November 4, 1998

COMMENTS OF ALAN B. MORRISON ON THE DRAFT REPORT
OF THE COMMISSION ON STRUCTURAL ALTERNATIVES
FOR THE FEDERAL COURTS OF APPEAL

On April 29, 1998, I submitted a statement to the Commission in which I focussed on matters other than the problem of the workload of the Ninth Circuit. My background and description of my practice are included in that statement and will not be repeated here. I submit these comments as an individual, and they should not be attributed to any organization with which I am affiliated.

These comments respond to both the recommendations of the majority and the dissent. With respect to the views of the majority, I agree generally with the proposal to divide the Ninth Circuit into three divisions, and I support the other recommendations in general, although I have a number of specific comments about the details of the various legislative proposals that are contained in the appendices. With respect to the dissent, I agree with its overall approach, in particular the criteria in its Appendix II defining the appropriate federal-state allocation of judicial business. However, as for its specific recommendation on substantially cutting back on diversity jurisdiction, I oppose the drastic limitation on diversity jurisdiction, in part because I am not convinced that the data shows that its proposal would produce a significant reduction in appellate caseload.

(a) Proposal Relating to Creating Divisions Within Circuits

Section 2(a)(4) deals with the issue of direct appeals from decisions of federal administrative agencies. It provides that they shall be taken "to the regional divisions specified in the rules of the courts of appeals." Several matters need clarification and/or change.

The provision specifies appeals, but it also should include petitions for review of matters such as tax court decisions. Second, it is not clear why the rules of the individual courts of appeals should establish the assignment of a division for cases coming on direct review. It would seem to be more consistent with the general thrust of the recommendation to provide that the appeal will be taken to the appropriate court of appeals, but assigned to the division to which the appeal would be taken if the division were a separate court of appeals. Thus, in a case in which an appeal can be taken to the Ninth Circuit because of the taxpayer's residence, the assignment to a division should be based on residence, and not on any other basis set forth a rule of the court of appeals. Similar amendments should be made to 28 U.S.C. ' 2112(a)(3) to deal with multiple petitions for review and to provide that the transfer shall be to the division (or among divisions) in the manner provided in that section.

2. Various statutes provide for decisions by three-judge courts under the procedures set forth in 28 U.S.C. ' 2284. That provision should also be amended to require that the circuit judge be from the division (probably one of the local rather than non-local judges), rather than simply the circuit.

3. Section 2 currently provides that a majority of the judges shall reside within the division's jurisdiction, but it might be preferable to have a somewhat greater number, such as two-thirds, if the purpose is to keep the law consistent. That change would also cut down on the travel and inconvenience for the circuit judges who are sitting outside their division's locale.

4. There are some problems regarding the circuit en banc procedures. First, there needs to be a procedure for replacements in the event of a recusal, which is always a possibility. Second, I oppose the requirement in section 3(a) that a circuit may grant en banc review only if the decision of the panel has been sought and denied by the division. If there is a clear conflict with another division, why should there be a third step in the process? The circuit would be free not to accept the case until there has been a division en banc, but that should be discretionary and not mandatory.

5. Section 7 provides for an automatic termination of the divisions in the Ninth Circuit after seven years, but new section 46 would make permanent the parallel provisions for dividing up other circuits into divisions where certain conditions are met. If there is an automatic termination of the Ninth Circuit after seven years, then section 46 would still apply, and the court would have to reorganize under those terms which do not precisely fit the Ninth Circuit. In my view, the more sensible resolution is to use the same monitoring and report provisions by the Federal Judicial Center after eight years in section 4 and apply them to the amendment covering the Ninth Circuit as well as to the amendment to section 46.

(b) Other Workload Reduction Proposals

The Commission makes two recommendations regarding other ways to deal with the problem of workload -- the two judge panel and the district court appellate panel. I recognize the necessity of taking steps to relieve the appellate workload, but the suggestions made below are necessary to assure that "appellate justice" is not an oxymoron for most litigants. Both of proposals have common elements that raise some concerns, principally issues relating to case selection.^*

First, there should be some form of public procedure under which each circuit (or division if these judgments are to be made on a division basis) establishes the criteria for cases that will and will not be sent to either a two judge panel or a district court appellate panel. It may be that some cases should go one route and others another, but at the very least the matters ought to be considered together in some form of public proceeding. The preferable format would be a proposal issued for public comment, like changes in local rules, so that there is a reasonable opportunity to be heard before the criteria are set.

Second, cases should be principally selected by issues rather than subject matter. For example, social security cases claiming that there is no substantial evidence to support a finding are clearly candidates for some form of lesser review, whereas questions dealing with the proper interpretation of the statute, such as those which led to the long battle over the non-acquiescence policy in the early 1980's, raise very different concerns and should be handled by a full appeals court panel. Similarly, in diversity cases, if the issue is whether the jury was properly instructed, received prejudicial evidence, or made a determination that was based upon insufficient evidence, those might be candidates for alternative appellate review, but other diversity cases, such as those raising important questions about the interpretation of the Federal Rules of Civil Procedure or various jurisdictional provisions, raise very different concerns.

Related to this is my view that the decisions about whether a particular case should or should not be assigned to either a two judge panel or a district court appellate panel should be made only after the briefs have been filed, not at an early stage of the appeal. This approach is a necessary corollary to my position that the allocation should be made based on the issues in a particular case, rather than the subject matter of the case itself. In addition, copies of the rules/criteria should be sent to the Federal Judicial Center for its evaluation and to assure consistency with other circuits, as well as for use in the report under section 4. Finally, there should be a method by which parties who believe that the case has been wrongly assigned can move to have the matter reset on the main docket.

The possibility of a two judge court also raises the question of whether both judges should be active judges or whether one of them can be a senior or visiting judge. In my view, except when a circuit court is substantially under its complement of judges, both members should be active judges, with senior and visiting judges used only to fill up a three member panel.

The issue of senior judges also calls to mind the question of why Rule 46(c) allows each court of appeals to decide whether senior judges should sit in en banc cases. I have never understood why this local option should exist and would prefer a rule under which a senior judge may not sit unless he or she sat on the panel which produced the opinion under en banc consideration. I would also continue the rule under which a senior judge who sat on the panel would not be permitted to vote on whether the case should be heard en banc (whether at the division or circuit level) but would be permitted to participate if the case were reargued.

With respect to the statute authorizing district court appellate panels, there is no reason for appeals to be formally taken to those panels, and imposing such a requirement creates the possibility of errors and the necessity of motions to transfer. Rather, consistent with my views on the assignment of cases based on the substantive issues, and then only after they have been fully briefed, the appeal should continue to be taken to the appropriate court of appeals, with an assignment thereafter to a district court appellate panel or a two judge panel, as they case may be. This might also help with some of the administrative matters, but it is principally for substantive reasons that I make the recommendation.

If a decision of a district court appellate panel is subject to review, there should be time limits set forth in new section 145(d), as well as some further statement as to procedures to be followed, including who will make the decision as to whether the appeal by the circuit court should be granted. I oppose the part of new section 145(d) which limits the circumstances in which the court of appeals has jurisdiction to review a district court appellate panel. Since review is discretionary, attempts to limit jurisdiction will only engender further litigation for no benefit. The question for the circuit court should be, ought the first decision be reviewed, and attempts to cabin in the discretion of the courts of appeals are unwise and counter-productive. The overburdened courts of appeals will not routinely take these cases, but they should have the discretion to do so whenever they want and for whatever reason.

Moreover, the rules should not be made, as proposed section 145(d) suggests, by the courts of appeals but should instead either be Federal Rules of Appellate Procedure, or they should be included in the statute. There may be reasons to have some minor variations among the circuits, but basic issues, such as the amount of time and whether the matter is pursued by motion or the filing of formal briefs, ought to be in the basic rules governing all circuits.^**

The portion of the dissenting statement on the appropriate federal-state allocation of judicial business is both a useful and a correct statement as to that issue, but the principal difficulty is in persuading the Congress to adopt it, particularly with respect to the issue of narrowing diversity jurisdiction. Given the grave difficulties in obtaining even a modest change in diversity jurisdiction (other than occasionally raising the amount in controversy, which can easily be evaded in most cases), it seems highly unlikely that Congress is going to enact legislation to achieve this end.

But assuming that Congress were inclined to do so, it is not at all clear that the "breathing room" for the courts of appeals that the dissent believes would be achieved will actually occur as a result of the changes that are being proposed. In my prior statement, I suggested that further statistical evidence needs to be gathered, and while I recognize that some of the statistics set forth in this section of dissent do point in the general direction that changes in diversity jurisdiction would help the courts of appeals, they seem to me to be inconclusive largely for reasons that I set forth in my earlier statement. In particular, many of the numbers focus only on civil cases and deal only with those that have been orally argued. While there is an assertion by Judge Posner that the diversity cases are "above average difficulty," that does not necessarily apply to the courts of appeals since they may mean only that they result in more jury trials or more discovery disputes than cases that are regularly decided on summary judgment. Moreover, the statement invites a comparison with many other kinds of cases which do not take up much time on a per case basis, such as pro se habeas cases. Some of the appellate caseload problems arising from diversity cases may be alleviated by the proposals made by the majority, but it is not clear that they would specifically assist in the reduction of appellate work load.

Finally, I strongly oppose the radical change in diversity jurisdiction to a regime of permissive jurisdiction, in large part because the proposal places almost insurmountable burdens on those who wish to use the federal courts in diversity cases because they must establish prejudice in the local court "with reasonable certainty." There are a number of lesser measures -- such as narrowing the definition of in-state defendants such that companies, like General Motors, with large plants in a state, could no longer remove to federal court, and forbidding state and local governments who are sued under 42 U.S.C. ' 1983 in state courts from removing to the federal courts (generally done to cause inconvenience to the plaintiff) that are at least doable if not sufficient to satisfy all of the needs of caseload reduction. But turning virtually all diversity jurisdiction into discretionary jurisdiction is far too sweeping a proposal, at least without having tried lesser measures.