Judging Justice on Appeal
Injustice on Appeal: The United States Courts of Appeals in Crisis
BY WILLIAM M. RICHMAN & WILLIAM L. REYNOLDS NEW YORK, NY: OXFORD UNIVERSITY PRESS, 2012, PP. 256. $89.50.
author. Associate Professor of Law, Duke University School of Law. Thanks to Will Baude, Ben Blocher, Joseph Blocher, Guido Calabresi, Josh Chafetz, Taylor Crabtree, Frank Levy, Stephen Sachs, Katherine Swartz, and the excellent editors of the Yale Law Journal, particularly Robert Quigley. All views expressed here, as well as any errors and solecisms are, of course, my own.
When recently asked what he thought was one of the greatest problems plaguing the federal judiciary, Supreme Court Justice Samuel Alito responded by saying the “crushing” workload faced by his former colleagues on the courts of appeals.1 The Justice’s statement should come as no surprise. For close to half a century, judges and scholars alike have spoken out about a critical problem facing the federal appellate courts: the caseload has grown at an exponential rate.2 Whereas in 1950 circuit judges had to review an average of only 73 appeals,3 their modern counterparts must decide more than four times as many, with an average of 329 appeals per annum today.4 Indeed, it is on account of this precipitous rise in caseload that the federal courts of appeals have so often been said to face a “crisis in volume.”5
Attempts to assess and ameliorate this crisis have come from the bench, bar, and academy. Several committees and commissions have studied the problem, including, most notably, the Commission on Revision of the Federal Court Appellate System (also known as the Hruska Commission) in the 1970s,6 the Federal Courts Study Committee in the 1980s,7 and the Commission on Structural Alternatives for the Federal Courts of Appeals (or the White Commission) in the 1990s.8 Throughout this time, prominent jurists including Chief Justice William Rehnquist,9 Judge Henry Friendly,10 Judge Jon Newman,11 and Judge Richard Posner12 have weighed in from the bench, writing that the rising caseload stands to harm the court system as we know it. Scholars including Paul Carrington and Daniel Meador have similarly written extensively about the predicament,13 discussing measures the courts could undertake to preserve their fundamental nature as the caseload mounts.These efforts to better understand our court system in the face of this crisis helped create, and ultimately define, the field that is now called “judicial administration.”14
Over the past thirty years, no one has contributed more to this field than two court scholars together—William M. Richman and William L. Reynolds. Through a series of critical articles,15 Richman and Reynolds were able to pinpoint the precise effects of the caseload crisis, both on litigants and the system as a whole. Furthermore, they were able to show the interplay of these various effects, providing a holistic account of the problem in a way that no one else had done.16 Their recent book, Injustice on Appeal: The United States Courts of Appeals in Crisis,17 stands as a culmination of their earlier work,18 bringing together vital analysis of the caseload crisis, the ways in which appellate review has suffered as a result of that crisis, and potential solutions. More broadly, Injustice on Appeal stands as one of the most comprehensive and thoughtful accounts of the largest problem facing the federal judiciary today.
Part I of this Review assesses the major contributions of the book—namely Richman and Reynolds’s detailed discussion of the effects of the courts’ staggering caseload. As the Injustice authors were among the first to point out, the courts’ increased workload has meant that only a fraction of all appeals now receive what one might call “traditional” appellate adjudication in the form of oral argument, consideration by a judge and his or her clerks, and then a published opinion. Instead, the vast majority of appellate litigants currently receive no oral argument,19 have their cases worked up primarily by staff attorneys,20 and then have their cases disposed of via unpublished order or summary judgment.21 Injustice delineates the losses associated with each of these case-management developments, but then makes two larger, profound points. First, it is not simply that some cases receive less judicial attention overall, but rather that some kinds of cases receive less attention—namely, social security cases, prisoner cases, and criminal cases (or, as the authors point out, cases brought by parties who are arguably the most vulnerable in our legal system). Second, by deciding which cases will receive significant judicial attention and which cases will receive very little, the courts of appeals have begun to resemble certioraricourts—a move they have made entirely on their own. Part I concludes by assessing the significance of these two contributions and their theoretical implications.
While the Review by and large concurs with Richman and Reynolds’s diagnosis of the problems facing the federal judiciary, it parts company on the cure. Part II assesses the final portion of Injustice, which provides the authors’ would-be prescriptive measures. Specifically, Richman and Reynolds argue that the obvious solution to the problem of an overworked judiciary is simply to increase the size of the judiciary—and substantially, by as much as one hundred percent.22 While increasing the resources of a resource-constrained court system might seem like a fitting response, the authors’ proposal falls short in two respects. First, it fails to provide any sort of detail about how this change would be implemented, including whether it would involve increasing the size of the existing circuits or adding new circuits altogether, and whether it would happen gradually or in a condensed timeframe. Problems afflict all of these implementation mechanisms, and because the authors fail to articulate a full proposal for expanding the judiciary, it is possible that Richman and Reynolds’s solution might amount to a cure that is worse than the disease.
Yet beyond these ground-level practical problems is a greater problem still—the problem of political reality. By the authors’ own account, proposals to expand the bench have stalled because the judiciary has been opposed to them and Congress has not been moved to act on its own.23 Barring an explanation of why the motivations of both branches have changed—something Richman and Reynolds do not provide—it seems highly unlikely that these same branches would now support a plan to add over 150 new judges (at an initial cost of over $150 million24) any time soon.
The final Part of the Review therefore focuses on more fruitful avenues for improving the judiciary. In particular, Part III considers ways of enhancing “non-traditional” review through means such as altering judicial voting practices and increasing the specialization of staff attorneys. It also sets out potential ways to improve appellate adjudication more generally, focusing on the use of visiting judges, senior judges, and other so-called “housekeeping” practices of the courts that together have a significant impact on tens of thousands of cases each year.25
The Review concludes that although Richman and Reynolds may not have provided the optimal prescription for improving the state of the federal courts of appeals, their book serves a critical function by providing the most comprehensive account of the problems plaguing those courts today and over the past several decades. That account stands to make a significant contribution on its own, and will no doubt inspire a second wave of scholars to contemplate one of the key questions for the judiciary and the academy: how can we improve the quality of appellate review in this country?
Injustice on Appeal begins “BCE” or “before the caseload explosion.”26 The authors briefly trace the history of the appellate courts, reminding the reader that the courts as we know them today did not come into existence until little more than a century ago, with the Evarts Act of 1891.27 Richman and Reynolds’s key observation is that during the judiciary’s salad days, judges could decide appeals at their leisure. Indeed, through the middle of the twentieth century, judges provided what we might call traditional appellate review, or what the authors dub the “Learned Hand Treatment” after the famed judge of the Second Circuit, to nearly every case on their docket.28 Specifically, this meant that appeals were almost universally given oral argument and decided by published opinion, and that law clerks played a limited role in the decisionmaking process.29
But as the authors point out, traditional appellate review soon gave way as the caseload began a period of rapid growth. In 1960 the number of cases filed in the geographic circuit courts was 3,899;30 by 1970 that figure had swelled to 11,662;31 by 1980 it had more than doubled to 23,200;32 and by 2010 it had more than doubled again to 55,992.33 The book devotes little space to what caused the expansion, but it briefly identifies a few contributing factors. First, Richman and Reynolds touch on the general growth in population.34 (Specifically, in 1960 the population of the United States was just under 180 million and within twenty years it had grown to over 225 million.35) As the authors succinctly put it, “more people generate more litigation and more economic activity, which in turn creates more and more complex litigation.”36 A second factor is the increased legislative activity of Congress, causing, in turn, an increase in federal law—both in amount and in complexity.37 And a third factor is the heightened activity of the courts themselves—particularly the Warren Court—which read “many new federal rights” into the Constitution.38 The combination of at least these factors created a seismic change in the caseload of the federal courts. To sharpen the point, Richman and Reynolds note that “over the past fifty years, the circuit courts’ caseload has increased by 1436 percent, a more than fourteenfold increase.”39 By contrast, the number of judgeships has only grown from 68 to 167, meaning that the workload per judgeship has increased by almost 600% in this time.40
Given the magnitude of the increase in caseload vis-à-vis the increase in the number of judges, the federal appellate courts clearly could not have continued to give every case a leisurely review. And in fact, beginning in the 1960s, the courts began developing a series of “case-management” measures to cope with the increase in volume. The main portion of Injustice is devoted to documenting and assessing these case-management tools and the effects of their implementation.
First, as Richman and Reynolds detail over several chapters, federal appellate judges began publishing full opinions in a smaller percentage of total cases.41 Judges have often reported that writing full-length, precedential opinions is the most time-consuming aspect of their job, with some sources suggesting that opinion-writing takes up over half of judges’ working hours.42 It should come as no surprise, then, that as the caseload began increasing, judges considered whether some cases could be disposed of in a shorter (and faster-to-execute) form. The upshot was a 1964 decision by the Judicial Conference of the United States that only opinions of “general precedential value” should be published,43 meaning that cases that would not give rise to such opinions could be decided by more succinct unpublished orders.44 Soon after, each of the circuit courts devised its own rules for when cases could be decided by unpublished disposition.45 Within only a few years, more cases were being decided by unpublished disposition than by published opinion46—a fact that remains true today.47
Second, judges began to hold oral argument in a smaller percentage of total cases.48 Following opinion writing, oral argument is one of the most time-intensive components of a court’s business. There is, of course, the direct cost of oral argument, which is typically thirty minutes per case in many circuits.49 This is a significant amount of time if the court were to hold argument in all or even the majority of cases, considering that the average number of filings per judgeship is 329 and that appellate judges hear cases in panels of three (thereby effectively tripling that figure).50 And yet, the indirect costs associated with oral argument are arguably just as great. Beyond the time judges spend in the courtroom questioning the parties is the time spent preparing lines of inquiry and, though easy to overlook, the time most judges spend traveling to the courthouse from another city and possibly another state at multiple points throughout the year.51 Accordingly, as with traditional opinions, one could predict that when the courts became inundated with cases, they would begin to limit the share of cases that would receive oral argument. Indeed, beginning in 1968 with the Fifth Circuit, the courts did just that.52 This change in appellate process was formalized with a 1979 amendment to Federal Rule of Appellate Procedure 34, which authorized the resolution of an appeal without oral argument when a three-judge panel determined that the appeal was “frivolous,” the dispositive issue had already been “authoritatively decided,” or the decisionmaking process “would not be significantly aided by oral argument.”53 And today, the vast majority of appeals that are decided on the merits are decided solely on the briefs.54
Third and finally, judges began to rely more on what the authors call “additional decision makers”—particularly law clerks and staff attorneys—to assist in performing their official responsibilities.55 The increase in the number of law clerks or “elbow clerks”56 allotted to each appellate judge roughly tracked the increase in caseload. As Richman and Reynolds note, appellate judges went from receiving one clerk in the 1930s to two in 1969, three in 1979, and four today.57 Additionally, beginning in 1973, courts began to receive funding to hire staff law clerks, as distinct from elbow clerks, to assist with certain kinds of cases including pro se and habeas appeals.58 Within less than a decade, Congress authorized the creation of official staff attorney offices to further assist with the caseload.59 The authors report that as of 2008, there were five hundred staff attorneys—or nearly three for every judgeship—helping to defray the work in the federal appellate courts.60
In addition to documenting these marked changes in the federal appellate process, Richman and Reynolds identify the key costs associated with them. Beginning with limiting the publication of full-length opinions, the authors argue that relying to a great extent on unpublished opinions ultimately lessens judicial accountability, in that such dispositions generally receive less scrutiny (by other members of the bench and the public) and are often unsigned, meaning that no single judge has to take ownership of the decision.61 The authors further make the case that disposing of so many appeals without precedential opinions adversely affects the development of the law, in that it places the majority of decisions outside the realm of stare decisis.62 As a result, they argue, these cases may not be decided as carefully (because the panel knows they are not binding on future cases) and, in the opposite direction, the absence of these datapoints in the Federal Reporter may hinder the proper decision of future cases (because future panels will have fewer examples to draw upon or distinguish from).63
Turning to the reduction in oral argument, Richman and Reynolds assess the value of having in-person arguments in each case, thereby highlighting the losses attendant on limiting such arguments as a timesaving technique. Those benefits include ensuring that the judges are fully informed about the facts of the case and the relevant caselaw when reaching a decision,64 and providing legitimacy to their decisionmaking (by showing the parties and the public that they have considered the matter at hand).65 Finally, Injustice delineates the problems that come with having additional decisionmakers in the circuit courts, focusing mostly on “overdelegation”—i.e., having staff, be they law clerks or staff attorneys, take on what are truly Article III responsibilities (such as drafting opinions) without sufficient oversight.66
By canvassing the vast changes in the appellate courts in the last sixty years and the effects of those changes, Injustice makes a significant contribution to the literature on court administration and indeed, to the collective understanding of the federal judiciary more generally. One might question some of the authors’ conclusions—specifically regarding how great the losses attendant on these case management techniques truly are—or whether the authors sufficiently consider the benefits that also come with such techniques.67 Yet overall, these chapters by Richman and Reynolds arguably provide the most comprehensive account of how the federal appellate courts have been changed in the face of the workload increase in the past few decades.
But beyond these points, Injustice makes a greater contribution still by showing the macro effect of these case-management decisions—namely, that certain kinds of cases tend to receive traditional appellate treatment while others tend to receive less (and, indeed, little) judicial attention. In the words of the authors, “a litigant in an ‘important’ antitrust or securities case, one who is represented by serious counsel, will get the full Learned Hand Treatment” (meaning oral argument, careful consideration by a panel of judges, and a published opinion).68 In contrast, the authors note that “[a] litigant who is poor, without counsel, and with a boring, repetitive problem . . . can expect only . . . second-hand treatment . . . .”69 Examples of cases that consistently receive what the authors call “track two” treatment include those involving prisoner rights, social security, and criminal convictions70 (and though the authors do not specifically mention them here, immigration and pro se cases generally receive this limited treatment as well71). By combining this insight with their earlier analysis, Richman and Reynolds are able to argue convincingly that because they receive less in the way of judicial resources, the track two set of litigants also receive less in the way of overall decisionmaking quality.
Accordingly, it is not simply that some litigants bear the costs of the judiciary’s decision72 to pare down the process that it generally affords appellants, but rather that certain litigants bear the majority of these costs. In this way, the case-management practices of the courts have something of a disparate impact on the different appellants who come before them, with, as the authors argue, those who are socially and economically disadvantaged receiving a lower quality of justice.73 From the authors’ standpoint, the federal courts that exist for the benefit of society now underserve a huge swath of the population—the same swath that is underprivileged more generally.74 It is this inequality (reinforcing a broader inequality) that stands as the “injustice” in Injustice on Appeal.
It would be difficult to overstate the significance of Richman and Reynolds’s contribution with this analysis. They pinpoint a critical issue with the way justice is delivered in the appellate courts—one that impacts thousands of parties each year.75 This will undoubtedly help pave the way for numerous articles examining court process, as their previous work has already done,76 which will only add to our substantive understanding of the federal courts. Now here again, the authors could have done more with their analysis—specifically, to address the fact that there are arguably valid reasons for the courts to treat these parties differently, in light of the caseload pressures that courts face.77 For example, as the authors themselves intimate, many of the classes of cases that receive less judicial attention raise repetitive legal claims,78 thereby making them prime candidates for a limited form of review. Still, Injustice fundamentally adds to our knowledge of the appellate courts, regarding both the changes that those courts made and the ramifications of those changes (both individually and cumulatively).
Finally, in making these critical substantive contributions, Injustice makes an important process-based point as well. The authors note that in creating what essentially amounts to a two-track system of treatment, the judges have decided which cases to give full attention to and which cases at least initially will be handled primarily by staff. Richman and Reynolds point out, quite powerfully, that the courts of appeals now resemble certioraricourts, with the judges essentially having the meta-ability to decide what they will decide.79 This is a critical change from the judiciary’s earlier days in two respects. First, it has a fundamental impact on the nature of our court system. It is one thing not to guarantee review by the Supreme Court, as that usually amounts to a second review of the initial district court decision. But to not guarantee full review by even the intermediate appellate courts means that some portion of district court decisions will be virtually unchecked, thereby calling into question the very notion of an appellate system.80 Second, this shift has separation-of-powers implications. As the authors note, the appellate courts have shifted towards a certiorarimodel without the assent of Congress.81 This is in direct contrast to the way the Supreme Court’s docket became largely discretionary,82 and is arguably a usurpation of the legislative role.83
While these are vital points, one can again argue that the authors’ analysis is incomplete. Specifically, though the authors acknowledge that in dubbing the circuits “certiorari courts” a generation ago they had given in somewhat to hyperbole,84 they fail to explore the limitations of their analogy. For example, even cases that receive track two treatment are still decided by the court, with the judges still involved in the decision (even if not to the full extent they are involved in track one cases). That said, the point is well taken—as Injustice shows, the circuit courts have altered the way that they review cases and have done so without the explicit authorization of Congress.
In short, the majority of Injustice provides court scholars and judges with key ways to think about how the courts have changed over the last half century. In so doing, Richman and Reynolds have offered invaluable knowledge about, and insight into, the federal courts.
Injustice does not conclude after its compelling descriptive analysis, however. The last several chapters turn to the normative, with the authors prescribing what the courts should do to solve their current problems. And it is here that the book falls short, with a recommendation that is incomplete at best and inconsistent with its earlier analysis at worst.
Having spent ten chapters thoroughly diagnosing the problem of the federal courts as being too many cases for the number of judges, the authors promise to spend the final four recommending what appears to be a natural solution: increasing the size of the federal appellate bench.85 And indeed, because Richman and Reynolds have provided such a comprehensive and compelling account of the caseload crisis, the reader is anxious at this point to learn how that crisis can be abated. Unfortunately, though, the remedy is more assumed than proposed. The book devotes only a few paragraphs to a subsection entitled “The Obvious Solution: More Judges,”86 and nowhere in that space actually makes a direct call for that solution. Instead, the authors discuss how overworked the current judiciary is, simply reinforcing the message of the previous chapters.
To be clear, the authors not only fail to make an affirmative case for their solution, they fail to specify precisely what their solution is. This is particularly problematic because there are significant decisions to make as far as how one should implement a judicial expansion. First, by how much exactly should the bench be increased? Second, how should the increase occur—by adding judges to the existing circuits or by creating new circuits altogether? Third, should this expansion take place gradually or as quickly as possible? What’s more, decisions at any of these junctures are open to criticism. For example, in deciding how the increase should occur, one could argue, as some have, that adding to the existing circuits would make those courts less efficient and harm the collegiality of their members.87 As it stands, many commentators already question the functionality of the current Ninth Circuit;88 generating ten more such circuits hardly seems ideal. On the other hand, the creation of new circuits poses problems of its own. Carving up existing circuits is a daunting task at best (again, one need only look to discussions about the Ninth Circuit89). And in light of the fact that there are already concerns about too many circuit splits that are left unresolved by the Supreme Court,90 the country would arguably face even greater disuniformity of federal law with more circuits.91 The authors consider some of these objections later in Injustice, when critiquing concerns about expansion more generally.92 But without having a clear sense of what the authors are truly proposing93 and how that proposal would be implemented, it is impossible to evaluate whether their would-be solution might actually be a cure that is worse than the disease.94
Instead of detailing a solution, the authors quickly pivot to why such a solution has not already come to pass.95 The main culprit, according to the authors, is the judiciary itself. Specifically, Richman and Reynolds argue that appellate judges have helped keep their own numbers down through, among other means, lobbying efforts (both writing and speaking out against creating new judgeships)96 and even directly voting against recommendations to add more judgeships to their circuits.97
The fact that appellate judges appear to be resisting the idea of increasing the size of the bench raises the question of why, in light of the sizeable caseload, they are not embracing the possibility of an expansion. Here, Richman and Reynolds provide several accounts, beginning with those that touch upon the quality of the bench. For example, the authors note the expressed concern of numerous judges that there simply would not be a sufficient number of qualified applicants to fill an increased number of judgeships and that some of those who are qualified might no longer be interested since the prestige of the office would have diminished.98 The accounts then extend to the stability of the law. Specifically, Richman and Reynolds discuss the fear on the part of the bench (as touched on above) that adding judgeships would create inconsistencies in the law within circuits or increase those that exist between circuits.99 Finally, the authors describe the concern regarding the financial cost of implementing a significant expansion in judgeships (a cost of approximately $1 million per new judge or as much as $150 million altogether).100
The authors then respond with mixed success to what they see as the judges’ reasons for opposing an expansion of the bench. They are on their strongest footing when they counter the concerns, lodged by several prominent judges,101 that an increased bench would necessarily result in a lower quality bench. Here, Richman and Reynolds point to the numbers, noting that in 1960 there were 4,205 attorneys for every circuit judgeship and in 2010 there were 5,600 attorneys for every circuit judgeship.102 They persuasively argue that even if the bench were expanded considerably, there would still be a robust number of lawyers per judgeship, meaning that there would certainly be a sizeable enough pool in which to find qualified applicants.103
The authors are less successful when they dismiss concerns about the effects an expanded judiciary would have on the coherence of the law. Focusing on adding judgeships to existing circuits, Richman and Reynolds claim that “[t]here is simply no evidence that increasing the number of judgeships within a circuit reduces the stability of circuit law.”104 While this statement might be technically true, more than two-thirds of the appellate bench reported that they believed the maximum number of judges for a court to “function as a single decisional unit,” creating a cohesive body of law, is between eleven and seventeen.105 It is not difficult to imagine why this would be so. As the White Commission detailed, judges of smaller circuits can read the opinions of all of their colleagues before they are published—thereby providing the opportunity to correct “inadvertent inconsistenc[ies]” in the law and also keeping everyone abreast of new legal developments.106 Additionally, once a circuit becomes quite large, as in the case of the Ninth, it becomes no longer practicable to hold en banc proceedings as a single, unified court.107 It seems quite possible that not being able to have the full court review any number of issues would lead to a less coherent body of law. There are some studies that call into question just how much of a problem expanding the existing appellate courts would be for intracircuit consistency, and Richman and Reynolds are right to cite to them.108 But it is too strong to state, as the authors do, that this concern is “the great red herring”109 of the expansion debate.
But beyond the problems that one can locate within their counter-counterarguments—that is, the arguments the authors preemptively make against the counterarguments that one could raise against expanding the judiciary—lies a fundamental problem with the main argument itself. Richman and Reynolds suggest that the size of the federal bench should be increased dramatically, and yet they provide no account of how this proposal could ever be realized. According to the authors, the state of the judiciary has not improved to date because of the positions of the relevant actors. Specifically, Congress, at best, appears to have little interest in expanding the bench (indeed, it has not created a single new appellate judgeship since 1990110) and the judiciary actively opposes it. And yet, the solution proposed in Injustice relies on these very same branches, for it requires Congress to create more judgeships and the judiciary to support the action (or at least not actively oppose it).
This inconsistency calls to mind what Eric Posner and Adrian Vermeule have described as a problem of “incentive-compatibility.”111 That is, the diagnosis supplied by Richman and Reynolds relies on an account of the judiciary’s motivations that is in direct contrast with the motivations that presumably would need to be present for the solution to be realized.112 If realpolitik is what has halted proposals to expand the judiciary to date, then there needs to be a compelling account of what has changed or what could change if that same proposal is being put forward as the solution today—an account that the book does not provide.113
One response to this concern is that perhaps the authors are simply taking on the position of reformers, arguing what would be best without concern for constraints.114 And yet, Richman and Reynolds reject at the end of their book what is arguably the second “obvious” solution to the caseload problem—namely, reducing the caseload by limiting jurisdiction—on the ground that this proposal is a “pipe dream.”115 As they write, “Congress makes changes to federal jurisdiction in response to political pressure from constituents and contributors. It is not about to make radical cuts in federal jurisdiction to accommodate the judiciary’s desire to remain small . . . .”116 Here, the authors recognize (and care about) the problems posed by the political reality inherent in some would-be solutions; they simply fail to carry that recognition over to their own.
Bound up in the intellectual consistency problem117 is the feasibility problem itself. The main proposal of Injustice is one that will not be realized in the foreseeable future. It is therefore unsatisfying, if not disheartening, to be left with this proposal as the only way out. The task, then, is to take up the contributions of the first portion of the book and begin to consider other ways to improve appellate review.
As the previous Parts describe, Richman and Reynolds were among the first to thoroughly diagnose the problem facing the federal appellate courts—namely that the workload has become unmanageable, and has caused appellate review of most cases to suffer. But the authors ultimately suggest an impossible cure: massively increasing the size of the appellate bench.
It is worth recognizing that Richman and Reynolds have not been alone in suggesting lofty solutions. Other court scholars and judges writing within the first wave of scholarship on judicial administration similarly suggested grand fixes—either by agreeing that the size of the bench must be increased118 or proposing that the cases coming into courts be reduced, such as by limiting or even eliminating diversity jurisdiction.119 It is not hard to see why such solutions have been put forth time and time again, despite the fact that Congress has consistently resisted them.120 A large-scale problem of the federal judiciary would seem to demand a large-scale solution. Moreover, singular solutions have the benefit of intellectual tidiness; they are both relatively easy to identify and easy to argue for. It is far more difficult to get into the weeds of appellate adjudication, as it were, and carefully determine the combination of smaller steps that must be taken to improve review, and then make a compelling case for that combination.
And yet, in light of the fact that the current problems facing the courts need some kind of solution, a more nuanced approach to improving appellate review is precisely what is called for. In the remainder of this Review, I consider the main challenge for the second wave of court administration scholars—how to ameliorate the current state of the courts—by examining non-argument review and then court practices more generally.
A prime concern regarding non-argument review, as identified by Richman and Reynolds, is that some cases (and really, some litigants) are failing to receive thorough judicial consideration.121 As a result, it is more likely that these cases, as compared to cases that receive what the authors call a traditional review, will have legal and factual errors that go uncorrected or confused points of law that go unclarified. Richman and Reynolds respond to this concern by proposing a solution that would do away with the need for non-argument review altogether. With that option as a non-starter, however, we must reframe the task at hand. Instead of finding a way to eliminate non-argument review, we must find ways of improving it. Specifically, these ways should focus on the source of many of the judgments in non-argument cases, namely the staff attorneys, and the oversight of that source, namely the review of judges.
The vast majority of cases in the federal appellate courts are diverted onto a non-argument track. Specifically, of the approximately 38,000 appeals decided on the merits in Fiscal Year 2012, nearly 28,000 were decided on the briefs alone.122 In this class of cases, staff attorneys generally play two critical roles. First, they serve as the primary gatekeepers in many circuits by reviewing cases and deciding if any should be routed onto the argument calendar and given traditional appellate treatment.123 Second, staff attorneys serve as primary drafters; in most circuits, many orders in non-argument cases are drafted in the first instance by these attorneys.124
Though there has been much discussion of improving appellate review by expanding the bench, there has been virtually no discussion of how to enhance review in this set of nearly 28,000 appeals per year. Based on extensive qualitative research on several of the circuit courts,125 a few key suggestions seem particularly promising."
First, as I have noted elsewhere, while a few of the circuit courts have staff attorneys who specialize in a particular subject matter—say, immigration law—many do not.126 The benefits of specialization are straightforward: as members of courts that rely on specialization point out, by developing expertise in one or a few areas of law, staff attorneys are better able to identify errors in decisions below as well as cases that raise particularly complex issues and thus need to be placed on the argument calendar.127 It would not be difficult for more circuits to encourage specialization in a host of substantive areas128—for example, sentencing, employment law, tax, and habeas appeals. Just which areas would be appropriate for specialization would depend on the circuit’s docket (a circuit that only receives a handful of immigration appeals per year would not need a staff attorney who focuses on immigration law alone). And this is not to suggest that the optimal workload should not include some variety in subject matter so as to remain engaging over time (it might make sense for one staff attorney in a given circuit to specialize in, say, two or three types of appeal). But the larger point remains: given that a considerable number of circuits currently fail to have staff attorneys develop an expertise in a subset of areas of law, increasing the extent of specialization in those circuits could be one promising way to improve the quality of review in non-argument cases.
Second, while some circuits (such as the First) have staff attorneys serve for at least several years, many (such as the Fourth) have attorneys who generally are given only one- or two-year terms.129 Just as specialization can be valuable because it increases one’s expertise, so too can a substantial tenure. Or to make the point somewhat finer, if one has concerns about the kind of initial review performed by a staff attorney with only one or two years of experience, a natural solution is to ensure that more staff attorney positions have lengthier terms or are even career appointments, as in other parts of the Clerk’s Office.
While neither of these proposals is as sweeping as abolishing diversity jurisdiction or significantly expanding the bench, their promise lies in their feasibility. They would not necessarily require any additional funding130 or any action from Congress more generally. These proposals could all be implemented by the judiciary itself and their implementation would be fairly straightforward. Moreover, given that staff attorneys are the first movers in the adjudication of the majority of cases in the federal appellate courts, it is vital to contemplate how to ensure that their review is of the highest quality possible. Focusing on the organization of the staff attorney offices—through both specialization and length of tenure—is a promising place to begin.131
The previous set of proposals focused on the initial source of non-argument review in the form of staff attorneys; I turn now to improving the check on that initial source, the judges. Though staff attorneys make recommendations about which cases should be placed on the argument calendar, and indeed draft the orders in tens of thousands of cases per year, three appellate judges ultimately review and must sign off on those decisions.132 Putting the judge in the role of manager (and not initial drafter) is not an unfamiliar practice. In many chambers across the country, the norm is that law clerks will write the first draft of opinions and the judges will review the work product.133 Though from time to time there has been some unease about the scope of the law clerks’ role,134 the dynamic between law clerks and judges is generally accepted because of the belief that there is considerable oversight by the judges. The concern with the dynamic between the staff attorneys and the judges, by contrast, is that the judges do not carefully review the work product in non-argument cases.135
One way to help ensure a thorough review of these proposed orders is through altering voting practices. Currently, in some of the circuits, judges review proposed orders on their own and then submit their vote in the case (and whether they would like to make any changes to the order), typically via fax or e-mail.136 The norm in these circuits is to conduct the voting process in serial or “round robin” fashion.137 This means that if a particular non-argument panel is assigned twelve cases to decide, the panel will be divided so that each judge—Judge 1, Judge 2, and Judge 3—receives four at the outset. The judges then vote on those cases and pass them on to the next judge (and so Judge 1 passes her voting sheet on to Judge 2, Judge 2 to Judge 3, and Judge 3 to Judge 1), and then the process repeats once more.138 The benefits of this kind of voting system are plain: it is orderly and helps ensure that cases are decided efficiently as each judge may be “nudged along” by the knowledge in most instances that at least one of his or her colleagues has already voted on the case. The problem, of course, is that this kind of practice naturally creates some path dependence. When voting, each judge already knows how one colleague has voted in one-third of all cases and how the other two panel members have voted in another third of all cases. The concern, then, is that if a judge sees that two of her colleagues have already voted to accept the proposed order of the staff attorney, she will then conduct a less thorough review of the order (and the case more generally) than if she were conducting the review without this knowledge.
One could respond to this concern by noting that the non-argument voting procedure is not so different from the voting procedure employed in conference following oral argument. In the latter situation, the judges discuss the case together and announce how they will vote and so again, most judges know how at least one other judge would decide the case before casting a vote. But in argument cases, there is little concern about the thoroughness of judicial review—by and large the expectation is that the judges have already read the briefs and any memoranda prepared by their law clerks, and certainly they have just heard the oral arguments of counsel. In the non-argument cases, there simply is not the same assurance that the judges have spent much time considering the case.
Given this concern, one potentially promising solution is to alter the voting pattern of the judges. As it currently stands, at least one circuit—the Third—purposefully does not utilize serial voting.139 Rather, each judge reviews the case on his or her own and then submits a voting sheet to the presiding judge of that panel.140 According to one Third Circuit judge, the decision to forgo serial voting was deliberate and made so as to minimize the extent to which judges would be influenced by their colleagues when casting votes.141 This process could be taken one step further, with each judge submitting his or her vote to the Clerk’s Office so that even the presiding judge would not be influenced by the other panel members (and then, once all of the judges had voted, the Clerk’s Office could inform the panel of the results). This kind of “blind” voting procedure would be easy to implement across the circuits and would help ensure that the judges are carefully considering each case and reviewing each drafted order.
Again, none of these proposed solutions are overarching; they are, to be sure, somewhat limited in scope. But in this context, this type of limitation should be understood to be a virtue and not a defect.142 It is the limited nature of the proposals—and, attendantly, the fact that they carry no new financial burden and require no action by anyone outside the judiciary—that makes them feasible. As such, they can actually stand to improve appellate review.
Stepping beyond non-argument review, the courts of appeals have implemented a range of practices and procedures to respond to the rising caseload. Some of these practices relate directly to the judges themselves—either by allocating judge hours where they are most needed, through the use of visiting judges, or by increasing judge hours, through practices relating to senior judges. There are also broader practices that relate to everything from mediation to filing deadlines. Given that those practices and procedures seem likely to remain in some form, it is worth theorizing about whether they can be improved as well.
One way in which circuits facing particularly high workloads cope is to import visiting judges from other courts for a short period of time.143 The visitors can be other circuit court judges, district judges, or other Article III judges altogether. As a result, a criminal appeal brought in the Second Circuit could be decided by two Second Circuit judges and a judge from the United States Court of International Trade. Though this practice is not widely discussed in the literature, it is fairly significant; overall, visiting judges participated in nearly 4,000 cases between September 2011 and September 2012.144
Richman and Reynolds raise several concerns about the use of visiting judges—including that their presence may disturb the collegiality of the court and that resident circuit judges may be resistant to having non-resident judges help shape their law.145 The proposal put forth by Richman and Reynolds would largely do away with the need for visitors, as there would simply be more judges for each court. But again, as it is unlikely that such a proposal will be realized in the near future, the key question becomes, how can we improve this practice?
One promising place to begin is in better understanding the dynamic between “home” judges and judges sitting by designation. Interestingly, there are differing views on what kinds of judges make the most useful visitors. For example, some appellate judges have stated a preference for having in-circuit district judges sit by designation, as they are the ones most familiar with circuit law.146 Accordingly, the argument goes, they can be most helpful in identifying an error below or an area of law that would benefit from development through a published opinion. Other appellate judges prefer sitting with visiting circuit judges, on the ground that those judges feel like “colleagues”147 (and in a related vein, some appellate judges have noted that district judges can be overly deferential to the other judges on the panel148). Another reason in favor of outside circuit judges is that they are used to the job, and have the time needed to do the work required of a sitting.149 Still others talk about the benefits of having Article III judges who do not sit on district courts or ordinary courts of appeals, such as judges from the United States Court of International Trade, sit by designation—namely that those judges do not have their own, competing body of law150 and that, given their relatively low workload, these judges are best positioned to accept a writing assignment and therefore serve as real members of the court.151
In light of how many panels include a judge sitting by designation, and in light of the concerns raised by the different classes of these judges, the reliance on visitors is clearly one promising area for research. It would be helpful for court scholars to explore, through qualitative research, just how the dynamics between the various kinds of judges play out in panel decision-making. Furthermore, it would be useful for scholars to assess, through quantitative research, if the rates of affirmance or denial change depending on whether the visiting panel member is a district judge from that circuit or not. If, based on these different methodologies, it became clear that one type of visitor helped contribute to a more thorough review than another, this would be critical information for circuits to have and could lead to improved appellate adjudication.
In addition to visiting judges, the other critical set of judges who can assist with the caseload is to be found within the circuit: senior and senior-eligible judges. There are two ways in which these judges can increase the productivity of the rest of the court. First, by taking senior status as soon as they are eligible, judges free up seats to be filled by newly appointed judges.152 And second, once judges have taken senior status, they can decide to continue working at near or even full capacity.153 It is worth assessing ways of encouraging both.
On the first point, it is important to keep in mind that judges are not obligated to “go senior” at the moment when they become eligible. Rather, they have the option of taking senior status whenever they reach the age and service requirements of the “Rule of Eighty” (meaning that the judge’s age and years of service total at least eighty, and, furthermore, that the judge is at least sixty-five years old and has been a judge for at least ten years).154 Accordingly, judges can—and often do—wait to elect senior service for several years after they are eligible. From a productivity perspective, this is a great loss. By going senior, a judge vacates an active judgeship, which can be filled by someone else.155 If the now-senior judge continues at even only fifty percent capacity, the court has the equivalent of one-and-a-half judges during this time instead of only one. If each year even only a handful of judges took senior status as soon as possible instead of delaying for several years, the benefit would be the equivalent of adding several new judgeships without any need for congressional action.
The question, then, is how to make taking senior status as soon as one is eligible as desirable as possible. Some incentives already exist. Beginning with financial considerations, unlike an active judge, a senior judge can accept payment for approved teaching beyond the standard outside income cap of fifteen percent of annual salary.156 Turning to the job itself, in at least the Second Circuit, it is a frequent practice for a panel to offer a senior judge the choice of one or more opinions to write.157 But there are also quite a few disincentives. A senior judge is not permitted to participate in en banc proceedings unless he or she was part of the panel that originally decided the case or was an active judge when the en banc proceedings began.158 Furthermore, there are logistical and status disadvantages as well. For example, I have been told that in some circuits, taking senior status means giving up current chambers and accepting less desirable chambers or being relegated toward the ends of the bench at court ceremonies.159 It would be valuable to study the practices in different circuits surrounding taking senior status—specifically, whether there are ways of eliminating some of the disincentives, of creating new incentives, or even of simply establishing a new norm of assuming senior status when eligible. There might be some practical constraints—for example, given space constraints in certain courthouses, it might not be possible to have senior judges retain their old chambers—but it would still be worth studying what progress could be made in having more senior-eligible judges take senior status.160
On the second point, it is likewise important to bear in mind that senior judges decide each year how great a workload they will take on.161 A judge could assume, say, a twenty-five percent workload or the full workload of an active judge. Plainly, the more sittings a senior judge volunteers for, the greater the benefit to the rest of the court. It would therefore be valuable to study the policies around treatment of senior judges, and to what extent additional service can be encouraged. In some circuits, for example, senior judges are given great latitude in choosing when they will sit. Specifically, a judge might say she would be happy to sit an additional week but only if that week could be the last in February. The Chief Judge or the Clerk’s or Circuit Executive’s Office then tries to accommodate those requests whenever possible to gain the additional help.162 Court scholars would do well to investigate practices like this one to see if, as an empirical matter, they are effective in encouraging senior judges to assume a greater workload. Additionally, scholarship is needed on the normative component of such policies. Regarding scheduling, if judges are allowed to request certain sitting dates, then the oral argument panels necessarily will not be randomly configured. Is it worth a slight erosion in perfect randomness if such a policy can actually achieve greater productivity? There is a strong argument to be made that it is, but again, court scholars need to further study these questions.
The key point is that assistance from this set of judges is entirely a net benefit.163 By providing for additional judge hours—either by vacating a seat or by assuming a greater workload—senior-eligible and senior judges can contribute significantly to the productivity of the courts. Accordingly, there should be scholarly inquiry into the practices surrounding this set of judges.
Beyond relying on visiting judges, appellate courts engage in a range of practices in an effort to increase their efficiency and thus reduce their backlog of cases. These practices range from the process-oriented, such as requiring that a portion of all cases go through mediation before proceeding to some form of judicial review,164 to the logistical, such as altering the timeframe in which briefs are to be filed.165 Moreover, these practices vary from circuit to circuit, and have varied within each circuit from time to time. For example, the Third Circuit includes appeals filed by pro se litigants in its mediation plan, while other circuits typically do not.166 Turning to filing deadlines, the Second Circuit recently adopted a new practice in which the parties effectively propose their own briefing schedule—a practice that reduces requests for extensions of time (and the administrative burdens attendant on such requests).167
It would be easy to dismiss changes to any individual practice as simply “housekeeping”168 measures, but in toto, they can have a significant impact on the appellate process. Accordingly, the second wave of court scholarship should include study of the process-oriented components of appellate review including, among others, the various mediation practices—just how efficient the different plans are and just how much they lead to (or detract from) the satisfaction of the parties.169 Likewise, there should be further study of the logistical components of appellate procedure—including whether shifts in timing, and in who sets the timing, create net gains, again along the dimensions of efficiency and party satisfaction. To be sure, these are in-the-weeds inquiries; yet they are precisely the inquiries that must be undertaken to improve the quality of appellate review that cases in the federal courts receive.
The story of Injustice on Appeal is one of ever-shrinking resources—the courts of appeals have had to perform the same set of critical functions with fewer and fewer means per appeal to do so. Yet there is another story here as well about the resources of the academy. Legal scholars in general spend a great deal of time devoted to theory and doctrine. And yet, we spend relatively few resources on studying the institutions that make up our legal system, particularly on the twin positive and normative questions about how they actually function and how they should function. Richman and Reynolds’s work serves as a call to arms for the academy to take up these critical inquiries.
Ultimately, Richman and Reynolds have provided a great deal for court scholars following in their wake. They have carefully and thoughtfully delineated the largest problem facing the federal judiciary in the past several decades—one that affects tens of thousands of litigants each year. With the quality of overall judicial review in doubt, it is for the academics to carefully study—using both qualitative and quantitative tools—the use of court practices. From judicial voting rules to visiting judges, from mediation to staff organization, there are numerous areas ripe for academic review about how to improve judicial review. In Injustice on Appeal, Richman and Reynolds have laid the groundwork; it is up to the next generation of court scholars to find the way.