of allowing full citation are open to question. Empirical research on this point would be very useful. I know of no reports, so far, that courts that allow citation of summary decisions are having trouble keeping up with their *591 caseloads, declining to explain the results in most of their cases, [FN158] or suffering from a particular lack of clarity in their binding caselaw. In sum, arguing both from principles and from doctrine, one can make a plausible case that citation bans unconstitutionally burden a litigant's right to be heard in support of her case. Reasonable counterarguments exist. But, at a minimum, the due process analysis shows that the restrictions no-citation rules place on litigants' right to be heard evoke significant constitutional doubts. The arbitrariness many advocates feel inheres in courts' use of no-citation rules can be translated into constitutional terms. The due process analysis also begins to hint at some of the systemic effects of those rules for courts. III. THE IMPACT OF NO-CITATION RULES ON COURTS' PRECEDENTIAL PRACTICE I want to shift now from a constitutional analysis to a policy discussion, and from a focus on the problems no-citation rules create for litigants to the effects they have on courts' precedential practices. Hart v. Massanari, the leading case defending courts' nonbinding treatment of previous decisions and no-citation rules, exposes a link between the development of modern doctrines of precedent and the increased availability of judicial decisions when official case reporting began. [FN159] The opinion ignores the implications of that connection, however, when it analyzes the institutional role of no-citation rules. [FN160] Just as the increased availability of judicial opinions through universal reporting helped create the modern practice of following binding precedent, the decreased availability of prior judicial decisions due to no-citation rules is bound to produce changes in courts' precedential practices. Hart manages to avoid the contradiction between banning citation and following precedent only by redefining precedential practice. The opinion *592 adopts a concept of precedent that focuses entirely on maintaining consistent legal principles through written caselaw and ignores the precedential goal of deciding all cases consistently. Hart's model of precedent is a set of legal rules embodied in caselaw that is untouched by most of the decisions that apply those principles. This begins to look not so much like a precedential system but an awful lot like a system that applies a limited set of broad legal principles and rules in a wide range of cases. In other words, it looks less like traditional common-law practice and more like nonprecedential, civil-code-based adjudication. When most cases are uncitable, judges are never institutionally required to harmonize the reasoning and results of new cases with most of their court's decisions in similar cases. Circuits that forbid citation have effectively abandoned the kind of case-by-case internal consistency that is generally considered a defining norm of a common-law precedential system. Or, at the very least, they have abandoned any institutional accountability for such consistency. Such a system is not necessarily an unfair or undemocratic one. Once again, there are models in civil law countries and in American administrative adjudications. But it is troubling that federal appellate courts have removed institutional safeguards for consistency, while still ostensibly maintaining an institutional goal of consistent case outcomes. To recognize that this is the model of justice being followed by the federal appellate courts with no-citation rules is, therefore, not necessarily to indict it. But it is to see that this is no longer a traditional precedential system. A. The Connection between Availability of Information about Prior Cases and Courts' Adherence to Precedent In Hart v. Massanari, the Ninth Circuit faced an Article III challenge to its practice of making summary dispositions nonbinding in future cases. [FN161] In defense of that practice, the opinion sets out to prove that Article III's framers had no concept of binding precedent comparable to today's doctrine (and thus that Article III does not oblige federal courts to make all their cases precedential). In an extensive historical analysis, Hart establishes that the modern concept of precedent developed in concert with *593 the practice of standardized case reporting. [FN162] Moreover, Hart argues convincingly that a concept of binding precedent could not have developed in a regime in which most previous judicial decisions were unreported. [FN163] Hart traces the development of case reporting through Anglo-American law, explaining that for hundreds of years, "published opinions were relatively few." [FN164] When case reporting became more common, it was not a transparent record of judges' rulings. According to Hart, early case reporters were private entrepreneurs who did not faithfully record what judges said so much as express their own views of the cases' arguments and decisions. [FN165] "[C]ase reporters routinely suppressed or altered cases they considered wrongly decided." [FN166] The upshot was that at the time of the American constitution nothing like today's reliable official system of verbatim case reports existed. "Early American reporters resembled their English ancestors--disorganized and meager ...." [FN167] From this, Hart concludes that the framers of Article III had no concept of binding precedent comparable to the operative doctrine today. [FN168] The concept of binding precedent could and did develop only with "monumental improvements in the collection and reporting of case authorities." [FN169] As "a more comprehensive reporting system began to take hold, it became possible for judicial decisions to serve as binding authority." [FN170] In order to follow previous decisions, judges have to be able to find out what those decisions say. Universal case reporting makes that knowledge possible by enabling judges to find relevant cases through their own research and lawyers to draw the court's attention to previous cases they believe are relevant. As Hart explains, "The concept of binding precedent could only develop" after there was "a case reporting system that enabled later courts to know precisely what was said in earlier opinions." [FN171] Like spotty case reporting, no-citation rules prevent judges from learning about and following most of their courts' prior decisions. [FN172] The whole purpose of no-citation rules is to keep judges from having to consider *594 previous "nonprecedential" rulings when they decide new cases with similar facts. Some judges may choose to find and peruse relevant unpublished precursors to the cases before them. But no-citation rules make confronting previous rulings optional and increase the possibility that judges will rule without knowledge of their court's decisions in similar cases. The practice of designating most decisions "nonprecedential" removes the formal institutional obligation to adhere to those rulings in subsequent cases. No-citation rules go a significant step further and ensure that judges will not actually be called upon to consider their summary rulings in subsequent cases. B. How No-Citation Rules Reshape Precedential Practices Hart's analysis of historical concepts of precedent presents the practical availability of most case decisions as a prerequisite for the modern doctrine of strict judicial precedent. But when the opinion considers the current practice of making 84% of a court's decisions unavailable for citation in later cases, it ignores the effects of this practice on the doctrine of precedent. 1. No-Citation Rules Shift the Precedential Focus from Deciding All Cases Consistently to Articulating Consistent Legal Principles The opinion in Hart continually emphasizes courts' duty to develop "a coherent and internally consistent body of caselaw to serve as binding authority for themselves." [FN173] In contrast, the opinion never once mentions the duty to decide like cases alike. Of course, Hart does not explicitly reject the judicial duty to strive for consistent case outcomes. The opinion simply avoids the issue. This is possible in part because the opinion treats precedent as more or less self-executing, so long as legal rules are clear and consistent. "If a court must decide an issue governed by a prior opinion that constitutes binding authority, the later court is bound to reach the same result," explains Hart. [FN174] But manifestly, if a court is applying legal rules articulated in only 16% of its cases, there will be some factual variations in the other 84% that require interpretation of those rules. And if those interpretations are never reviewed, inconsistencies are free to develop. By making it less likely that judges will have the opportunity and incentive to know and to harmonize the reasoning and results of all similar cases, citation bans work against consistency of outcomes in all routine *595 cases. Maintaining consistency with the legal "rules" (as Hart characterizes them) [FN175] in a few designated precedents, without striving for internal consistency among the applications of those rules, is giving up on a defining feature of precedential practice. No-citation courts no longer rationalize applications of the rules with one another. In fact the Ninth Circuit's approach is recognizably closer to the kind of consistency practiced in a nonprecedential system of decisionmaking, like civil-law adjudication, which applies a limited set of rules to a wide range of factual circumstances without ever shifting the rules. In the Ninth Circuit, and other no-citation courts, judges never consider most of their recent decisions when they decide a new case. Instead, they decide case after case with applications of the same limited set of precedents, without ever modifying or even reenforcing the rulings of those precedential cases. In special circumstances, a generally uncitable case may be called to the court's attention. The Ninth Circuit rules allow such citation "in a request to publish a disposition ... or in a petition for panel rehearing or rehearing en banc, in order to demonstrate the existence of a conflict among opinions, dispositions, or orders." [FN176] Those few uncitable cases may then come to influence the court's precedent. But most routine applications of precedential rules will vanish without a trace. They may be consistent with a limited body of designated precedent, but not with other applications of those precedents. A special goal of a precedential judicial system is consistency among all those applications, achieved by case-by-case comparison. The ability to monitor, and thus to produce, that kind of consistency is just what the no- citation rules give up. To be sure, Ninth Circuit judges deciding the 84% of cases that are "nonprecedential" are still bound to follow the "rules" in the 16% of the court's decisions that are designated precedents. But that is like saying that because a court's decisions are all supposed to follow the same legal doctrines or the principles of natural law, the decisions must also be consistent with each other. The practice of following precedent, I take it, was developed in part because it was understood that even when judges made their best efforts to decide cases consistent with known principles and rules, the reasoning and results of individual case decisions could still diverge from one another on a detailed level. An article co-authored by a Sixth Circuit judge uses empirical examples to make this point. [FN177] The authors examine a series of (citable) *596 Sixth Circuit cases interpreting the meaning of the National Labor Relations Act's term "supervisors" in the context of the nursing profession. [FN178] They conclude that "the evolution of the law" depends on courts' willingness to publish multiple decisions on similar fact patterns and that if the later decisions in the Sixth Circuit series on nursing supervisors had not been citable precedents, "it is clear that some judges of the court would have reached results that, while arguably reconcilable with the initially stated general rule, were inconsistent with the law as developed in detail by the later published decisions." [FN179] Judges who work with no-citation rules no longer have the opportunity--or the chore, as Hart presents it--of harmonizing all their decisions in like cases. In Hart's scheme--which equates precedential practice with writing legal rules--that freedom results in greater precedential consistency, because judges are free to concentrate on articulating coherent legal principles. But "general statements of legal rules often are insufficient by themselves to ensure consistent results in future cases." [FN180] If precedential practice is mainly about deciding like cases alike, no-citation rules undermine it. The traditional notion of common law's development through precedents' application to a multifarious real world is alien to Hart's analysis. Hart presents precedent as fully developed in judges' original writing process. The application of precedents in other cases is not creative in any useful sense. There is no sense that litigants, lawyers, and judges in subsequent cases have any creative or interpretive role to play in the definition of precedent. [FN181] Use doesn't refine precedent; it either leaves it untouched or degrades it. As Hart describes it, a precedential opinion is at its zenith of authority and utility at the moment it leaves the pen of the judge who writes it. The original creation of that precedent "is a solemn judicial act that sets the course of the law for hundreds of thousands of litigants and potential litigants." [FN182] It's all downhill from there. Hart never stops to consider the gap between coherent written legal rules and consistent application of those rules. Instead, Hart's model of how future courts apply precedents depicts a kind of automatic *597 determinative precedential effect that bypasses any need for interpretation. "If a court must decide an issue governed by a prior opinion that constitutes binding authority, the later court is bound to reach the same result ...." [FN183] It's as though once a judge recognizes that a previous case is precedential, the old ruling decides the new case for him. "[B]inding authority is very powerful medicine," explains Hart. [FN184] The metaphor of a drug, which acts regardless of the will or desire of the person who uses it, perfectly expresses the sense that precedent once invoked is all determining. Once the judge takes her precedential "medicine," the outcome of the case before her is set. Under this model of automatic application, any interpretive use of precedent is bound to be either useless or positively harmful. Practice can't improve the prescription's effects, it can only corrupt them: "Writing a second, third, or tenth opinion in the same area of the law, based on materially indistinguishable facts will, at best, clutter up the law books and databases with redundant and thus unhelpful authority." [FN185] Variety is nothing but trouble. Publishing opinions in cases with closely similar fact patterns "will multiply significantly the number of inadvertent and unnecessary conflicts" in courts' precedential caselaw, because judges will "use slightly different language to express the same idea" and those verbal differences may "later take on a substantive significance." [FN186] Rather than offering multiple perspectives or illuminating different aspects of complex problems, in Hart' s model differences only open up the possibility of confusion or deliberate manipulation of the meaning intended by the rule's originators. Open citation leads to incoherence--a clear precedential pool made "muddy" with "a needless torrent of published opinions." [FN187] 2. No-Citation Rules Split Legal Interpretation from Case Adjudication The work of common-law appellate judges is sometimes usefully conceived as a dual task of deciding the cases brought to the court and creating legal rules. [FN188] In a no-citation regime, this conceptual duality is reified. According to Hart, courts approach lawmaking and error correction as entirely separate tasks: lawmaking takes place in a second step after the "disposition" of the case at hand. [FN189] In disposing of uncitable, unpublished *598 cases, judges ostensibly never generate legal rules that reach beyond the case at hand. While in precedential cases, "[t]he rule of decision cannot simply be announced, it must be selected after due consideration of the relevant legal and policy considerations." [FN190] One problem with this framework is that, in practice, it can be very hard to separate legal interpretation from error correction. [FN191] True, there have been proposals in the past to formally divide these tasks into the work of separate tribunals. [FN192] But such schemes have never gained very serious institutional traction, probably because these two aspects of a court's work are practically entwined and sometimes not wholly separable even in an ideal sense. Eventually, most thoughtful observers of appellate practice are drawn to ask whether a practical distinction between error correction and lawmaking is possible in most cases. [FN193] The justification for rules that make some appellate decisions non-precedential and uncitable depends on judges being able to tell the difference between error correcting and lawmaking. In practice, however, that distinction is ambiguous. In many cases, deciding whether the trial judge erred depends on a judgment about the application of broad legal principles to a particular factual situation. [FN194] Uncitable summary decisions are not limited to affirmances these days. [FN195] An example from one of these uncitable reversals shows the difficulty of separating legal interpretation from error correction. In an appeal by a habeas petitioner, Donny Picazo, the Ninth Circuit panel disagreed with a district judge's holding (affirming a magistrate judge's conclusion) that the wrongful admission of expert evidence in Picazo's trial was harmless error. [FN196] The appellate court found the evidence was prejudicial, reversed the judgment below, and remanded the case for the district judge to grant the writ. [FN197] The panel agreed with the district and magistrate judges that testimony by the state's police expert that Picazo intended to kill the person he shot at for the benefit of his gang was *599 a violation of Picazo's due process rights because "[a] witness is not permitted to give a direct opinion about the defendant's guilt or innocence." [FN198] But the appellate judges disagreed with the district and magistrate judges' conclusion that the unconstitutionally admitted testimony was harmless. [FN199] Applying the standard for harmless error articulated by the Supreme Court in Brecht v. Abrahamson, [FN200] the appellate judges held that "[g]iven the evidence in the record, it is likely that [the expert's] testimony had a substantial and injurious effect on the jury's verdict." [FN201] The district judge who had denied the writ below plainly disagreed about Brecht's application to the facts of Picazo's conviction. Was her assessment error or legal interpretation? In either case, does preventing a subsequent district court judge from hearing about Picazo's successful appeal help keep the applicable harmless error standard clear? Apart from their practical connection, legal interpretation and individual case adjudication may need to be linked to warrant appointed federal judges' lawmaking in a democracy. A federal court's lawmaking power is conceived as flowing from the court's function of deciding "cases or controversies." [FN202] Law made by judges occurs as a kind of secondary effect of resolving legal conflicts. [FN203] Unlike legislators, judges are not supposed to begin with an idea of how they want to change the law and then work to make that happen. The idea is that judges are responsive to the cases brought to them, and in order to decide those cases, they make law. If judges can themselves decide when to use their lawmaking power, it begins to look quite disconnected from the business of deciding cases and the need to reach consistent results, and more like the sort of freewheeling discretion legislators enjoy to make law on the questions they believe are important. [FN204] *600 3. No-Citation Rules Eliminate the Reciprocal Nature of Precedent Even assuming it were possible, and constitutionally permissible, to separate cases in which lawmaking is necessary from cases in which the decision is purely a matter of applying existing law, such a separation seems fundamentally at odds with the reciprocal nature of a precedential system. Traditionally, precedent is a two-way street. Once a case has been decided, its ruling helps shape subsequent decisions in cases with similar facts, and in that subsequent use the meaning of the original ruling also develops. A precedent expands to cover new scenarios, is narrowed to exclude others and is rationalized and harmonized with other legal rules and principles with which it intersects in later cases. In a no-citation regime, however, the court can use its earlier precedents to affect the outcome of numerous subsequent cases without affecting those precedents. This is another way that no-citation rules tend to shift adjudication toward a civil-code system. In an important sense, precedents in a no-citation regime are no longer shaped by practice because most of the opinions in which they are used are not themselves precedential and cannot even be discussed in developing subsequent precedents. So, once written, a precedent rarely develops. Like statutes applied by civil law courts, the designated precedents of no-citation courts encode a kind of original meaning that is unaltered by their later application. Judges considering still later cases look back past the intermediate (uncitable) applications of precedents to divine and apply the original intent of the old precedents, much as they would look to interpret the words of a statute. In this system, ideally, designated precedents maintain a stable meaning. In contrast, in a fully precedential system judicial opinions circulate continuously. Like the objects in a traditional gift culture, precedents pass through formally determined networks and down through generations and acquire additional meaning with each new use. Like circulating Kula shells, the cited precedents "represent the histories of their movements through time and the histories of the individuals who fashioned them, owned them, and exchanged them." [FN205] Subsequent citations do not simply refer to the words and meaning of the original decision. An old precedent acquires additional weight and significance through its use in later cases. Its later uses confirm its quality and add layers of new significance, enriching its value. In effect, by lending its reasoning to the later cases, the old precedent acquires the additional power of those rulings. When those *601 later uses occur in opinions that are themselves uncitable, however, the exchange of power and meaning is never completed. The later cases are never able to give back their power to the old precedents because they are uncitable and thus immediately taken out of circulation. In this system, old precedents are strangely stable. They do not acquire the enhanced--or, sometimes, degraded--meaning and value associated with their continuous circulation in other precedential opinions. [FN206] In a thoroughly precedential system, opinions remain to some extent works in progress, long after they are first published. In a no-citation system, however, the legal rules are fully formed at birth. Once a rule is created, it is available for use but will never absorb the effects of most of its applications. This leaves the courts' precedents oddly unworn, unchanged by the different contexts of their use, and untouched by the facts of the lives they are used to change, like so many jurisprudential Dorian Greys. C. Allowing Citation of Nonbinding Opinions for a More Flexible Precedential Practice Courts that ban the citation of most of their decisions are affecting their precedential practice in important ways. Citation bans tend to shift the focus away from case-by-case consistency to articulating consistent legal rules. The bans disrupt or ignore the close relationship of judicial lawmaking and individual case decisions. And they do away with the dynamic reciprocal nature of precedential caselaw. Even if courts want to reverse those trends, however, treating all summary decisions as binding precedents may be too much to ask in a time of such heavy caseloads. Perhaps partly to restore a more integrated precedential practice, most federal appellate courts have moved to a middle position, allowing their summary opinions to be cited for persuasive or even precedential value, while explaining that those cases are still not binding precedents and are given less weight than full-fledged precedential opinions. These rules are less clear-cut than citation bans. They treat precedent as partly contextually determined, and acknowledge that practice, litigants, lawyers, and subsequent judges all have roles to play in determining the weight a previous judicial decision carries in any given subsequent case. Consider, for example, the Sixth Circuit's rule, which recognizes that precedential value cannot be entirely preordained by the issuing judges: *602 Citation of unpublished decisions in briefs and oral arguments in this Court and in the district courts within this Circuit is disfavored .... If a party believes, nevertheless, that an unpublished disposition has precedential value in relation to a material issue in a case, and that there is no published opinion that would serve as well, such decision may be cited .... [FN207] The District of Columbia Circuit has a rule that similarly recognizes that a decision's precedential power is not entirely within the control of the judges who wrote and signed it: While unpublished orders and judgments may be cited to the court in accordance with Circuit Rule 28(c)(1)(B), a panel's decision to issue an unpublished disposition means that the panel sees no precedential value in that disposition. [FN208] Such rules acknowledge the conventional power of judges' words when they are part of official rulings, a potential that reaches beyond the individual views of the judges themselves, but also treat that conventional power as varying depending on its original context and the context in which it is invoked. Once again, comparison with administrative law provides a parallel. The two-tiered scheme that retains. Skidmore deference even after Chevron is roughly analogous to the treatment of binding precedents and nonbinding summary decisions in courts that allow citation of summary opinions. The Supreme Court has explained that under Chevron, an agency decision that exercised delegated Congressional lawmaking authority is "binding in the courts unless procedurally defective, arbitrary or capricious in substance, or manifestly contrary to the statute." [FN209] But the fact that only some agency decisions are "intended to have the force of law" does not mean that all the other decisions merit no deference at all. [FN210] A nonbinding agency ruling deserves "respect proportional to its 'power to persuade."' [FN211] Thus, an agency ruling that does not purport to make law "may surely claim the merit of its writer's thoroughness, logic and expertness, its fit with prior interpretations, and any other sources of weight." [FN212] *603 Putting Chevron and Skidmore together, the Court has thus defined a scheme for reviewing administrative decisions that looks much like the approach of the federal appeals courts that now allow citation of their own summary decisions as nonbinding but potentially precedential or persuasive authority. Choosing to allow varying levels of deference, rather than requiring a strict on-or-off approach to the binding power of agency rulings, expresses the view that a decision's value as a guide to future action and legal interpretation does not begin and end with the consequential status its author accords to it. The importance of a decision's reasoning for future choices is affected both by the procedures through which it came about in the first place and by the future context in which it is invoked. In contrast, categorical citation bans make precedential power all or nothing. In a no-citation court, as explained in Hart, a precedential judicial decision exerts a power over future court rulings comparable to the automatic, indiscriminate effects of "powerful medicine," that can only be controlled with complete suppression. [FN213] In this view, a judicial precedent's power is entirely separate from the ordinary persuasive power of words and the general requirement of consistency in government decisions, entirely determined by the author of the opinion, and largely impervious to the effects of its application. Courts that persist in forbidding citation of most of their case decisions are thus bound to become more and more focused on crafting written legal rules and less and less aware of the way those legal rules are applied in hundreds of subsequent cases. Eventually these courts will have legal rules and principles encoded in a set of limited authority that the court applies in rulings which are not themselves legally binding. In effect, they will have ceased to operate as precedential systems. IV. PRECEDENT, NO-CITATION RULES, AND COURTS' POWER TO AFFECT SOCIETY Can't put the rain back in the sky Once it falls down Please don't cry Lucinda Williams [FN214] Looking at the effects of no-citation rules on courts' precedential practices sets me to thinking more broadly about the meaning of precedent and how it might help shape courts' relationships with the rest of society. *604 A. Precedent and Judges' Transformative Verbal Power Most commentary about the role of precedent in adjudication has focused on whether and how precedent limits judges' discretionary power. But it seems undeniable that in some way precedent amplifies judicial power. There is, after all, something extraordinarily effective about an utterance that can bind future utterances, one's own and those of other judges. I also wonder if the extraordinary attention Americans--both lawyers and nonlawyers--pay to the words of judges is related to our courts' precedential practice. Puzzling about how the fact that judges' words have precedential consequences might be related to their social impact, I thought of another area in my life where questions about long-term consequences have recently surfaced. 1. The Undoing Game My three-year-old daughter and I have an ongoing discussion about whether it is possible to undo things. It began one evening when her father washed her face when she wanted me to do it. Oh, the horror! After a couple of minutes of hysterical sobbing, I went into the bathroom and started wiping her face with the washcloth. But she cried on. "I'm doing what you wanted," I snapped at her, "What's the problem? I'm washing your face." "No you're not," she wailed, "Daddy already washed it. Now it doesn't need washing. I want Daddy not to have washed it and you to wash it." I sat back. She was right that my gratuitous swiping at her already shining countenance was not really washing her face in any meaningful sense. Now what? I tried for a tone of calm reason: "Honey, I'm doing the best I can. I understand that you want me to have washed it in the first place, but it's too late for that. I can't unwash your face." Not good enough. Great gasping sobs. "No, no, no." "Okay," I backpedaled, "hold on. Maybe there is something I can do. Now, you understand I've never done this before, but I'm going to try to unwash your face." And with that I stood up, threw my arms into the air and started whirling around and waving the wet washcloth over the tub and over my (now discernibly quieter) daughter in what must have looked like a blend of ersatz Isadora Duncan and someone trying to fend off a swarm of gnats. "Abracadabra alacazam," I chanted, "wishy-washy, wishy-washy, piff, poof, I now pronounce you ... unwashed!" She bought it. That is, she appreciated the effort and decided it was good enough to let me off the hook. But she clearly understood that *605 unwashing wasn't real. She accepted it in the spirit in which it was offered: as a little maternal gift, as a game or a trick or a bit of magic conjured to help her make peace with the tragic reality that what's done is done. After that episode, my daughter and I started identifying all the things we did that we couldn't undo--and all the things we could. And we found that most physical actions are not reversible. Eating? No, you can't uneat. Painting? Can't unpaint. Tearing? Can't untear. Of course you can wipe or scrape away paint, and you can tape a torn page back together. But that kind of erasing or fixing is not at all the same thing as actually being able to put things back the way they were before you acted on them in the first place. What you have when you are done is something that has been refinished or put back together, not something that was never painted or torn. And this resistance to reversibility turns out to be the general rule. It turns out that there are very few things we do that are truly undoable. And so far as my daughter and I have been able to discover, these undoable actions are all associated with manmade things designed to be undone. Tying is reversible. Locking and unlocking. Making and unmaking a bed. Zipping and unzipping. But in this embodied life, most acts are not reversible. Can't unscramble those eggs, call the train back to the station. Can't unhit, unpee, unrun, unlaugh, unrain, ungrow, uncook. Daily, we live in a world in which we can only go forward from what has taken place; except in rare instances, we cannot undo the past. From this perspective, the idea of precedent seems less like a unique feature of judicial language and more like a given of everyday existence. Maybe the genius of the precedential concept is that in practice it is both. Treating judges' words as if they are irrevocable makes them both more like ordinary actions and less like ordinary words. 2. The Irrevocability of Precedent Shifts Judges' Language Toward Action The practice of precedent, it seems to me, brings to judicial language a kind of heightened awareness of this extremely widespread phenomenon of our embodied human existence in linear time. It's a kind of recognition of the general irreversibility of physical actions transposed to a legal world where the only action is verbal. Precedent gives to judicial language some of the undoability that my daughter and I have been investigating in physical actions. And in so doing it imparts to judges' words something of action's efficacy. Borrowing the irrevocability of ordinary physical action makes courts' pronouncements more like life, less like an abstract exercise, more real in some sense. At the same time, because most words clearly are not *606 irrevocable, precedent sets judicial language apart, makes it extraordinarily powerful., The combination tends to give to court judgments a feeling of heightened consequence. When you stop to consider the demands that are placed on judicial language to bring people around to a different view of their world, to transform the way people act toward one another, you can see how important it might be to preserve this quality of heightened consequentialness generally associated with action. As Alexander Hamilton said, judges "have neither Force nor Will, but merely judgment." [FN215] For courts, saying is doing. [FN216] In fact, judges, as judges, never do anything else. If courts are to be effective, we have to treat their words like actions in some respects. The doctrine of precedent makes that treatment seem more plausible because it commits judges to their words the way ordinary people are committed by their actions. It's a kind of liability that goes with the power we attribute to judicial language to actively intervene in the world as we know it. Of course we can think of instances when judicial language, precedential though it was, did not do the trick. We still needed the National Guard in Little Rock. But I wonder whether such examples would not be more plentiful if courts' words lacked the activating, consequential quality imparted to them by the irrevocability of precedent. In most cases, a judicial decision is enough, no show of force is needed. Yes, the threat of force is there, but so is the force of the words themselves. In fact, looking at this the other way round for a moment, you can see how the notion of precedent might arise from the ascription of transformative power to judges' words. If judges' decisions are speech-acts--if their words are like deeds that transform reality--then judges' words once pronounced should be undoable, in the same way that physical actions are. If courts' words are more than commands to do as the judges say on pain of military or police enforcement--if they have some of the transformative effect of actions--then perhaps they must also partake of actions' irrevocability. None of this means that we have to live forever with the effects of every judicial speech-act, anymore than we are forced to wallow in every glass of milk we spill. We can't go back and undo most of our actions, but *607 we can go on from where we are. We can't wish away the consequences, but we can set about fixing them, deliberately, imperfectly trying to undo the damage they have done. But not without some effort. We can choose to repudiate the old action and set a new course. But what we cannot do is make the world go back the way it was before. A three-year-old knows this. 3. A Cross-Cultural Comparison: "Hard Words" If it seems too farfetched to think that precedential judicial language borrows its extraordinary effectiveness from physical action by first borrowing action's irrevocability, consider a parallel phenomenon in the traditional culture of the Trobriand Islands. Anthropologist Annette Weiner wrote of the unusual efficacy of the special category of speech Trobriand Islanders call "hard words," which "are weighty," like physical objects, and thus carry "the ability to penetrate the personal space of others" and to influence their thinking. [FN217] Such hard words have no more direct physical influence than judges' precedential decisions. Both are, after all, just words. But both hard words and precedents have acquired an extraordinary capacity to shape reality, a capacity generally reserved for physical phenomena. In the Trobriands, apparently, ordinary speech is indirect and nonconfrontational. Direct expression is generally avoided, and, when employed, is understood as a special category of expression: "speaking what one truly thinks about something is called saying 'hard words' (biga peula)." [FN218] The way Trobrianders understand the extraordinary weightiness of hard words seems quite like the extraordinary irrevocability we attribute to precedential language. Weiner explains that truthful "hard words" are speech that has acquired some of the weight of physical objects. "Like weighted arrows, words ... are projected toward another person's personal space." [FN219] Both hard words and precedential opinions obtain a heightened ability to reshape reality more typical of physical phenomena through their association with other physical attributes. Because they are heavy, like physical objects, hard words are seen as able to penetrate the boundaries of personal space that ordinarily protect individuals' autonomy. [FN220] Because *608 they are irrevocable, like physical actions, precedential opinions acquire by association action's reality-shaping power. Weighty, penetrating hard words can influence individuals and social situations in ways that ordinary words cannot. But how do words become hard like objects? As I understand the process, in the Trobriand Islands truthful words, spoken publicly, are thought to expose reality. [FN221] And exposure is a quality Trobrianders associate with the appearance of physical objects, which cannot hide their nature the way words may dissemble thoughts. (Weiner points out that in our post-industrial society, objects can be faked, forged, or counterfeited, but in traditional small-scale societies it is extremely difficult if not impossible to disguise an object's origins and qualities. [FN222]) Their authenticity makes truthful words like physical objects. Then comes the crucial leap. Because for Trobrianders truthful words are like objects in their genuineness, such words are treated like objects in other ways as well. Truthful statements are harder to ignore, more weighty, and more penetrating and influential than ordinary words. [FN223] I am proposing that a similar process operates in our society for precedential judicial language. Because judges' precedential words are like physical actions in their irrevocability, we tend to see them as more "active" in other ways as well, and thus more able to affect the world. As a result of their heightened association with embodied life, both Trobriand "hard words" and American judges' precedential words are understood to have a power to transform reality ordinarily characteristic of the physical realm of objects and actions. B. No-Citation Rules and Judicial Action The selectivity of precedent under no-citation rules works against the association of judicial language with physical action. We can't pick and *609 choose which actions will have lasting effects. I can't decree, "Now this swing of the axe or this swipe of the sponge will be particularly effective because I say so." I have to actually swing harder or clean with greater alacrity. When judges can selectively control the precedential effects of their decisions, not by changing the nature of those decisions themselves but by prohibiting their citation, it makes it harder to see those decisions as actions. A crucial element of risk has been lost. When we act, we aim for the results we want, we gauge the possible consequences, and then, we take our chances. [FN224] To the extent no-citation rules eliminate that risk, they weaken the analogy of judicial words to physical action and the sense of consequence that flows from that connection. The on-off no-citation approach severs the connection between judicial words and ordinary actions that I have argued is a source of judicial power to affect social reality. For when judges can preordain the binding power of their words, their decisions no longer risk the kind of consequences the rest of us face in our everyday activities. [FN225] In contrast, judges in a fully precedential system cannot know or control all the future ramifications of their many decisions. It may be that this lack of control makes their words worthy of the transformative effects we ascribe to them. Judges are forced to commit themselves moment by moment to lasting effects they cannot foresee or control, like ordinary people plunging ahead in the dark. Acceptance of that sort of active, ongoing risk is part of what entitles judicial words to transform ordinary people's lives. The irony is that by trying to control their words' effect on other judicial decisions, courts may be losing some of their verbal power to affect people's lives. Courts use no-citation rules to protect their binding precedents from the ambiguity, uncertainty, and nuance that would result if those precedents absorbed the effects of their subsequent use in thousands of routine cases. Citation bans undeniably stabilize precedents, and in this sense may make them more powerful as legal rules. But using citation bans to shore up precedential opinions' power to affect subsequent legal outcomes may unwittingly deplete precedents' power to affect the world at large. The ability of judges to turn their precedential power on and off at *610 will runs distinctly counter to the unforgiving, onrushing effectualness of physical action. The no-citation courts' selective approach to precedent thus works against the identification with physical action that I have suggested is part of what gives judicial words their extraordinary ability to change how we see our world. By conserving their verbal power, no-citation courts may be sacrificing some of that power's transformative social effect. [FNa1]. Associate Counsel, Brennan Center for Justice at New York University School of Law; Doctoral Candidate, Columbia University School of Law; J.D. 1996, Brooklyn School of Law. Thanks to Brennan Center colleagues who provided insightful suggestions on approaches to the subject early in the process, and especially to Deborah Goldberg for nurturing the writing process. My heartfelt appreciation to Peter Strauss, Ursula Bentele, Sasha Samberg-Champion, Nathan Newman, Gillian Metzger, Stephen Barnett and Lawrence Solan who read drafts and offered much excellent criticism and advice, and to Doug Schulkind and Lincoln Alkind for their support and forbearance. [FN1]. PETER M. TIERSMA, LEGAL LANGUAGE 1 (1999). [FN2]. Oliver Wendell Holmes, The Path of the Law, 10 HARV. L. REV. 457, 457 (1897). [FN3]. They are the Second, Seventh, Ninth, and Federal Circuits. See 2D CIR. R. 0.23 (stating that the court's summary order statements "shall not be cited or otherwise used in unrelated cases before this or any other court"); 7TH CIR. R. 53(b)(2)(iv) (providing that the court's unpublished orders "shall not be cited or used as precedent"); 9TH CIR. R. 36-3(b) (stating that the court's "[u]npublished dispositions and orders ... may not be cited to or by the courts of this circuit"); Fed. Cir. R. App.
9.9 ("Nonprecedential opinions ... shall not be employed as precedent by this court, nor be cited as precedent by counsel ...."). In addition, some twenty-five state appellate courts likewise forbid citation of their unpublished opinions. Stephen R. Barnett, No-Citation Rules Under Siege: A Battlefield Report and Analysis, 5 J. App. Prac. & Process 473, 485 (2003). [FN4]. STATISTICS DIV., ADMIN. OFFICE OF THE U.S. COURTS, 2003 ANNUAL REPORT OF THE DIRECTOR: JUDICIAL BUSINESS OF THE UNITED STATES COURTS 36 tbl.S-3 (2003) [hereinafter 2003 ANNUAL REPORT], available at http:// www.uscourts.gov/judbus2003/tables/s3.pdf. [FN5]. Id. [FN6]. Id. The Federal Circuit does not provide data on its disposition of cases appealed via citable (published) and uncitable (unpublished) decisions. See id. (stating that the data regarding published verses unpublished decisions for the U.S. Court of Appeals for the Federal Circuit was not included in the report). [FN7]. Proposed Federal Rule of Appellate Procedure 32.1 would provide: No prohibition or restriction may be imposed upon the citation of judicial opinions, orders, judgments, or other written dispositions that have been designated as "unpublished," "not for publication," "non-precedential," "not precedent," or the like, unless that prohibition or restriction is generally imposed upon the citation of all judicial opinions, orders, judgments, or other written dispositions. Memorandum from Judge Samuel A. Alito, Jr., Chair, Advisory Committee on Appellate Rules, to Judge Anthony J. Scirica, Chair, Standing Committee on Rules of Practice and Procedure 28-29 (May 22, 2003) [hereinafter Alito Memorandum], available at http://www.uscourts.gov/rules/app0803.pdf. The rule was originally proposed to the U.S. Judicial Conference, Advisory Committee on Appellate Rules by the U.S. Department of Justice. JUDICIAL CONFERENCE OF THE UNITED STATES, MINUTES OF THE SPRING 2001 MEETING OF THE ADVISORY COMMITTEE ON APPELLATE RULES 47 (April 11, 2001). [FN8]. Even the original proponents of the rule, the U.S. Department of Justice, have tempered their support. At a meeting of the Appellate Rules Committee, the DOJ representative asserted that DOJ continues to favor the rule. He reported, however, that "the Solicitor General received a phone call from Judge Alex Kozinski of the Ninth Circuit and other opponents of the rule, and he is troubled by some of the concerns that they raised. The Solicitor General believes it essential that this Committee fully consult with the Ninth Circuit regarding its concerns." JUDICIAL CONFERENCE OF THE UNITED STATES, MINUTES OF THE FALL 2002 MEETING OF THE ADVISORY COMMITTEE ON APPELLATE RULES 26 (November 18, 2002). Proposed Rule 32.1 was put out for public comment from August 2003 to February 2004. Judicial Conference of the United States, Report of the Judicial Conference Committee on Rules of Practice and Procedure 2 (Sept. 2004). Nearly 500 comments were received, the great majority opposed to the rule. Id. These included near universal condemnation of the rule from judges of the no-citation courts. [FN9]. See Hart v. Massanari, 266 F.3d 1155, 1177 (9th Cir. 2001) ("It goes without saying that few, if any, appellate courts have the resources to write precedential opinions in every case that comes before them."); Alex Kozinski & Stephen Reinhardt, Please Don't Cite This! Why We Don't Allow Citation to Unpublished Dispositions, CAL. LAW., June 2000, at 44 ("[I]t would be impossible to do this without neglecting our other responsibilities."). [FN10]. See Kozinski & Reinhardt, supra note 9, at 44 (explaining that 40% of the Ninth Circuit's uncitable opinions were drafted by staff and presented to a panel of three judges who review 100-150 cases at a time). Judge Kozinski explained in his comments on Proposed Rule 32.1 that judges devote only five or ten minutes to each such case and that the remaining uncitable dispositions produced in chambers from law clerks' drafts are frequently not checked by judges "for latent ambiguities or misinterpretations." Letter Comments from Judge Alex Kozinski, U.S. Court of Appeals for the Ninth Circuit, to Judge Samuel A. Alito, Jr., Chairman, Advisory Committee on Appellate Rules 5 (Jan. 16, 2004) [hereinafter Kozinski Comments]. But see the comments of Judge Edward Becker of the Third Circuit, who testified that, in that court, summary opinions are "prepared in chambers under the close supervision of the judge," and though "usually drafted by clerks" they are "carefully reviewed and edited by the Judges." Judge Edward Becker, Statement to Advisory Committee on Appellate Rules 2 (April 13, 2004). [FN11]. See Hart, 266 F.3d at 1178 ("Deciding a large portion of our cases in this fashion frees us to spend the requisite time drafting precedential opinions in the remaining cases."); Letter Comments from Chief Judge Haldane Robert Mayer, on behalf of all the judges of the U.S. Court of Appeals for the Federal Cir., to Peter G. McCabe, Secretary, Committee on Rules of Practice and Procedure 1 (Jan. 6, 2004) ("The adoption of the practice allows the judges to concentrate their efforts on opinion writing in cases involving important and precedent-setting issues."). [FN12]. The court in Hart insisted: "That a case is decided without a precedential opinion does not mean it is not fully considered, or that the disposition does not reflect a reasoned analysis of the issues presented." Hart, 266 F.3d at 1177. But the court explained that in unpublished dispositions, [l]anguage adequate to inform the parties how their case has been decided might well be inadequate if applied to future cases arising from different facts. And, although three judges might agree on the outcome of the case before them, they might not agree on the precise reasoning or the rule to be applied to future cases. Id. at 1178. [FN13]. See infra notes 57-59. [FN14]. See supra note 3. [FN15]. Hart, 266 F.3d 1155, 1179. [FN16]. Id.; Letter Comments from Chief Judge John M. Walker, Jr., on behalf of himself and eighteen other judges of the U.S. Court of Appeals for the Second Circuit, to Peter G. McCabe, Secretary, Committee on Rules of Practice and Procedure 7 (Feb. 11, 2004) [hereinafter Second Circuit Comments]. [FN17]. See Patricia M. Wald, The Rhetoric of Results and the Results of Rhetoric: Judicial Writing, 62 U. Chi. L. Rev. 1371, 1374 (1995). This article by a member of the D.C. Circuit asserts that unpublished opinions there, which were then covered by a no-citation rule, sometimes "deal with issues not clearly controlled by prior precedent" and notes that a study applying the court's own written criteria "questioned the decision not to publish in 40 percent of the cases." Id. at 1374; see also Lauren K. Robel, The Myth of the Disposable Opinion: Unpublished Opinions and Government Litigants in the United States Courts of Appeals, 87 U. Mich. L. Rev. 940, 950-52 (1989) (describing useful but uncitable Ninth Circuit opinions regarding pleading requirements for RICO violations); Pamela Foa, Comment, A Snake in the Path of the Law: the Seventh Circuit's Non-Publication Rule, 39 U. Pitt. L. Rev. 309, 315 (1977) (describing how the Seventh Circuit failed to accurately distinguish between trivial and important cases). [FN18]. The Ninth Circuit is the clear leader in producing uncitable reversals and dissents. A Westlaw search for all Ninth Circuit decisions issued between January 1, 2004 and March 1, 2004 that contain the words "this case was not selected for publication" found that in the first two months of 2004 that court issued 15 uncitable dissents and 72 reversals. See also Michael Hannon, A Closer Look at Unpublished Opinions in the United States Courts of Appeals, 3 J. APP. PRAC. & PROCESS 199, 216 (2001) (finding that in 2000, about one in five unpublished opinions was "non-affirming"); Deborah Jones Merritt & James J. Brudney, Stalking Secret Law: What Predicts Publication in the United States Courts of Appeals, 54 VAND. L. REV. 71, 113 (2001) (finding in a study of "unpublished' labor law opinions from all federal circuits from 1986-1993, most of which were then covered by no-citation rules, an overall reversal rate of 7% and a 2% rate of concurrence or dissent) [FN19]. 9TH CIR. R. 36-2(a). [FN20]. See David Greenwald & Frederick A.O. Schwarz, Jr., The Censorial Judiciary, 35 U.C. Davis L. Rev. 1133, 1141-42 (2002) (stating that "[i]n 1971 ... [the] Federal Judicial Center ... [noted a] 'widespread consensus that too many opinions are being printed or published or otherwise disseminated,"' and in 1972 recommended that each circuit reduce the number of published decisions and restrict the citation of unpublished opinions.); Salem M. Katsh & Alex V. Chachkes, Constitutionality of "No-Citation" Rules, 3 J. App. Prac. & Process 287, 291-92 (2001) (noting the 1971 report by the Federal Judicial Center and its 1972 recommendations). [FN21]. See RICHARD A. POSNER, THE FEDERAL COURTS: CHALLENGE AND REFORM 66- 79, 162-165 (1996) (discussing statistics on the growth of the federal appellate workload and the corresponding use of unpublished decisions); Greenwald & Schwarz, supra note 20, at 1145-47 (citing the disparate growth rates of the appellate judiciary and their caseload over the past four decades). [FN22]. In the eighteenth century, Lord Coke complained that case reporters were becoming "elephantini libri" and called for limited publication. See Greenwald & Schwarz, supra note 20, at 1144 (quoting 2 Coke's Rep. iii-iv (1777)); see also Symbol Tech., Inc. v. Lemelson Med., 277 F.3d 1361, 1367 (Fed. Cir. 2002) (noting Lord Coke's concern about "too much precedent"). [FN23]. ADMIN. OFFICE OF THE UNITED STATES COURTS, ANNUAL REPORT OF THE DIRECTOR (1964), in REPORTS OF THE PROCEEDINGS OF THE JUDICIAL CONFERENCE OF THE UNITED STATES 11 (1964). [FN24]. Greenwald & Schwarz, supra note 20, at 1142. [FN25]. Id. [FN26]. Merritt & Brudney, supra note 18, at 75-76. [FN27]. See Katsh & Chachkes, supra note 20, at 292-93 (citing testimony from the Commission on Revision of the Federal Court Appellate System in the 1970s). The Second Circuit's no-citation rule still avowedly rests on this rationale: "Since these statements do not constitute formal opinions of the court and are unreported or not uniformly available to all parties, they shall not be cited ...." 2D CIR. R. 0.23. [FN28]. See, e.g., Merritt & Brudney, supra note 18, at 76 ("By 1978-79, half of federal appellate dispositions were unpublished; the figure reached two-thirds by 1989."); Wald, supra note 17, at 1373 ("Due to the pressure of accelerating caseloads, the majority of federal cases now get this unpublished treatment."). [FN29]. See POSNER, supra note 21, at 130 ("No effort has been made to expand the number of judges in proportion to the increase in caseload."). [FN30]. See id. at 130-39 (providing a sophisticated discussion of the problems entailed in expansion). But see William M. Richman & William L. Reynolds, Elitism, Expediency, and the New Certiorari: Requiem for the Learned Hand Tradition, 81 CORNELL L. REV. 273, 299-300 (1996) (pointing out that the judiciary has vigorously opposed any proposed solution ["to the caseload glut"] that would expand the number of judges). [FN31]. According to Hart, "Congress would have to increase the number of judges by something like a factor of five to allocate to each judge a manageable number of opinions each year." Hart v. Massanari, 266 F.3d 1155, 1179 n.39 (9th Cir. 2001). [FN32]. See Richman & Reynolds, supra note 30 at 301-03 (discussing the merits of the prestige argument); cf. HENRY J. FRIENDLY, FEDERAL JURISDICTION: A GENERAL VIEW 44-46 (1973) (arguing against expanding courts of appeals because of the need to maintain collegiality and personal associations among judges); Felix Frankfurter, Distribution of Judicial Power Between United States and State Courts, 13 CORNELL L.Q. 499, 515 (1928) ("A powerful judiciary implies a relatively small number of judges."), quoted in FRIENDLY, supra, at 28. [FN33]. But see Marc Galanter The Turn Against Law: The Recoil Against Expanding Accountability, 81 TEX. L. REV. 285, 285 (2002) [hereinafter Galanter, The Turn Against Law] (arguing that "[a] century ago, Americans invoked the civil courts more frequently on a per capita basis than they do now"); Marc Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (and Think We Know) About Our Allegedly Contentious and Litigious Society, 31 U.C.L.A. L. Rev. 4, 5 (1983) [hereinafter Galanter, Reading the Landscape] (contending that "present [litigation] levels are not historically unprecedented"). [FN34]. Justice Macklin Fleming, Court Survival in the Litigation Explosion, 54 JUDICATURE 109, 109 (1970), quoted in Galanter, The Turn Against Law, supra note 33, at 292. [FN35]. William J. McGill, Litigation-Prone Society: Protection of Professional Life, 78 N.Y. St. J. Med. 658, 661 (1978), quoted in Galanter, Reading the Landscape, supra note 33, at 7. McGill was President of Columbia University at the time. McGill, supra, at 658. [FN36]. Galanter, Reading the Landscape, supra note 33, at 7-8 (quoting [Third Circuit Judge] Ruggero J. Aldisert, An American View of the Judicial Function, in LEGAL INSTITUTIONS TODAY: ENGLISH AND AMERICAN APPROACHES COMPARED 31, 56 (H. Jones ed., 1977)). [FN37]. The term "too much law" appeared in a 1977 Newsweek cover story referenced by Galanter, Reading the Landscape, supra note 33, at 6 n.4 (citing Jerrold K. Footlick, Too Much Law?, NEWSWEEK, Jan. 10, 1977, at 42. [FN38]. Marc Galanter has suggested that when the "too much law" critique was launched, total per capita court filings were actually fewer than they had been a century earlier. Galanter, Reading the Landscape, supra note 33, at 38. Taking a longer view, he points out that from about 1930 to 1960 Americans were litigating much less often than they had previously, and argues that the rising rates thereafter were more a return to historic levels than a unique "litigation explosion." Id. at 38 n.171. Whether or not that is the case, it is clear that rising rates of litigation in the 1960s came toward the end of a significant shift in the kinds of cases brought to court. As Galanter explains, the twentieth century saw a shift from civil to criminal cases and, within the civil docket, "there has been a shift from cases involving market transactions (contract, property, and debt collection) to family and tort cases." Id. at 42. [FN39]. Civil Rights Act of 1964, Pub. L. No. 88-352, 78 Stat. 241 (codified as amended in scattered sections of 28 U.S.C. and 42 U.S.C.). [FN40]. Voting Rights Act of 1965, 42 U.S.C.
1973 (2000). [FN41]. Age Discrimination in Employment Act of 1967, 29 U.S.C.
621-34 (2000). [FN42]. Civil Rights Act of 1968, 42 U.S.C.
3601-19 (2000). [FN43]. POSNER, supra note 21, at 104; see also FRIENDLY, supra note 32, at 19-20 (discussing how the Supreme Court decisions in James v. Alfred H. Mayer Co., 397 U.S. 409 (1968), and Griffin v. Breckenridge, 403 U.S. 88 (1971), would lead to increased civil rights litigation). [FN44]. Brown v. Bd. of Educ., 347 U.S. 483 (1954). [FN45]. Criminal Justice Act of 1964, 18 U.S.C.
3006A (2000); see FRIENDLY, supra note 32, at 31-32 (citing a more than 400% increase in the number of appeals from district court decisions between 1961 and 1972); POSNER, supra note 21, at 75-77 (attributing the rise of criminal appeals to the Criminal Justice Act of 1964). [FN46]. See FRIENDLY, supra note 32, at 23-24 (listing various statutes that enabled parties to appeal not only the procedures, but also the substantive parts of an agency's acts or decisions). [FN47]. Galanter, Reading the Landscape; supra note 33, at 43. [FN48]. Galanter, The Turn Against Law, supra note 33, at 287. [FN49]. The disproportionate use of the new summary procedures for particular kinds of cases may have been unintentional. It could be a secondary effect of the criteria the courts were using to decide which cases to make nonprecedential and uncitable. As a matter of fact, however, in at least some of the circuits, summary procedures were avowedly intended to cover certain kinds of cases. For instance, the Fourth Circuit has explained that its adoption of unpublished decisions was a response to rising levels of habeas claims. Jones v. Superintendent, Va. State Farm, 465 F.2d 1091, 1094-96 (4th Cir. 1972). The First Circuit's rule on "unpublished" opinions, whose citation was, until recently, forbidden and is now "disfavored," explains that "[m]ost opinions dealing with claims for benefits under the Social Security Act, 42 U.S.C.
250(g), will clearly fall within the exception." 1ST CIR. R. 36(b)(1). [FN50]. See David F. Pike, Why Everybody is Suing Everybody, U.S. NEWS & WORLD REP., Dec. 4, 1978, at 50 (describing "the impulse to litigate anything and everything" in America, and stating that "Americans in all walks of life are being buried under an avalanche of lawsuits"). [FN51]. See Galanter, The Turn Against Law, supra note 33, at 292 nn.42 & 44 (citing several uses of the "explosion" metaphor, including HARRY W. JONES, THE COURTS, THE PUBLIC, AND THE LAW EXPLOSION 2 (1965); Fleming, supra note 34, at 109; and Eleanore Carruth, The "Legal Explosion" Has Left Business Shell-Shocked, FORTUNE, Apr. 1973, at 65). [FN52]. FRIENDLY, supra note 32, at 4. [FN53]. It is hard to say whether there is anything wrong with issuing avowedly nonprecedential, uncitable opinions more often in certain types of cases than in others. On the one hand, such categorical differences create at least the appearance of a kind of second-class adjudication for poor people appealing the denial of Social Security benefits and prisoners challenging their convictions or the conditions of their incarceration. On the other, there might be strong rational reasons why these sorts of cases are more appropriately adjudicated without precedential status. Habeas cases are likely to be uncounseled, so that the appellate court has not had the assistance of adversarial lawyering in analyzing the legal issues. In direct appeals from federal criminal convictions, lawyers are provided free of charge, so these cases are not subject to the ordinary market pressures that would generally discourage bringing appeals that are highly unlikely to succeed. Then again, most habeas petitions and criminal appeals are being litigated for the highest personal stakes imaginable: liberty or life, which would seem to mandate using the most thorough analytic processes in their adjudication. [FN54]. See, e.g., Pike, supra note 50, at 50 (arguing that the rise in lawsuits results in "social and economic repercussions rippling through society, touching virtually every aspect of American life in one way or another, often for the worse"). [FN55]. The reported percentages of unpublished opinions for 2003 were: 9th Cir., 84%; 7th Cir., 57%; 2d Cir., 75%. 2003 ANNUAL REPORT, supra note 4, at 36 tbl. S-3. Data for the Federal Circuit was not reported. Id. [FN56]. See id. (reporting that across the country 80% of decisions rendered by the federal appellate courts, not including the Federal Circuit, were done by "unpublished" dispositions, which most of those courts assert are "nonprecedential," though they allow their citation). [FN57]. 3D CIR. R. APP. 5.3; 5TH CIR. R. 47.5.4; 8TH CIR. R. 28A. [FN58]. 1ST CIR. R. 32.3(a)(2); 11TH CIR. R. 36-2. The D.C. Circuit does not assert an objective precedential or nonprecedential status for its unpublished decisions but explains that "a panel's decision to issue an unpublished disposition means that the panel sees no precedential value in that disposition." D.C. CIR. R. 36(c)(2). [FN59]. 4TH CIR. R. 36(c); 6TH CIR. R. 28(g); 10TH CIR. R. 36.3(B). [FN60]. Of course, as public documents, even now the unpublished opinions can be obtained by request from the court clerk's office. But since no one knows to look for them, and their contents are not indexed for researchers' use, the lack of official publication takes them out of circulation for all but the expert advocates who may keep private databases of them. After 2005, however, all circuits will be required to post their decisions on the web and the last vestiges of the actual practice of nonpublication will disappear. E-Government Act of 2002, Pub. L. No. 107-347,
205(a)(5), 116 Stat. 2899, 2913. [FN61]. See supra note 7 and accompanying text. [FN62]. See supra note 8 and accompanying text. [FN63]. See, e.g., Kozinski Comments, supra note 10, at 20-21 (stating that the citation bans "cannot conceivably be viewed as a First Amendment violation," and that many courts of appeals judges "believe that the noncitation rule is an important tool in the fair administration of justice"). [FN64]. See, e.g., Symbol Tech., Inc. v. Lemelson Med., 277 F.3d 1361, 1368 (Fed. Cir. 2002) (defending unpublished opinions as a safeguard against "the growing imprecision, uncertainty and unpredictability of the law"); Hart v. Massanari, 266 F.3d 1155, 1179 (9th Cir. 2001) (asserting that unpublished decisions prevent redundancy and clutter); see also Kozinski Comments, supra note 10, at 1 (stating that disallowing unpublished decisions would, inter alia, "make more difficult our job of keeping the law of the circuit clear and consistent"). [FN65]. Thomas v. Newton Int'l Enters., 42 F.3d 1266, 1272 (9th Cir. 1994) (quoting In re Burns, 974 F.2d 1064, 1068 (9th Cir. 1992)). [FN66]. Symbol Tech., Inc., 277 F.3d at 1368. [FN67]. Id. [FN68]. Letter Comments from Judge Ferdinand F. Fernandez, U.S. Court of Appeals for the Ninth Circuit, to Peter G. McCabe, Secretary, Committee on Rules of Practice and Procedure 1 (Dec. 4, 2003) [hereinafter Fernandez Comments]. [FN69]. Hart, 266 F.3d at 1179. [FN70]. Kozinski Comments, supra note 10, at 2-3. [FN71]. Letter from nine judges of the U.S. Court of Appeals for the Seventh Circuit, to Judge Samuel A. Alito, Jr., Chairman, Advisory Committee on Appellate Rules 1 (Feb. 11, 2004) [hereinafter Seventh Circuit Comments]. [FN72]. Fernandez Comments, supra note 68, at 1-2. [FN73]. Unpublished Judicial Opinions: Hearing Before the Subcomm, on Courts, the Internet, and Intellectual Property of the House Comm. on the Judiciary, 107th Cong. 13 (2002) (statement of Honorable Alex Kozinski, Judge, United States Court of Appeals for the Ninth Circuit) [hereinafter Kozinski Statement]. [FN74]. See Seventh Circuit Comments, supra note 71, at 1 (arguing that unpublished decisions will become precedent "because the judges of a court will be naturally reluctant to repudiate or ignore previous decisions"). [FN75]. Kozinski Comments, supra note 10, at 4. [FN76]. Seventh Circuit Comments, supra note 71, at 1. [FN77]. Hart v. Massanari, 266 F.3d 1155, 1170 (9th Cir. 2001). [FN78]. Jones v. Superintendent, Va. State Farm, 465 F.2d 1091, 1094 (4th Cir. 1972); see also Re Rules of the United States Court of Appeals for the Tenth Circuit, 955 F.2d 36, 37 (10th Cir. 1992) (Holloway, C.J., dissenting) ("[A]ll rulings of this court are precedents, like it or not, and we cannot consign any of them to oblivion by merely banning their citation."). [FN79]. Seventh Circuit Comments, supra note 71, at 1. [FN80]. Letter Comments from Judge Frank H. Easterbrook, U.S. Court of Appeals for the Seventh Circuit, to Peter G. McCabe, Secretary, Committee on Rules of Practice and Procedure 2 (Feb. 13, 2004). [FN81]. See Lauren Robel, The Practice of Precedent: Anastasoff, Noncitation Rules, and the Meaning of Precedent in an Interpretive Community, 35 IND. L. REV. 399, 408-09 (2002) ("[W]ith few exceptions, unpublished opinions are generally available in the searchable electronic databases [,] ... on court websites and in specialty reporters."). [FN82]. The numbers were as follows: 2d Cir., 49%; 7th Cir., 43%; 9th Cir., 47%; and Federal Circuit, 34%. 1998 FJC SURVEY OF APPELLATE COUNSEL, in COMMISSION ON STRUCTURAL ALTERNATIVES FOR THE FEDEREAL COURTS OF APPEALS, WORKING PAPERS 78 (1998). [FN83]. See Kozinski Statement, supra note 73, at 13 ("[Courts might] reduce our unpublished dispositions to one-word judgment orders, as have other circuits."). [FN84]. As Michael C. Dorf has observed, "For the judiciary, giving reasons justifies the exercise of governmental authority, much as elections justify its exercise by the political branches." Michael C. Dorf, Dicta and Article III, 142 U. Pa. L. Rev. 1997, 2029 (1994). [FN85]. Planned Parenthood v. Casey, 505 U.S. 833, 865 (1992). [FN86]. See supra note 18. [FN87]. Kozinski Statement, supra note 73, at 63. [FN88]. See Fuentes v. Shevin, 407 U.S. 67, 80 (1972) (characterizing the right to be heard as the "central meaning of procedural due process"); Goldberg v. Kelly, 397 U.S. 254, 267 (1970) ("'The fundamental requisite of due process of law is the opportunity to be heard."' (quoting Grannis v. Ordean, 234 U.S. 385, 394 (1914))); Joint Anti-Fascist Refugee Comm. v. McGrath, 341 U.S. 123, 161-62 (1951) (stating that one of the fundamental principles of due process "'is that no person shall be deprived of his liberty without opportunity, at some time, to be heard ...."' (omission in original) (quoting Kaoru v. Yamataya, 189 U.S. 86, 101 (1903))). [FN89]. Fuentes, 407 U.S. at 80 ("The constitutional right to be heard is a basic aspect of the duty of government to follow a fair process of decisionmaking ...."). [FN90]. Joint Anti-Fascist Refugee Comm., 341 U.S. at 162 (quoting Hagar v. Reclamation District, No. 108, 111 U.S. 701, 708 (1884)). [FN91]. Fuentes, 407 U.S. at 81. [FN92]. Henry J. Friendly, Some Kind of Hearing, 123 U. Pa. L. Rev. 1267, 1281 (1975), cited in Mathews v. Eldridge, 424 U.S. 319, 343 (1976). Judge Friendly was describing the elements of procedure necessary for a fair administrative hearing, a context in which the amount of process "due" should be significantly lower than that required in a federal appeals court. There is a long tradition of high levels of process in federal court. Indeed, the context for Judge Friendly's enumeration of hearing requirements was a discussion of the extent to which administrative hearings should provide the levels of process required in courts. Moreover, the right of appeal to federal court is one of the reasons initial administrative hearings need not provide such high levels of process. Also, the stakes are often higher in court, where, for instance, a criminal defendant may face incarceration or even death if convicted. [FN93]. See Cafeteria and Rest. Workers Union, Local 473 v. McElroy, 367 U.S. 886, 894 (1961) ("One may not have a constitutional right to go to Bagdad, but the Government may not prohibit one from going there unless by means consonant with due process of law." (quoting Homer v. Richmond, 292 F.2d 719, 722 (D.C. Cir. 1961))). [FN94]. See Kozinski Comments, supra note 10, at 4 ("In other words, by citing what a court has done on a previous occasion, a party is saying: This is what that court did in very similar circumstances, and therefore, under the doctrine of stare decisis, this court ought to do the same."). [FN95]. Frederick Schauer, Precedent, 39 STAN. L. REV. 571, 596 (1987) (footnotes omitted). [FN96]. For instance, discussing China's recent entry to the World Trade Organization, one author explains, "To meet the fundamental requirement for transparency under the WTO, ... Chinese courts and judges must consciously adopt the modern spirit of rule of law, embodied by public and fair trials, and consistency." Chris X. Lin, A Quiet Revolution: An Overview of China's Judicial Reform, 4 ASIAN-PAC. L. & POL'Y, J. 255, 291 (2003). [FN97]. BENJAMIN N. CARDOZO, THE NATURE OF THE JUDICIAL PROCESS 33 (1921). [FN98]. Henry J. Friendly, Indiscretion about Discretion, 31 EMORY L. J. 747, 758 (1982) (quoting Ward v. James [1966] 1 Q.B. 273, 294 (C.A.)). [FN99]. Friendly himself recognized this, explaining that "[w]ith the volume of cases in the federal system and the consequent necessity for twelve courts of appeals with an incalculable number of different panels, and with the limited availability of Supreme Court review, the goal of achieving complete consistency even within that system is unattainable." Id. at 758. "But," he wrote, "that is no reason for not doing what we can." Id. As Judge Easterbrook has pointed out, the fact that courts are represented by multiple decisionmakers who must compromise in order to form majorities makes complete consistency over time an institutional impossibility. Frank H. Easterbrook, Ways of Criticizing the Court, 95 HARV. L. REV. 802, 831 (1982). [FN100]. Named for the Supreme Court case that first articulated it, Skidmore v. Swift & Co., 323 U.S. 134 (1944), the principle is that a nonbinding agency decision will nevertheless command judicial deference depending on "the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control." Id. at 140. [FN101]. Cf. PETER L. STRAUSS ET AL., GELLHORN AND BYSE'S ADMINISTRATIVE LAW: CASES AND COMMENTS xxii (10th ed. 2003) (listing "Consistency of Application" and "Consistency of Rule Declaration" as the two subheadings in the section on "The Baseline Norm of Legal Regularity"). [FN102]. Skidmore, 323 U.S. at 140 (emphasis added). [FN103]. Chevron, U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837 (1984). [FN104]. See United States v. Mead Corp., 533 U.S. 218, 227-28 (2001). [FN105]. Chevron, 467 U.S. at 863-64; see also Good Samaritan Hosp. v. Shalala, 508 U.S. 402, 417 (1993) ("[T]he consistency of an agency's position is a factor in assessing the weight that position is due"), quoted in Mead, 533 U.S. at 228 n.8; Immigration & Naturalization Serv. v. Cardoza-Fonseca, 480 U.S. 421, 446 n.30 (1987) ("An agency interpretation of a relevant provision which conflicts with the agency's earlier interpretation is 'entitled to considerably less deference' than a consistently held agency view." (quoting Watt v. Alaska, 451 U.S. 259, 273 (1981))); see also Smiley v. Citibank, 517 U.S. 735, 742 (1996) (explaining that, while inconsistency does not necessarily invalidate an agency interpretation, a "[s]udden and unexplained" change in position may). [FN106]. See Mead, 533 U.S. at 228 ("The fair measure of deference to an agency administering its own statute has been understood to vary with circumstances and courts have looked to the degree of the agency's care, [and] its consistency ...." (footnote omitted) (referencing Skidmore, 323 U.S. at 140)). [FN107]. Smiley, 517 U.S. at 742 (second alteration in original) (citations omitted) (quoting 5 U.S.C.
706(2)(A)). [FN108]. Cardoza-Fonseca, 480 U.S. at 446 n.30. [FN109]. Id. (quoting Watt, 451 U.S. at 273). [FN110]. Id. Another example is the Court's analysis of the apparently inconsistent interpretations of a statute rendered by the National Labor Relations Board in a series of cases involving unions' right to picket. In National Labor Relations Bd. v. Local Union No. 103, Int'l Assoc. of Bridge, Structural & Ornamental Iron Workers, the union challenged the Board's construction of the relevant statute, saying it merely reflected the principle enunciated in one particular Board adjudication that was itself inconsistent with a previous Board decision. 434 U.S. 335, 350 (1978). The Court responded that "the Board has plainly not adhered to" the approach of that previous decision. Id. at 351. Instead, "[i]ts contrary view has been expressed on more than one occasion." Id. The Court's reasoning in Local Union No. 103 goes to show that while "[a]n administrative agency is not disqualified from changing its mind," consistency remains an important factor in determining the reasonableness of a challenged agency decision. Id. [FN111]. Atchison, Topeka & Santa Fe Ry. Co. v. Wichita Bd. of Trade, 412 U.S. 800, 807 (1973). [FN112]. Id. at 808. [FN113]. Shaw's Supermarkets, Inc. v. Nat'l Labor Relations Bd., 884 F.2d 34, 41 (1st Cir. 1989). [FN114]. Hart v. Massanari, 266 F.3d 1155, 1170 (9th Cir. 2001). [FN115]. Id. [FN116]. Wolff v. McDonnell, 418 U.S. 539, 558 (1974) (citing Dent v. West Virginia, 129 U.S. 114, 123 (1889)). Reliance is one of the policy reasons for this concern with administrative consistency. Concern about reliance, however, seems largely inapplicable when courts barring citation of their unpublished opinions clearly announce that those opinions are not precedential and not to be treated as binding in apparently similar future cases. There may be both constitutional and pragmatic problems with such disclaimers, but, assuming for the moment that it is proper and reasonable for the courts to label some of their decisions nonprecedential, potential litigants have been warned not to rely on uncitable opinions as guides to conduct. [FN117]. Smiley v. Citibank, 517 U.S. 735, 743 (1996). [FN118]. Id. [FN119]. Hart, 266 F.3d at 1178. [FN120]. Id. at 1177. [FN121]. In support of his court's no-citation rule, Judge Kozinski of the Ninth Circuit has stressed the extent to which the summary dispositions it covers are the work of staff attorneys and law clerks. They only "appear to have been written ... by three circuit judges," he says. Kozinski Comments, supra note 10, at 2. Moreover, because these decisions get no meaningful en banc review, they cannot be truthfully represented as the view of the full court. Id. at 7. Again, though, it cannot be that the results of these cases are not judgments of the court. Otherwise, they have not adjudicated the case. And while their unpublished, uncitable status certainly makes en banc or certiorari review less likely, these judgments are from time to time appealed and reversed. See MICHAEL E. TIGAR & JANE B. TIGAR, FEDERAL APPEALS JURISDICTION AND PRACTICE
10.13 (3d ed. 1999) (indicating a decreased likelihood of review). [FN122]. See Hart, 266 F.3d at 1178 ("Language adequate to inform the parties how their case has been decided might well be inadequate if applied to future cases arising from different facts. And, although three judges might agree on the outcome of the case before them, they might not agree on the precise reasoning."); see also Kozinski & Reinhardt, supra note 9, at 44 ("[T]he result is what matters in those cases, not the precise wording of the disposition .... Using the language of the memdispo to predict how the court would decide a different case would be highly misleading."). In his comments opposing the proposed Federal Rule of Appellate Procedure 32-1 mandating open citation, Judge Kozinski suggests, metaphorically, that summary decisions are not merely useless as precedent but potentially toxic: "When the people making the sausage tell you it's not safe for human consumption, it seems strange indeed to have a committee in Washington tell people to go ahead and eat it anyway." Kozinski Comments, supra note 10, at 2. [FN123]. Kozinski Comments, supra note 10, at 4. [FN124]. Hart, 266 F.2d at 1179. [FN125]. Id.; see also Symbol Tech., Inc. v. Lemelson Med., 277 F.3d 1361, 1368 (Fed. Cir. 2002) ("Courts contribute to the growing imprecision, uncertainty and unpredictability of the law by issuing repetitive opinions on subjects that have been thoroughly irrigated."). [FN126]. See, for example, the Federal Circuit's rules explaining that disposition by uncitable nonprecedential opinions means the opinion "would not add significantly to the body of law or would otherwise fail to meet" one of the following criteria: the opinion treats an issue of first impression; establishes a new rule of law; criticizes, clarifies, alters or modifies an existing rule of law; or applies an existing rule to facts significantly different from previous cases. Fed. Cir. R.
10(3)-(4). The fact that there are significant numbers of dissents in at least one no-citation circuit, the Ninth, suggests that this may be somewhat overstated, but it is still likely that these cases are relatively uncontroversial. [FN127]. For example, between Jan. 1, 2004 and Mar. 1, 2004, the Ninth Circuit had a reversal rate of 34% in its citable precedents and only a 13% reversal rate in its uncitable cases. The comparable figures for the Second Circuit were 48% and 11%, and for the Seventh Circuit, 36% and 2%. [FN128]. Medina v. California, 505 U.S. 437, 445 (1992); Patterson v. New York, 432 U.S. 197, 202 (1977); Speiser v. Randall, 357 U.S. 513, 523 (1958); Snyder v. Massachusetts, 291 U.S. 97, 105 (1934), quoted in Leland v. Oregon, 343 U.S. 790, 798 (1952). [FN129]. Rock v. Arkansas, 483 U.S. 44, 56 (1987). [FN130]. Mathews v. Eldridge, 424 U.S. 319 (1976). [FN131]. It is hard to say why the Court has used balancing tests for some litigation-based due process challenges and the "fundamental fairness" approach in others. Deciding which test would apply here is further complicated by the fact that many of the previous cases involve challenges to state criminal procedure, which, under principles of federalism, presumably is entitled to considerably more deference from the federal judiciary than the Federal Rules of Appellate Procedure. An entirely different standard might apply when the Supreme Court is putting its own federal judicial house in order. It is also quite possible the Court would apply different doctrines depending on whether the no-citation rules were challenged in the context of a civil or a criminal case. [FN132]. Medina, 505 U.S. at 448 (quoting Dowling v. United States, 493 U.S. 342, 352 (1990)). [FN133]. Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1, 17 (1991) (quoting Jackman v. Rosenbaum Co., 260 U.S. 22, 31 (1922)). [FN134]. For at least one member of the Court, a practice found to comport with common law tradition that does not otherwise violate the Bill of Rights is per se all the process that is due. See Pac. Mut. Life Ins. Co., 499 U.S. at 24-25 (Scalia, J., concurring). [FN135]. Honda Motor Co. v. Oberg, 512 U.S. 415, 430 (1994). [FN136]. See Hart v. Massanari, 266 F.3d 1155, 1165-69 (9th Cir. 2001); see also notes for Section III, infra. [FN137]. Lance A. Wade, Honda Meets Anastasoff: The Procedural Due Process Argument Against Rules Prohibiting Citation to Unpublished Judicial Decisions, 42 B.C. L. Rev. 695, 722-23 (2001). This is the only previous published analysis of the procedural due process implications of no-citation rules that I was able to find. It is limited to the fundamental fairness approach. [FN138]. Hart, 266 F.2d at 1168-69; see also Thomas Healey, Stare Decisis as a Constitutional Requirement, 104 W. Va. L. Rev. 43, 89 (2001). [FN139]. See Anastasoff v. United States, 223 F.3d 898, 903 (8th Cir. 2000) (noting that the absence of a reporting system in the eighteenth century did not impede precedential authority of cases). [FN140]. Hart, 266 F.3d at 1180. [FN141]. Anastasoff, 223 F.3d at 903 n.14 ("[T]he common-law view ... considered entry on the official court record sufficient to give a decision precedential authority whether or not the decision was subsequently reported."). [FN142]. Id. at 903 ("[Eighteenth-century] judges and lawyers ... recognized the authority of unpublished decisions even when they were established only by memory or by a lawyer's unpublished memorandum."). [FN143]. Re Rules of the United States Court of Appeals for the Tenth Cir., Adopted November 18, 1986, 955 F.2d 36, 37 (1992) (Holloway, C.J., Barrett, J., and Baldock, J., concurring in part and dissenting in part) (The 1986 dissent was reprinted in the 1992 opinion). Since that decision, the Tenth Circuit has revoked its no-citation rule. Under today's rules, unpublished cases are still "not binding precedents," and their citation is "disfavored." 10TH CIR. R. 36.3(A)-(B). "But an unpublished decision may be cited if: (1) it has persuasive value with respect to a material issue that has not been addressed in a published opinion; and (2) it would assist the court in its disposition." 10TH CIR. R. 36.3(B). [FN144]. Re Rules of the United States Court of Appeals for the Tenth Cir., 955 F2d at 37. See also Harris v. United Fed'n. of Teachers, NYC Local 2, 2002 WL 1880391 (S.D.N.Y. 2002), in which District Judge Gerald Lynch cites a Second Circuit opinion covered by that court's citation ban, and criticizes the appellate court for "pretending that this decision never happened and that it remains free to decide an identical case in the opposite manner," explaining that he "finds the opinion of a distinguished Second Circuit panel highly persuasive, at least as worthy of citation as law review student notes, and eminently predictive of how the Court would in fact decide a future case such as this one." Id. at *1 n.2. [FN145]. Outside of the due process analysis, the Supreme Court has held that statutory restrictions on Legal Services lawyers' constitutional arguments violated the First Amendment. See Legal Servs. Corp. v. Velazquez, 531 U.S. 533, 545 (2001). The Court explained that the restrictions were "inconsistent with the proposition that attorneys should present all the reasonable and well-grounded arguments necessary for proper resolution of the case," and that "[b]y seeking to prohibit the analysis of certain legal issues and to truncate presentation to the courts, the [statute restricted] expression upon which courts must depend for the proper exercise of judicial power." Id. [FN146]. Rock v. Arkansas, 483 U.S. 44, 55-56 (1987). [FN147]. In Rock, the Supreme Court reviewed Arkansas's evidentiary rule excluding a homicide defendant's hypnotically refreshed testimony. Id. at 45. The Court noted that a criminal defendant's right to testify in her own behalf has several constitutional sources, including the due process clause. Id. at 51. The Court listed the Sixth Amendment's Compulsory Process Clause, guaranteeing the defendant's right to call "witnesses in his favor;" the Fifth Amendment's guarantee against compelled testimony; and the guarantee of due process under the Fourteenth Amendment. Id. at 51-53. [FN148]. Id. at 56. On the way to developing a standard, the Court reviewed both Sixth Amendment and due process cases. Id. at 51-53. One of the due process examples was the exclusion of a defense witness's exonerating testimony under a state hearsay rule, which the Court had struck down because it had been applied "mechanistically." Id. at 55 (citing Chambers v. Mississippi, 410 U.S. 284 (1973)). [FN149]. Rock, 483 U.S. at 56. [FN150]. Id. at 61. [FN151]. See Kozinski Comments, supra note 10, at 20 (analogizing no-citation rules to other types of rules that restrict advocacy). [FN152]. Mathews v. Eldridge, 424 U.S. 319, 335 (1976); see also Logan v. Zimmerman Brush Co., 455 U.S. 422, 430 n.5 (1982) ("[H]aving made access to the courts an entitlement or a necessity, the State may not deprive someone of that access unless the balance of state and private interests favors the government scheme."). [FN153]. A recent article discusses the special problems of the courts' use of unpublished and uncitable opinions in death penalty cases. David R. Dow & Bridget T. McNeese, Invisible Executions: A Preliminary Analysis of Publication Rates in Death Penalty Cases in Selected Jurisdictions, 8 TEX. J. ON C. L. & C.R. 149 (2003). [FN154]. Cf. Atchison, Topeka & Santa Fe Ry. Co. v. Wichita Bd. of Trade, 412 U.S. 800, 807-08 (1973) (finding that a government agency's settled course of behavior leads to a presumption that policies committed to it are best carried out when a settled rule is adhered to). [FN155]. Davis v. Comm'r, 65 T.C. 1014, 1022 (1976); see also Margaret Gilhooley, The Availability of Decisions and Precedents in Agency Adjudications: The Impact of the Freedom of Information Act Publication Requirements, 3 ADMIN. L.J. 53, 58 (1989) (noting that precedential effect is withheld from decisions that are viewed as "merely a repetitive application of settled law"). [FN156]. See generally JERRY L. MASHAW, BUREAUCRATIC JUSTICE: MANAGING SOCIAL SECURITY DISABILITY CLAIMS (1983). [FN157]. See id. at 29-30 (contrasting the goals of the adjudicative process with those of bureaucratic justice). [FN158]. The Fifth Circuit does have a process for issuing decisions without any opinions, but it was in place before they allowed citation of nonprecedential opinions. One of the no-citation courts, the Second Circuit, at one time disposed of "up to 70%" of its cases through brief oral decisions from the bench. Second Circuit Comments, supra note 16, at 3. But, assuming that those oral dispositions can be transcribed and made available in searchable form, that is not the same thing as resolution without reasoning. Indeed, two commentators have suggested that such oral dispositions could be a good substitute for uncitable written summary decisions. See Greenwald & Schwarz, supra note 20, at 1169-73. They suggest that such a practice would save judicial time and might have additional benefits, in particular "the opportunity to put a face on justice" and "satisfy a desire on the part of litigants and their counsel to interact directly with their decision makers." Id. at 1173. And as they note, such extemporaneous, or planned, oral disposition is common today in trial courts. Id. at 1171. [FN159]. Hart v. Massanari, 266 F.3d 1155, 1165-69 (9th Cir. 2001). [FN160]. Id. at 1178. [FN161]. Id. at 1158-59. Though Hart's main analytic focus was on refuting Anastasoff's theory that nonprecedential opinions violate Article III, Hart's actual holding was directed to the ban against attorneys' citing such decisions. In fact, the case came up because the court ordered an attorney who had cited a "nonprecedential" case to show cause why he should not be sanctioned. The lawyer responded that the rule might be unconstitutional, citing Anastasoff, and it was off to the races. [FN162]. Id. at 1165-69. [FN163]. Id. "As the concept of law changed and a more comprehensive reporting system began to take hold, it became possible for judicial decisions to serve as binding authority." Id. at 1168. [FN164]. Id. at 1165. [FN165]. Id. at 1166-67. [FN166]. Id. at 1167. [FN167]. Id. at 1168. [FN168]. Id. at 1167. [FN169]. Id. at 1168. [FN170]. Id. [FN171]. Id. at 1175. [FN172]. Cf. Gilhooley, supra note 155, at 58 ("The availability of decisions makes possible the precedential effect of decisions ...."). [FN173]. Hart, 266 F.3d at 1176. [FN174]. Id. at 1170. [FN175]. "The rule of decision cannot simply be announced, it must be selected after due consideration of the relevant legal and policy considerations." Id. at 1176. [FN176]. 9TH CIR. R. 36-3(b)(iii). [FN177]. Danny J. Boggs & Brian P. Brooks, Unpublished Opinions & the Nature of Precedent, 4 GREEN BAG 17 (2000). [FN178]. Id. at 19-20. [FN179]. Id. at 20. [FN180]. Id.; see also Richard B. Cappalli, The Common Law's Case Against Non-Precedential Opinions, 76 S. Cal. L. Rev. 755, 768-69 (2003) ("[T]he actual scope of a doctrinal formulation is learned through its applications and not through the words chosen to express the doctrine."). [FN181]. Cf. Robel, supra note 81, at 401 (suggesting a cultural conception of stare decisis that includes lawyers and judges, rather than a purely rule-bound conception). [FN182]. Hart, 266 F.3d at 1177. [FN183]. Id. at 1170. [FN184]. Id. at 1171. [FN185]. Id. at 1179. [FN186]. Id. [FN187]. Id. [FN188]. Cf. FRIENDLY, supra note 32, at 42-43 (discussing a proposed dual nature appellate review, namely, corrective and interpretive). [FN189]. Hart, 266 F.3d at 1176-77. [FN190]. Id. at 1176. [FN191]. FRIENDLY, supra note 32, at 42-44. [FN192]. Id. at 43-44. [FN193]. "Is the distinction between the 'corrective' and the 'institutional' [or lawmaking] function viable?" Id. at 44. [FN194]. Judge Friendly cites a case in which "[t]he district judge and the three judges of the court of appeals agreed on the applicable 'principles'-- that a serviceman did not give up his First Amendment rights but that the interest of maintaining discipline justified curtailments that would not have been permissible for the ordinary citizen." Id. at 44 n.149. Still, the district court came out differently than the appellate panel. Friendly asks, "Is this 'error' or 'policy?"' Id. [FN195]. For instance, in a two-month period in 2004, the Ninth Circuit issued 537 uncitable opinions: 72 of these reversed or vacated all or part of the judgment being appealed. See supra note 18. [FN196]. Picazo v. Alameida, No. 03-55497, 2004 WL 326399 (9th Cir. Dec. 19, 2004). [FN197]. Id. at **3-4. [FN198]. Id. at *1 (quoting United States v. Lockett, 919 F.2d 585, 590 (9th Cir. 1990)). [FN199]. Id. [FN200]. Brecht v. Abrahamson, 507 U.S. 619 (1993). [FN201]. Picazo, 2004 WL 326399 at *3. [FN202]. U.S. CONST. art. III
2, cl. 1 ("The judicial Power shall extend to all Cases, in Law and Equity, arising under this Constitution, ... [and] controversies ...."). [FN203]. See Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803) ("It is emphatically the province and duty of the judicial department to say what the law is .... If two laws conflict with each other, the court must decide on the operation of each."). [FN204]. See Boggs & Brooks, supra note 177, at 22 (making a similar argument in response to an article written by Ninth Circuit Judges Reinhardt and Kozinski (author of Hart), that by contending that "a judicial decision has only as much precedent as the writing judges intend to give it," they are privileging opinion writing beyond its supporting role in "the core judicial function of deciding cases or controversies" and making judges' decisionmaking more like legislation). [FN205]. Annette B. Weiner, From Words to Objects to Magic: "Hard Words" and the Boundaries of Social Interaction, in DANGEROUS WORDS: LANGUAGE & POLITICS IN THE PACIFIC 161, 162 (Donald Lawrence Brenneis & Fred R. Myers eds., 1984). [FN206]. No doubt supporters of citation bans would choose a different metaphor for the effects of unchecked precedential circulation. One reader of this article suggested that summary decisions are better imagined as viruses than as gifts. Maybe the point is that if you want the potential of the gift, you have to be willing to risk the virus. [FN207]. 6TH CIR. R. 28(g). [FN208]. D.C. CIR. R. 36(c)(2). [FN209]. United States v. Mead Corp., 533 U.S. 218, 227 (2001). [FN210]. Id. at 219. [FN211]. Id. at 235 (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) ). [FN212]. Id. [FN213]. Hart v. Massanari, 266 F.3d 1155, 1171 (9th Cir. 2001). [FN214]. Lucinda Williams, Are You Down, on ESSENCE (Lost Highway 2001). [FN215]. THE FEDERALIST NO. 78, at 523 (Alexander Hamilton) (Jacob E. Cooke ed., 1961). [FN216]. Much of judges' official speech exemplifies the "performative" aspect of language identified by J.L. Austin and other mid-twentieth-century ordinary language philosophers, i.e., the capacity of language to change rather than simply to describe reality. J.L. AUSTIN, HOW TO DO THINGS WITH WORDS 6-7 (2d ed. 1975). Note, however, that the active quality I am attributing to precedential language is a different phenomenon--a kind of hyper-persuasiveness. [FN217]. Weiner, supra note 205, at 167. [FN218]. Id. [FN219]. Id. at 165. [FN220]. Id. at 167. [FN221]. Id. [FN222]. Id. at 166. [FN223]. The story of Trobrianders' hard words also suggests another perspective on judges' concerns about allowing citation of their summary opinions. Both Trobriand precautions regarding hard words and judges' use of no-citation rules respond to the aspect of language that partakes of a power its users themselves cannot fully control. Like Trobrianders' "hard words," judges' words spoken in explanation of a judgment acquire a kind of potential to move future courts in some direction, even if the very judges who originally wrote the decision subsequently disclaim any intention to initiate such movement. Here, then, is another reason for courts' use of no-citation rules. The prohibition on citation of previous "nonprecedential" rulings in official documents or proceedings before the court that made those rulings recognizes the intractable effectiveness of judges' "hard" words. On this view, because those effects flow from conventional and institutional sources judges cannot fully control, the only way to quench the precedential effect of judges' words is to ensure that they will never be spoken again in a context that could trigger that effect. [FN224]. Another way to think about this is to think of that other, specialized sense of the term "acting." In the theater young actors quickly learn that the audience sees everything they do (and do not do) and understands their characters and the play based on everything that happens on the stage, not just the particular gestures and expressions the actors planned. Onstage, as in life, we cannot select just a few of our actions to be consequential. [FN225]. Again, the analogy to Trobriand hard words works, for Weiner tells us that "'[h]ard words' once spoken cannot be recalled." Weiner, supra note 205, at 167. Flipping this observation, we can see that words whose effects can be recalled (or suppressed) cannot be hard. END OF DOCUMENT