Statement of
Professor Arthur D. Hellman
University of Pittsburgh School of Law

House Committee on the Judiciary
Subcommittee on Courts and
Intellectual Property

Oversight Hearing on the
Final Report of the
Commission on Structural Alternatives
for the Federal Courts of Appeals

July 22, 1999

Executive Summary

1. Although the Commission states that the Ninth Circuit Court of Appeals "should continue to provide the West a single body of federal decisional law," its plan subverts that goal by abandoning circuit-wide stare decisis. This radical step would authorize, if not encourage, the creation of intracircuit conflicts.

2. The proposed "Circuit Division" would do little to preserve uniformity. The Commission's plan places substantial constraints on the Division's authority. In all likelihood, decisions of the Circuit Division would be so infrequent, and their effect on the law of the division so limited, that "the law of the circuit" would shrink to near-insignificance.

3. The Commission plan is thus not a compromise. Those who want to divide "the Ninth Circuit" have never cared about the circuit as such; what they have sought is a division of the court of appeals. And that, for all but a handful of cases, is what the Commission plan would give them.

4. The rationale for the Commission plan is that "the law-declaring function of appellate courts requires groups of judges smaller than the present Ninth Circuit Court of Appeals." But the arguments offered in support of the rationale do not stand up under scrutiny.

5. The Commission insists that judges on a large appellate court cannot adequately monitor other judges' decisions. The flaw is that the Commission lumps together two very different activities: keeping up with circuit law, which is something done by individual judges, and monitoring panel opinions, which is done by the court as an institution.

(a) Judges today need not keep up with circuit law in order to make use of opinions when they are relevant. And reading an opinion today will not help in avoiding a conflict when the judge confronts a similar issue months or years from now.

(b) Effective monitoring does not require that all judges keep up with all opinions. The evidence indicates that Ninth Circuit judges can and do monitor the opinions rendered by their colleagues.

6. The Commission argues that "large appellate units have difficulty developing and maintaining consistent and coherent law." But it disdains empirical research and relies instead on "perceptions" and its own (unspecified) experience. That is far too little to justify the radical restructuring that it proposes.

Mr. Chairman and Members of the Subcommittee:

I appreciate your invitation to express my views at this oversight hearing on the Final Report of the Commission on Structural Alternatives for the Federal Courts of Appeals (White Commission). The principal focus of the Commission report is the largest of the federal judicial circuits, the Ninth. The Commission recommends legislation that would keep the circuit intact but divide the Ninth Circuit Court of Appeals into three "semi-autonomous" adjudicative units.

The Commission's plan gives the appearance of compromise and moderation. But appearances are deceiving. The Commission plan is not a compromise; it gives one side almost everything it wants. And far from being moderate, it embodies a novel approach to federal appellate structure that is flawed both in conception and in execution.

This leap into the unknown might be justified if the Commission had demonstrated the existence of a problem of serious dimensions that could not be dealt with in any other way. On the contrary, in explaining its key conclusion - that "the law-declaring function of appellate courts requires groups of judges smaller than the present Ninth Circuit Court of Appeals" - the Commission offers remarkably little in the way of proof. The Commission simply does not make the case for the radical restructuring that it proposes.

Introduction

Five experiences have shaped my views on S. 253 and the White Commission report. First, from 1973 through 1975 I served as deputy executive director of the Commission on Revision of the Federal Court Appellate System (Hruska Commission). In that capacity I drafted the report that recommended that the Ninth Circuit be divided into two new circuits. (For a discussion of why that recommendation is no longer persuasive, see Hellman, Dividing the Ninth Circuit: An Idea Whose Time Has Not Yet Come, 57 Mont. L. Rev. 261, 264-74 (1996).)

Second, in 1978-79 I was the director of the central legal staff of the Ninth Circuit Court of Appeals. My responsibilities included devising and implementing procedures that would assist the court to do its work more effectively, and in particular to meet the new needs created by the expansion of the court from thirteen to twenty-three active judges.

Third, in the late 1980s I directed a study by fourteen legal scholars and political scientists of the structural and procedural innovations implemented by the Ninth Circuit during the period 1976-1988. The fruits of that study were published by Cornell University Press in 1990; the title of the book is Restructuring Justice: The Innovations of the Ninth Circuit and the Future of the Federal Courts.

Fourth, as stated by the Federal Judicial Center in the report submitted to Congress on Structural and Other Alternatives for the Federal Courts of Appeals, I have conducted "the only systematic study of the operation of precedent in a large circuit." This research has been published in several articles, including Jumboism and Jurisprudence: The Theory and Practice of Precedent in the Large Appellate Court, 56 U. Chi. L. Rev. 541 (1989).

Finally, earlier this year, Chief Judge Hug appointed me to a 10-member Evaluation Committee whose mission is "to examine the existing policies, practices and administrative structure of the Ninth Circuit Court of Appeals, in order to make recommendations to its judges to improve the delivery of justice in the region it serves."

It is an honor to serve on the Evaluation Committee and to work with the Ninth Circuit Court of Appeals in seeking better ways of carrying out the processes of appellate adjudication. However, I do not speak for the court or any other institution; the views expressed here are my own.

I. The Commission's Plan: Contradictions and Conundrums

The Commission offers a plan that would retain the Ninth Circuit but divide its court of appeals into three "semi-autonomous" divisions. The plan contains four elements:

1. Regional jurisdiction over appeals. The present Ninth Circuit Court of Appeals would be reorganized into three "regionally based adjudicative divisions." Each division would hear the appeals filed from that geographical area.

2. Regional assignment of judges. Each division would include seven to eleven court of appeals judges in active status. "A majority of [the] judges serving on each division would be residents of the districts over which that division has jurisdiction, but each division would also include some judges not residing within the division, assigned randomly or by lot for specified terms of at least three years."

3. Regional performance of the law-declaring function. "Each regional division would function as a semi-autonomous decisional unit." This entails two changes from the current arrangement. The circuit-wide en banc process would be abolished; the functions now performed by the Ninth Circuit's en banc court would be performed by en banc courts for each division. More important, divisional decisions - whether by panels or by the en banc court - would be binding only within the division.

4. Conflict resolution by a "Circuit Division." In addition to the three regional divisions, the Commission plan would establish a "Circuit Division … whose sole mission would be to resolve conflicting decisions between the regional divisions." The Circuit Division would be composed of the chief judge of the circuit and twelve active judges - four from each of the regional divisions - who would be selected by lot and who would serve for staggered three-year terms.

The Commission argues that its plan "is the most principled and effective way to resolve the debate about the Ninth Circuit and its court of appeals." (Final Report at 57.) However, analysis of the various elements leads to a very different conclusion. The Commission may be correct in saying that its proposal "addresses the adjudicative concerns that have animated calls to split the circuit." But its confidence that the plan "will achieve the legitimate ends of …those who seek to preserve [the circuit]" is sorely misplaced.

A. Abandonment of circuit-wide stare decisis

The most radical aspect of the Commission's proposal is the abandonment of circuit-wide stare decisis. Today, the Ninth Circuit, like all of the other federal courts of appeals, follows the rule that panel decisions are binding on all subsequent panels unless overruled by the Supreme Court or by the court of appeals en banc. Under the Commission's plan, decisions handed down in one division would be binding only within that division.

If there was any doubt about the Commission's commitment to this element of its plan, it is eliminated by the Commission's response to the comments by Chief Judge Hug on the preliminary draft of the Commission report. Judge Hug, speaking for a majority of the judges of his court, urged the Commission to modify its plan by making panel decisions binding throughout the circuit "unless …overruled by a circuit-wide en banc court." The Commission emphatically rejected this suggestion, stating that this modification "would leave the court of appeals essentially unchanged as an adjudicative body, and would defeat the purpose of the divisional structure that we recommend."

Abandonment of circuit-wide stare decisis would be a logical step if the Commission were recommending that the Ninth Circuit be kept intact solely for administrative purposes and that three separate courts be created within the circuit for adjudication. But that is not the Commission's plan, nor does the Commission reject the premise that the law within the Ninth Circuit should be uniform. On the contrary, the Commission states at the outset that the Ninth Circuit Court of Appeals "should continue to provide the West a single body of federal decisional law." (Final Report at iii.)

How, then, can the Commission propose a regime under which "[d]ecisions made in one division would not bind any other division"? The Commission gives two answers, perhaps three (with the third buried in a footnote).

First, the report contemplates that decisions of other divisions would "be accorded substantial weight as the judges endeavor to keep circuit law consistent." As a prediction of judicial behavior, this is well grounded in experience. Circuit judges today generally respect the decisions of other circuits, and there is no reason to think that judges in a restructured Ninth Circuit would not accord similar weight to decisions of other divisions.

On the other hand, there is a difference between respecting precedent and being obliged to follow it. I have no doubt that judges today often follow precedents they do not like, simply because it is their obligation to do so. If stare decisis did not operate circuit-wide, judges would be free simply to reject precedent from another division. The Commission plan would thus authorize, if not encourage, the creation of intracircuit conflicts.

This brings us to the Commission's second and more important response: the creation of a "Circuit Division." The Commission insists that the Circuit Division - "a small, stable, but still representative subset of the court's judges … focused on conflict resolution" - can insure the maintenance of "desirable circuit-wide uniformity." (Final Report at 51.) This response raises two questions. What does the Commission mean by "desirable circuit-wide uniformity"? And how much uniformity would the Circuit Division bring? To those questions I now turn.

B. Jurisdiction and authority of the Circuit Division

The keystone of the Commission plan is the Circuit Division. Without the Circuit Division, there could be no pretense that the Ninth Circuit Court of Appeals remained intact in anything but name. Each of the regional divisions would be totally autonomous except for the cumbersome process of rotating judges among the regions. Thus, it is essential to understand how the Circuit Division would operate.

The first thing that stands out is the extraordinary constraints the Commission's plan places on the authority of the Circuit Division. The jurisdiction of the Division would be limited to resolving "square" conflicts between the regionally organized divisions. Further, the Circuit Division could not take any case on its own motion; it could act only in response to an application for review filed by a party.

1. Only "square" conflicts

What does the Commission mean by "square" conflicts? One plausible interpretation is that the Commission refers to situations in which one division explicitly refuses to follow a decision handed down in another division. Explicit rejection is the only treatment of circuit precedent now forbidden to court of appeals panels. It would be logical to say that when a panel does take advantage of the freedom conferred by the divisional arrangement, the decision would be subject to review by the Circuit Division to eliminate the disagreement.

Suppose, though, that the panel (or the regional en banc court) distinguishes a decision from another division that reached a contrary result in a similar case. The losing litigant argues that, notwithstanding the purported grounds of distinction, the panel's resolution conflicts with the other division's ruling. Could the Circuit Division find that a square conflict exists and accept the application for review?

If the answer is "yes," that is an invitation to tiresome wrangling over whether two decisions really are in conflict. In this regard, it is instructive to consider the experience of the Florida Supreme Court. That court is vested with jurisdiction to review "any decision of a district court of appeal … that expressly and directly conflicts with a decision of another district court of appeal … on the same question of law." Commentators describe the jurisdiction as "disputatious" and note that "the existence of conflict often is not so certain, meaning that a brief [seeking review] must engage in a lengthier and more convoluted argument to establish the Court's discretion to hear the case." See Gerald Kogan & Robert Craig Waters, The Operation and Jurisdiction of the Florida Supreme Court, 18 Nova L. Rev. 1151, 1225, 1238 (1994). That is hardly a model to be emulated.

What makes the arrangement even more problematic in the Ninth Circuit context is that the judges of the Circuit Division would be questioning the good faith or competence of their own colleagues. If the Circuit Division agrees to review a decision that has distinguished an opinion handed down by another regional division, that would be tantamount to saying that the later panel has failed to recognize that the earlier opinion involved the same issue and required the same result. I suspect that the Circuit Division judges, taking into account the effect of such a declaration on collegiality within the circuit and on the legitimacy of the system, would be reluctant to take that step.

These considerations suggest that the jurisdiction of the Circuit Division would be limited to acknowledged conflicts - conflicts created by the explicit refusal of one regional division to follow the precedent established by another division. That limitation, however, would substantially undercut the effectiveness of the mechanism. Indeed, the Circuit Division would be far less able than the existing limited en banc court to maintain uniformity within the circuit - a mechanism that the White Commission finds wanting.

Under the existing arrangement, Ninth Circuit judges can and do grant rehearing en banc to resolve tensions in circuit law caused by inconsistencies in doctrines or outcomes less blatant than explicit rejection. See, e.g., Hale v. Arizona, 993 F.2d 1387, 1389 (9th Cir. 1993) (en banc) ("We consider these questions en banc to resolve the tension between [two panel decisions]."); Townsend v. Holman Consulting Corp., 929 F.2d 1358, 1365 (9th Cir. 1990) (en banc) (overruling panel decision, thus obviating need to maintain "unstable and awkward" distinction drawn by later case).

When the Commission issued its draft report in October 1998, it was unclear whether the narrower or broader interpretation of the Circuit Division's jurisdiction was intended. The Final Report appears to endorse the narrower reading. In explaining how the arrangement it proposes "will ensure clearer, more consistent circuit law," the Commission states that "conflicts …between divisions will be more sharply highlighted," and that the Circuit Division will "choose between articulated conflicting points of view." (Final Report at 49; emphasis added.) This language implies that the Circuit Division would be limited to cases in which a panel explicitly rejected the "point of view" adopted by one of the other divisions. As long as the panel found grounds of distinction - even "unstable and awkward" grounds - the Circuit Division would stay its hand.

This interpretation is confirmed by the testimony of Judge Pamela Ann Rymer, a member of the Commission, at a hearing of a Senate Judiciary subcommittee on July 16, 1999. Speaking on behalf of the Commission, Judge Rymer said that "it will be the rare case that qualifies" for review by the Circuit Division. She added: "Inconsistency alone is not sufficient for Circuit Division review. There must be square and significant conflict." (Emphasis added.)

2. Only upon litigant request

The authority of the Circuit Division would be further constrained by the Commission's insistence that the jurisdiction of the Division could be invoked only by a party to a case - and "only after the panel decision had been reviewed by the division en banc or a divisional en banc had been sought and denied." Here, too, the Commission plan casts aside one of the mechanisms used by the Ninth Circuit today to maintain uniformity: the sua sponte panel-initiated en banc call.

Recent decisions illustrate the utility of this procedure. In 1998, the court took a group of cases en banc sua sponte "to rethink our previous decisions" on the preemption of state tort claims by the Airline Deregulation Act. Charas v. Trans World Airlines, Inc., 160 F.3d 1259 (9th Cir. 1998). The en banc opinion explained, "Because of the need to clarify the law in this area, these cases were taken en banc after they were assigned to a three-judge panel, but prior to the panel's rendering a decision." The en banc court issued a unanimous opinion overruling two panel decisions and establishing the law for the entire circuit. This process would not have been possible under the Commission's plan. More recently, the court accepted a panel's sua sponte en banc call to resolve "an irreconcilable conflict in this circuit's case law regarding the standard of review for rulings on the prosecution's use of peremptory challenges." Tolbert v. Gomez, -- F.3d - (9th Cir. 1999) (No. 97-55004). The court eliminated the inconsistency without waiting for a litigant's request and without waiting for the panel to issue an opinion.

C. The Commission's narrow vision of uniformity

Supporters of the Commission plan are caught on the horns of a dilemma. If the Circuit Division can review decisions even when the regional panel insists that the other division's ruling is distinguishable, it opens the door to time-consuming and uncollegial disputation over whether the new case creates a "square" conflict. But if the Circuit Division is limited to hearing cases in which one division has explicitly rejected another division's precedent, it will be powerless to eliminate less blatant inconsistencies of the kind that arouse concern today.

Is there any escape from this quandary? The White Commission gives what it may regard as a partial answer. In a little-noticed footnote - not included in the October 1998 draft report - the Commission reveals that its vision of "uniformity" is a narrow one. The Commission's text refers to "conflicts on issues for which circuit-wide (or state-wide) uniformity is important." (Emphasis added.) The footnote explains:

[W]e envision that [the function of the Circuit Division] will be focused on maintaining uniformity on issues of law that matter to the entire circuit or to a state (such as California) that is in more than one division. For example, it would be highly undesirable if the Northern and Southern Divisions established different rules on an admiralty issue. On the other hand, it would not appear to matter whether all divisions had the same rule of law with respect to the factors to be considered in granting an adjustment for abuse of trust under the Sentencing Guidelines. (Final Report at 44 n.99.)

Although the Commission does not generalize from its two examples, this passage implicitly draws a distinction emphasized by the Federal Courts Study Committee in its analysis of conflicts between circuits. The Study Committee recognized that not all intercircuit conflicts are "intolerable," and it posited that one criterion for identifying "intolerable" conflicts is that they "impose economic costs or other harm to multi-circuit actors." The White Commission's examples suggest that it draws the line in the same way.

In the aftermath of the Study Committee report (and at the request of Congress), I conducted a study of unresolved conflicts between federal judicial circuits. The study concluded that, more often than not, unresolved conflicts do not pose a serious threat to the activities of multi-circuit actors. Indeed, on many issues the subject matter alone virtually forecloses any effect on multi-circuit actors. This is true of sentencing issues, as suggested by the White Commission; it is also true of most civil rights issues and most issues involving the elements of federal crimes. In disclaiming the importance of circuit-wide uniformity on these issues, the Commission is implicitly telling us that the Circuit Division need not resolve even "square" conflicts in large and important areas of federal law.

Two other aspects of the distinction also warrant mention. First, "square" conflicts on issues affecting multi-circuit actors are probably less common than "square" conflicts on issues such as the interpretation of federal criminal statutes or sentencing guidelines. Second, the concerns that underlie the desire for uniformity between divisions on matters of admiralty law and other issues affecting multi-circuit actors apply equally to uniformity between circuits. For that reason, these concerns often guide the Supreme Court in the exercise of its certiorari jurisdiction. If the Circuit Division is confined to resolving "square" conflicts on issues affecting multi-circuit actors, it will have little to do, and that little may well be overtaken in short order by Supreme Court decisions.

D. The shriveled "law of the circuit"

In sum, there is less to the Circuit Division than meets the eye. The Circuit Division would resolve only "square" conflicts - a category apparently limited to cases in which one division has explicitly rejected another's precedent. It would act only upon the request of a party, and it would probably limit itself to issues that affect the operations of multi-circuit actors - a circumstance that is the exception rather than the rule. In all other respects, the law in each division would be left to develop separately.

One other element of the Commission plan comes into play here. The Commission contemplates that after granting review, "the Circuit Division will simply resolve the issue in conflict, and return the case to the regional division for such other proceedings as are necessary." (Final Report at 46; emphasis added.) This too suggests a narrow view of the Circuit Division's field of operation, and it reinforces the supposition that the Circuit Division would confine itself to discrete issues on which there is an explicit disagreement.

What would the consequences of this arrangement be? I believe that, before very long, the three divisions would be carrying out their law-declaring functions almost as separate courts. Decisions of the Circuit Division would be so infrequent, and their effect on the law of the division so limited, that "the law of the circuit" would shrink to near-insignificance.

E. Isolation of the divisions

The scenario I have described is made even more likely by the probable fate of another element of the Commission's plan, the long-term random rotation of judges among the divisions. Here is what the Commission has to say about the rotation feature in its report:

A majority of judges serving on each division would be residents of the districts over which that division has jurisdiction, but each division would also include some judges not residing within the division, assigned randomly or by lot for specified terms of at least three years. (Final Report at 43.)

The draft statute is somewhat more open-ended:

A majority of the judges assigned to each division shall reside within the judicial districts that are within the division's jurisdiction …; provided, however, that judges may be assigned to serve for specified, staggered terms of three years or more, in a division in which they do not reside. Such judges shall be assigned at random, by means determined by the court, in such numbers as necessary to enable the divisions to function effectively. (Final Report at 94.)

Even here, there is some ambivalence about long-term cross-division assignment of judges. (Compare "would" in the report text with "may" in the draft statute.) And when Senator Murkowski (joined by Senator Gorton) introduced the legislation implementing the Commission proposal, he offered the "strong suggestion" that the Senate Judiciary Committee eliminate the rotation requirement altogether.

I believe that if the Commission plan were to be enacted into law, the Murkowski view would prevail. I say this because there is simply no constituency for the long-term random rotation of judges among divisions. The northwestern senators - who until now have been the most ardent advocates of splitting the circuit - have already made clear their opposition to this feature. And the circuit judges, most of whom do not want any division of the circuit or the court, would be equally opposed to long-term cross-division assignment. A judge living in Alaska would hardly relish the prospect of flying to Pasadena or Phoenix for every argument calendar for three long years. A judge from Los Angeles would not want to hear all of his or her cases in the northwest.

I am not suggesting that judges would hear cases only in their own region. On the contrary, short-term cross-division assignment of judges would certainly be a feature of the arrangement, if only because caseloads will seldom be proportional to the number of judges residing in each of the regions. But that is little different from current use of, for example, district judges and senior judges from other circuits. The judges regularly sitting in each division would be the judges who reside there.

F. Conclusion: the compromise that isn't

What happens when you put all of this together? In all likelihood, the result would be something like this. In each division, cases would be adjudicated largely by a self-contained group of judges bound only by the precedents they themselves have handed down. The Circuit Division would intervene to provide circuit-wide law only on the rare occasions when a panel or en banc court in one division has explicitly rejected another division's precedent on an issue that affects multi-circuit actors. In many - perhaps most - areas of the law, each division would develop its own line of precedent. The "law of the circuit" would become almost an irrelevance.

This analysis explains why the Commission plan is not a compromise. Those who want to divide "the Ninth Circuit" have never cared about the circuit as such. It is a matter of indifference to them whether the circuit council, the Bankruptcy Appellate Panel, the circuit conference, and other circuit institutions remain as they are. What they have sought is a division of the court of appeals. And that, for all but a handful of cases, is what the Commission plan would give them.

II. The Commission's Faulty Diagnosis

Notwithstanding its flaws and limitations, the divisional structure plan might be worth pursuing if the Commission had identified a serious problem in the Ninth Circuit Court of Appeals that could be solved only through reliance on smaller adjudicative units. But on the evidence of the Commission report, no such problem exists.

The rationale for the Commission plan is that "the law-declaring function of appellate courts requires groups of judges smaller than the present Ninth Circuit Court of Appeals." This rationale rests in turn on two overlapping arguments. First, judges in a large appellate court are unable "to monitor all the decisions the entire court of appeals renders." Second, "large appellate units have difficulty developing and maintaining consistent and coherent law." Neither argument stands up under scrutiny.

A. Monitoring of panel opinions

Central to the Commission's vision of effective appellate adjudication is the "monitoring" of panel opinions by other judges of the court. The Commission puts it this way:

Courts of appeals rely on their judges to monitor the decisions of all panels of the court so that their own decisions are consistent with earlier decisions of the court and so that the court can identify and correct any misapplication or misstatements of the law. …The volume of opinions produced by the Ninth Circuit's Court of Appeals and the judges' overall workload combine to make it impossible for all the court's judges to read all the court's published opinions when they are issued. (Final Report at 47.)

For several reasons, the Commission's reliance on this theory is misplaced.

First, as Chief Judge Hug and his colleagues have aptly stated, the assumption that judges cannot keep sufficiently abreast of circuit law without reviewing opinions as they come out "is a relic of the pre-computer era." Before computers, opinions would not appear in the advance sheets for weeks or months; digests, citators, and other research tools lagged even further behind. On a large court, the only way a judge could avoid an inadvertent conflict with another panel's decision was to read opinions as they came out, sort them into piles by subject matter, and perhaps keep a personal index of important rulings.

Today, conditions are very different. If a judge is considering a case involving NEPA or FOIA or Miranda or Noerr or any other issue, all of the court's decisions on point, no matter how recent, can be accessed in seconds through Westlaw and Lexis. In addition, the Ninth Circuit has its own computerized case inventory tools. A judge may scan newly filed opinions simply to get a sense of what is going on in the court, but to collect cases in an effort to replicate the computerized databases would be a waste of time.

Second, the Commission lumps together two very different activities: keeping up with circuit law and monitoring panel opinions. Keeping up with circuit law is something done by individual judges; it is an activity that looks to the future. With all circuit law now easily retrievable by computer when it is needed, there is no particular reason for individual judges to acquire familiarity with decisions that have no relevance for any of their current cases. And reading an opinion today will not help in avoiding a conflict when, months or years from now, the judge does confront a case presenting a similar issue.

Monitoring panel opinions, in contrast, is something that the court does as an institution. The purpose of monitoring, as the Commission suggests, is to identify panel decisions that conflict with earlier decisions of the court or that misstate the law. But effective monitoring does not require that all judges keep up with all opinions. As long as each opinion receives some scrutiny by off-panel judges, the objectives can be met.

Third, the Commission goes off track by referring to "[t]he volume of opinions produced by the Ninth Circuit's Court of Appeals." (Emphasis added.) What the Commission fails to mention is that the volume of published opinions does not correlate with circuit size. In 1998, three other circuits produced a larger number of published opinions than did the Ninth Circuit.

(The analysis is limited to published opinions because only published opinions contribute to the law of the circuit. Also, I recognize that 1998 may have been aberrational for the Ninth Circuit, in that the court's output of published opinions was probably reduced by its high vacancy rate. However, it is not uncommon for other circuits to approach or exceed the output of the Ninth Circuit.)

One would think that, other things being equal, an annual output of 800 opinions could be monitored more easily by 28 judges than by 14. Opinions are not fungible, and neither are judges. The larger the number of judges engaged in the monitoring process, the greater the likelihood that a particular error or inconsistency will catch the eye of at least one member of the court.

Finally, the evidence leaves no doubt that the judges of the Ninth Circuit Court of Appeals engage in a substantial amount of opinion monitoring. In the four-year period ending in 1997, there were more than 300 cases in which an off-panel judge initiated en banc activity. (This figure includes only cases in which the off-panel judge formally invoked the en banc procedures of the court's General Orders. It does not include cases - perhaps quite numerous - in which the off-panel judge communicated only with the panel members.) Even when the court did not vote on an en banc call, the off-panel judge's comments often resulted in modification of the panel opinion and sometimes in a modification of the disposition.

In this light, the Commission's concerns about the supposed difficulties of opinion monitoring in the "large appellate unit" ring hollow. Judges today need not read opinions as they come out in order to make use of them when they are relevant. As for monitoring, the evidence indicates that the judges of the Ninth Circuit can and do monitor the opinions rendered by their colleagues.

B. Maintaining Coherent and Consistent Law

Monitoring, of course, is not an end in itself, but a means to an end. The Commission's principal argument is that "large appellate units have difficulty developing and maintaining consistent and coherent law." (Final Report at 47.) The Commission thus aligns itself with those who believe that inconsistencies in panel decisions are more common in the Ninth Circuit than in other circuits.

What is the basis for this conclusion, so critical to the Commission's recommendation? The Commission refers to "perceptions" of inconsistency and to its own "judgment, based on experience." The "experience" is not specified or described. This is a remarkably weak foundation on which to build so substantial a structure.

The Commission acknowledges "the literature on [the] subject," including my own empirical studies of inconsistency in the Ninth Circuit. The Commission's only response is to say that consistency and predictability cannot be "reduce[d] … to statistical analysis" because the "concepts are too subtle, the decline in quality too incremental, and the effects of size too difficult to isolate, to allow evaluation in a freeze-framed moment." (Final Report at 39-40 & n.39.)

It is the Commission's prerogative to reject the methods or conclusions of empirical research, but it is regrettable that the Commission simply gives up and declares that the concepts are too subtle to warrant analysis. For example, what does the Commission mean by "evaluation in a freeze-framed moment"? The research I conducted, and which the Commission cites, embraced two distinct years of the Ninth Circuit's work, and the evaluation involved decisions rendered over a much longer period of time.

Although the Commission is not willing to credit systematic empirical research, it is willing to rely on "perceptions." The reference to "perceptions" apparently incorporates the brief account earlier in the report of the Commission's survey of district judges and lawyers in the Ninth Circuit and nationwide. The survey is a valuable - indeed unique - source of information, and happily the Commission has made available a complete account of the findings in its Working Papers. Those findings raise some doubts about the conclusions drawn by the Commission.

Consider one of the specific points cited by the Commission in its report:

Ninth Circuit lawyers, more often than others, reported as a "large" or "grave" problem the difficulty of discerning circuit law due to conflicting precedents, and the unpredictability of appellate results until the panel's identity is known. (Final Report at 40.)

When we look at the corresponding table in the Working Papers, we find that, indeed, Ninth Circuit lawyers were more likely than lawyers in other regional circuits to have experienced problems in "discerning circuit law due to conflicting precedents." (Working Papers at 86, Item 20g.) But two other points also stand out:

· The Ninth Circuit lawyers who viewed the problem as "large" or "grave" constituted only one-quarter of the respondents.

· The highest proportion of lawyers giving this response came not from the Ninth Circuit, but from the Federal Circuit - a court of 12 judges, all of whom have their chambers in the same building.

A similar pattern can be seen in the responses to the question "how big a problem is the unpredictability of results until the panel's identity is known?" (Working Papers at 87, Item 20j.) Ninth Circuit lawyers were more likely to have experienced problems than lawyers in other regional circuits, but so were lawyers practicing before the Federal Circuit. Interestingly, one out of seven lawyers experienced a "large" or "grave" problem of unpredictability in the First Circuit, which has only six judgeships and enjoys a reputation for collegiality. (The Commission, in explaining what it means by "collegiality," quotes at length from a book by the former chief judge of the First Circuit.)

These findings point to the need for caution in interpreting the survey results. The question is not whether particular phenomena are associated with the Ninth Circuit Court of Appeals, but whether those phenomena are causally linked to circuit size. On this score, a recent news story about the Court of Appeals for the Federal Circuit provides a useful perspective. (National Law Journal, Aug. 3, 1998, at A-1.) The story notes that some members of the intellectual property bar "accuse the specialized court of unpredictability, claiming that judges are deeply divided on basic patent doctrine, [and] that results are often panel-dependent." The story elaborates:

This factionalism leads to a crap-shoot mentality among lawyers who say the outcome of their cases depends too heavily on who sits on a particular panel. Because the U.S. Supreme Court rarely reviews patent cases, the panels' inconsistent rulings remain unresolved. … Some say the court should take more cases en banc.

To anyone who has followed the debate over dividing the Ninth Circuit, these comments will sound uncannily familiar. They are precisely the kinds of comments that give rise to the "perceptions" that the Commission relies on. Yet no one would argue that the Court of Appeals for the Federal Circuit is too large and should be divided into smaller adjudicative units.

I do not know whether the criticisms of the Federal Circuit are justified. Nor would I want the Ninth Circuit to view the survey findings with complacency. I do suggest that the "perception" evidence drawn from the survey offers little support for the Commission's conclusion that "large appellate units have difficulty developing and maintaining consistent and coherent law."

Finally, there is (to borrow a favorite allusion of Chief Justice Rehnquist) the evidence of the dog that did not bark in the night-time. If inconsistency is as much of a problem as the Commission believes it is, examples should be easy to find. The Commission compiled a voluminous record of testimony and statements dealing with the Ninth Circuit, yet not a single witness came forward with examples - systematic or even anecdotal - of conflicts between Ninth Circuit panel decisions. It is not even clear what kinds of conflicts the Commission has in mind - whether it believes that panels are ignoring relevant precedents, or that panels are drawing unpersuasive distinctions, or some combination of the two.

The absence of examples and the lack of specificity are emblematic of the flimsy evidentiary support that underlies the Commission's plan. At most, the Commission has shown that there is some dissatisfaction with the Ninth Circuit Court of Appeals' performance of its law-declaring function. The Commission has not demonstrated the existence of problems that would be cured by dividing the court into three largely autonomous decisional units.

III. Conclusion

The Commission's proposal for regionally based adjudicative divisions reflects a conscientious attempt to respond to criticisms of the Ninth Circuit Court of Appeals "while preserving [an] administrative structure that no one has seriously challenged." Unfortunately, the plan is flawed both in conception and in execution. It is unlikely to accomplish its goals, and it has the capacity to produce much mischief. I urge the Committee to reject the proposal and to allow the Ninth Circuit Court of Appeals to continue its course of productive experimentation "to improve the delivery of justice in the region it serves."

Professor Arthur D. Hellman

University of Pittsburgh School of Law

Pittsburgh, PA 15260

Phone: (412) 648-1340

Fax: (412) 648-2649

E-mail: Hellman@law.pitt.edu

Curriculum Vitae

Arthur D. Hellman is a professor at the University of Pittsburgh School of Law, where he teaches courses in Federal Courts, Civil Procedure, and Constitutional Law. As a scholar, Professor Hellman is best known for a unique series of empirical studies that describe and analyze the actual operation of precedent in the Supreme Court and the federal courts of appeals. These studies have been relied on by policymakers in Washington and elsewhere. His most recent work is Light on a Darkling Plain: Intercircuit Conflicts in the Perspective of Time and Experience, 1998 Supreme Court Review 247 (1999); and Precedent, Predictability, and Federal Appellate Structure, 60 U. Pitt. L. Rev. No. 4 (forthcoming 1999).

Prior to joining the faculty at the University of Pittsburgh, Professor Hellman served as Deputy Executive Director of the Commission on Revision of the Federal Court Appellate System (Hruska Commission). He directed the Commission's research projects and drafted reports to Congress on the realignment of the federal judicial circuits and on the structure and internal procedures of the courts of appeals.

In 1977-79, Professor Hellman took a leave of absence to serve as the director of the Ninth Circuit's central legal staff. He devised the structure of the office and established its procedures. His article, Central Staff in Appellate Courts: The Experience of the Ninth Circuit, 68 Calif. L. Rev. 937 (1980), is a widely cited authority on the effective use of central legal staff and on internal appellate processes generally.

In the late 1980s Professor Hellman directed a study by fourteen leading legal scholars and political scientists of the structural and procedural innovations implemented by the Ninth Circuit during the period 1976-1988. The fruits of that study were published by Cornell University Press in Restructuring Justice: The Innovations of the Ninth Circuit and the Future of the Federal Courts (1990).

In 1999, Professor Hellman was appointed by Chief Judge Hug to a 10-member Evaluation Committee whose mission is "to examine the existing policies, practices and administrative structure of the Ninth Circuit Court of Appeals, in order to make recommendations to its judges to improve the delivery of justice in the region it serves."

Professor Hellman received his B.A., magna cum laude, from Harvard University in 1963 and his J.D. from Yale Law School in 1966. He has been a member of the faculty at the University of Pittsburgh School of Law since 1975.

Disclosure statement: I have not received any federal grant funds during the current fiscal year. In the preceding fiscal year I received a contract payment of $10,000 from the Federal Judicial Center for research on federal appellate structure.