I now have had an opportunity to carefully study the Report and to receive memoranda and comments from many of the judges and lawyers, and I thought it would be helpful to give my evaluation of the Report. The Report recommends that the Ninth Circuit not be split and does acknowledge some important goals and objectives.

Some conclusions by the Commission are particularly important:

The Commission also recommends some experimental innovations that could reduce the burden on the courts of appeals, such as district court appellate panels, two-judge circuit court of appeals panels, possible discretionary review of certain cases, and the possible transfer of certain appeals to the Federal Circuit. The Commission also pointed out the additional burden on the courts of appeals if the Bankruptcy Commission=s recommendation for direct appeals from the bankruptcy courts to the courts of appeals were adopted, and recommended legislation to preserve the participation of the Bankruptcy Appellate Panels (ABAP@).

The difficulty I find is with the recommendation that Congress enact a statute requiring the Ninth Circuit to adopt very specific adjudicatory divisions, composed in very specific ways, with significant changes in stare decisis rules and en banc functions, rather than letting the circuit have the flexibility to achieve the objectives identified by the Commission in less disruptive ways. The Ninth Circuit has always been willing to reexamine its procedures and to experiment in working out better methods of operation. The Commission has identified the Ninth Circuit as an experimental model for how a large circuit can best operate. Allowing the Ninth Circuit to work out the objectives of the Commission with flexibility, rather than a rigid statutory divisional structure, would provide the best model for the future. It also would not saddle the Ninth Circuit during the seven-year experimental period with a structure that gives up a proven successful process for intangible benefits that could be better achieved in less disruptive ways.

Although the Commission does not recommend a split of the Ninth Circuit for administrative purposes, its draft recommendation goes very far in splitting the more-vital adjudicative functions to be performed within the circuit. The present recommendation is that Congress enact a statute, organizing the Ninth Circuit Court of Appeals into three regionally based adjudicative divisions, each division with a majority of its judges resident in its region but providing that judges from other regions will be assigned into the regional division. Each division is to have its exclusive jurisdiction over appeals from judicial districts within that division. Each division would function as a virtually autonomous decisional unit. There is no unifying circuit-wide en banc structure to maintain uniformity of the circuit decisions or to resolve questions of exceptional importance, as has been the traditional function of an en banc court under Rule 35 of the Federal Rules of Appellate Procedure. Instead, there are divisional en bancs, which will finally resolve for the division, all issues that do not conflict with another division=s cases. Any further review of those determinations would be by the United States Supreme Court. The unifying adjudicative structure is designated as the Circuit Division, which is designed to resolve only conflicts between divisions. This Circuit Division panel is composed of seven judges, the Chief Judge, the presiding judge from each division, and one other judge from each division. Thus, the traditional unifying function of an en banc court does not take place unless a conflict has been generated between the divisions. This unnecessarily complicates the adjudicative process, significantly deters the rational development of circuit law, and presents particular problems for the State of California, which is divided into two divisions.

The motivation for this adjudicative structure seems to be to accomplish three major objectives:

1. To have a greater territorial connection between the judges deciding the appeals in the territory from which the appeals come.
2. To allow for a smaller group of judges to work together as a court in developing the law applicable to a particular area of the country.
3. To have more careful monitoring of the opinions of panels within the division, particularly to adopt the pre-filing review procedures that are used in some of the smaller circuits around the country.

In my view, these objectives could be accomplished, to a large extent, in much less intrusive ways, which would not impinge upon the development and maintenance of a consistent body of law throughout the circuit, and without creating another tier of review with the attendant delay and expense that would be involved.

This virtually autonomous divisional structure was adopted despite the overwhelming support of the judges and lawyers within our Ninth Circuit for the adjudicative structure that we have in place. This approval is confirmed by the survey of the judges of our court with 70% of the circuit judges and 68% of the district judges opposing restructuring of the circuit. There was virtually no advocacy for divisions within the Ninth Circuit, although suggestions were made by Chief District Judge Terry Hatter and attorney Sanford Svetcov about possible divisions. Neither of their suggestions would operate at all in the fashion suggested by the Commission Report.

One might ask how the Commission arrived at this particular division of the circuit into these virtually autonomous divisional units. I think it is instructive to review the report of Professor John Oakley, who was asked to prepare a paper on this subject for the Commission. It discusses the California and other state appellate systems. Sections III and IV of that report discuss recent proposals for divisional organization of the Ninth Circuit. The paper discusses the proposals of Judge Weis, Sanford Svetcov, Professor Meador, and Professor Oakley.

These proposals differ significantly from the one adopted by the Commission, as I later discuss. The major differences are: they propose a circuit-wide stare decisis, an en banc process that deals with both conflict and error correction, and assignment of out-of-division judges to division panels, rather than long-term assignments as members of the division.

I think it is important to note that Professor Oakley, after reviewing the other proposals, states in his Concluding Proposal the following:

Thus, in Professor Oakley=s view, this divisional structure of the court of appeals is not necessary because of any deficiency in the operation of the Ninth Circuit, but rather is a experimental model that could be tried by the Ninth Circuit as a possible structure for it and other circuits to try as the case load and need for other judges grow. He also expresses the view that this might encourage other circuits that are taking unacceptable case loads per judge to avoid their fear of asking for additional judges because the court would be too large.

From my viewpoint, the Ninth Circuit has been conducting an experiment with a large court, which Professor Oakley acknowledges has been successful, with the use of the limited en banc process. The Ninth Circuit has never been adverse to new ideas or new approaches to more effective case management or to special concerns with its operational procedures. Our view and the view of a great majority of the judges and lawyers in our circuit is that the experiment, as we have developed it, has worked very successfully. However, because of the criticisms and perceived problems that are sought to be remedied by this divisional approach, I think that, we as the Ninth Circuit, should attempt to address the Commission=s suggested improvements. If we are to be an experimental model for the country, we should be allowed much greater flexibility in determining how best a large circuit can operate. Major motivations for this divisional approach are as follows and their objectives could be accommodated as I indicate.

1. Appeals from a Particular Region to be Heard Principally by Judges from that Region. This could be accommodated by panels in a division being constituted with two judges from the region and one judge from without. This would involve the great majority of the decisions that affect the area. The small percentage of cases that are taken en banc, as I later propose, would be before a limited en banc panel that would have equal representation on the panel from the divisions.
2. More Careful Monitoring of Panel Decisions, Perhaps Before They are Filed. This could easily be accomplished by having judges of a particular division entrusted with the special responsibility of monitoring more carefully the decisions in their own division. However, I propose that any judge from any division would be able to call for a circuit-wide en banc. The Automation Committee of our circuit has developed a particular technology that will permit judges to have a Pre-circulation Report of decisions that have been sent for filing before they are actually filed. If it is deemed desirable, the five-day period between the opinion being sent to the Clerk for filing and the ultimate filing date of the opinion could be extended by a few days.

1. Improvement of the Limited En Banc Process. There has been expression that the number of judges on our limited en banc court is too small. This number could be increased to 13 or 15 judges. If it were to be 13 judges, it could be 4 judges from each division, with the Chief Judge presiding. If it were to be 15 judges, it could be 5 judges from each division, with the presiding officer to be the senior active judge. Other suggestions have been made for improving the en banc process that could be adopted. For example, some judges have suggested frustration with having been the judge that did a great deal of work in calling for the en banc, but not being selected on the en banc court. Judge Hawkins suggests that the author of the panel opinion and the person calling for the en banc automatically be included on the panel. One of the criticisms has been that we do not have enough en bancs. I question whether this is a legitimate criticism, in that proportionately, we have as many en bancs as other circuits. In 1996-97, we had 19 en bancs and 40 calls for en banc on which we took votes. However, we might want to consider adopting the Justice Department=s proposal that we have a requirement for less than a majority of the court to require an en banc review for cases where we are creating or there exists an intercircuit split. Perhaps there should be less than a majority for other types of situations.

As I indicate later, I would recommend to the court that we commence implementing these suggestions now.

Strengths And Weaknesses

1. To decide some selected cases with two-judge panels.
2. To authorize the Judicial Council to establish district court appellate panels to provide the first-level review of the designated categories of cases with discretionary review to the court of appeals.
3. Also, the Commission made a very important recommendation that the direct appeal of bankruptcy cases to the circuit courts, as proposed by the National Bankruptcy Commission, not be adopted without more careful consideration. It proposes three methods by which our Bankruptcy Appellate Panels (ABAP@) would still be involved in the appellate process. This is vitally important because if the direct appeal were adopted without the involvement of the BAP, we would have an additional 1,000 cases in our court of appeals because that number of cases were finally resolved by either the district courts or the BAP decisions.

The major problems that I see with the specific proposal of the divisional adjudicative structure of the Ninth Circuit Court of Appeals are as follows:

1. Under the Commission proposal, the decisions of a panel or an en banc panel in a division are not of stare decisis value to panels in other divisions or to the district courts in other divisions. This means that a decision of the Northern Division has exactly the same standing before the Southern Division as does a decision from the Second Circuit. This differs from Professor Oakley=s proposal, Judge Weis=s proposal, and Sanford Svetcov >s proposal, and does not contribute to the development of circuit law. Stare decisis among panels of a court is essential to the maintenance and development of the law of a circuit.
2. There is no circuit-wide en banc process that can be invoked unless there is a conflict between the divisions. Thus, unless there is a conflict with another division, the en banc decision of the division would be reviewed only by certiorari to the United States Supreme Court. This would be a particular problem between the Middle and Southern Divisions, which divide the State of California. An example would be one of the many ballot issues that the State of California passes. If Proposition 200 were declared to be unconstitutional by the Middle Division and if certiorari were denied by the United States Supreme Court, that would be the rule for the Northern Division of California. Unless a conflict were to arise in the Southern Division by a separate lawsuit being brought, that ruling of the Middle Division would remain intact for the Middle Division, but not for the Southern Division. If there were a conflict, this would be resolved, not between the two divisions that divide California, but by a seven-judge court with a very limited representation from the full circuit.
3. The Commission draws a bright-line between error correction and circuit conflict. This could lead to considerable litigation as to whether it is a true conflict or merely error correction. None of the proposals submitted to the Commission by Judge Weis, Sanford Svetcov, Professor Meador, or Professor Oakley made this distinction between error correction and conflict. I consider it crucial that there be a circuit-wide en banc court, although it could be revised in composition and its procedures could be changed. However, it is essential that the en banc court maintain the law of the circuit whether it is considered a conflict or error correction. (I have difficulty with the concept of error correction because it seems to me that a panel decision interpreting a rule of law or a fact situation may not necessarily be error just because a higher court interprets the facts or the law in a different manner. It is just that the rule of the higher court is the final rule.)
4. The composition of the divisional units, with the majority of the judges being resident judges but other judges, not from that division, being assigned to the division for at least a year causes unnecessary problems without any real benefit. For example, Judge Thomas and Judge Kleinfeld could be assigned to the Southern Division, and Judge Fernandez and Judge Reinhardt could be assigned to the Northern Division because the designation is by lot or by random. Under the Commission proposal, the out-of-division judges would be precluded from taking part in the en banc processes in their own residential area. A better way to accomplish the objective of having judges from throughout the circuit sit in the various divisions would be through panel assignments, not through more permanent assignments. The territorial aspect would be accommodated by the fact that a majority of the panel would be from the division; the monitoring aspect would be accommodated by the fact that the judges within the division would have the particular responsibility to look to the panel decisions within their division.
5. In my view, the divisional en bancs should be eliminated. They create the prospect of additional cost to litigants and additional delay for an ultimate decision. Even if the divisional en bancs are retained, the circuit-wide en banc process should be the method of further review and should be undertaken in accordance with Rule 35 of the Federal Rules of Appellate Procedure. Although judges of a division would have the special responsibility to review with care the panel decisions within their division, any judge within the circuit should be able to call for an en banc. The composition of the en banc court could be changed as above-indicated and the voting requirements in order to take a case en banc could also be changed as indicated in my Summary below. The important aspect of this matter, however, is that the en banc court should speak for the entire circuit and any review to the United States Supreme Court should be from the en banc court or from a panel decision where en banc review has been denied.

1. The Commission has rendered a valuable service in determining that the Ninth Circuit should not be split for administrative purposes. However, the proposal, without revision, would effectively divide the Ninth Circuit for adjudicative purposes. The Commission has recommended a number of valuable suggestions for experimentation by the circuits throughout the country but has not allowed sufficient latitude for experimentation with the divisional approach suggested.
2. The Ninth Circuit has, more than any other circuit, been willing to experiment with new approaches and innovative concepts. The Ninth Circuit has been a laboratory of experimentation with a large circuit, the results of which have been well accepted by the great majority of the lawyers and judges within the circuit. We should recommend to the Commission that the Report, and the legislation it proposes, be revised to authorize and direct the Ninth Circuit to continue to experiment in order to find the best ways to address the particular problems the Commission perceives, without the rigid statutory confines of the suggested divisional approach.

In my opinion, a viable experiment that could be tried, even at the present time, would be as follows:

1. That we establish the three divisions with resident judges.
2. That oral argument panels in each division be composed of two resident judges, and the third judge being from another division.
3. That there be a circuit-wide en banc composed of 13 or 15 judges as I indicated above.
4. That the death penalty cases be handled in the manner we have established--to be on the regular argument calendars.
5. That the motions and screening process be handled circuit-wide in the manner in which it is currently done, rather than trying to separate out those cases that pertain to particular divisions. We may, however, wish to revise the action that could be taken by motion and screening panels on cases of particular importance so that those would be referred to the divisional argument panels.