January 26, 1998

Justice Byron R. White United States Supreme Court One First Street, NE Washington, DC 20543-0001	Honorable Gilbert S. Merritt 303 United States Customs House 701 Broadway Nashville, TN 37203
Honorable Pamela Ann Rymer Richard H. Chambers Court of Appeals Building 125 South Grand Avenue Pasadena, CA 91105-1652	Honorable William D. Browning 301 James A. Walsh Courthouse 55 East Broadway Boulevard Tucson, AZ 85701

N. Lee Cooper, Esq.
Maynard, Cooper & Gale, PC
1901 6th Avenue N.
AmSouth Harbert Plaza, Suite 2400
Birmingham, AL 35203-2602

Dear Commission Members:

I take the liberty of writing to express my views on the subject of your charge. I do so because it is a subject with which I was heavily involved as a member of the Judicial Conference Committee on Long Range Planning, and hence I hope that I can offer a perspective that will be of value to you.

First and foremost, I urge you to keep in mind in your deliberations the Long Range Plan for the Federal Courts, which, as you know, has been adopted by the Judicial Conference of the United States.⁽¹⁾ I refer specifically to two facets of the Long Range Plan. First, I commend to you the statistical projections for the caseload of the courts of appeals into the next century. Even if the staggering projections for the year 2020, see Long Range Plan at 15-16, are exaggerated,⁽²⁾ the Commission must, I submit, be thinking of the federal appellate courts of the future, not just those of the present. As a corollary to this point, I suggest that if, in order to permit the courts of appeals circa 2020 to handle a vastly increased workload, the Commission deems it necessary to redraw some circuit lines, the republic will stand. After all, it is over a century since the basic appellate structure was created, and revisiting the whole structure once every hundred years does not seem inappropriate.

Second, I call your attention the commentary to the provisions of Recommendations 16 through 19 of the Long Range Plan, which read as follows:

RECOMMENTION 16: The federal appellate function should be performed primarily in:

(a) a generalist court of appeals established in each regional judicial circuit; and

(b) a Court of Appeals for the Federal Circuit with nationwide jurisdiction in certain subject-matter areas.

RECOMMENDATION 17: Each court of appeals should comprise a number of judges sufficient to maintain access to and excellence of federal appellate justice. Circuit restructuring should occur only if compelling empirical evidence demonstrates adjudicative or administrative dysfunction in a court so that it cannot continue to deliver quality justice and coherent, consistent circuit law in the face of increasing workload.

RECOMMENDATION 18: To the extent practicable, workload should be equalized among judges of the courts of appeals nationally.

RECOMMENDATION 19: The United States Supreme Court should continue to be the sole arbiter of conflicting precedents among the courts of appeals.

In particular, I stress the following commentary to Recommendation 17:

In principle, each court of appeals should consist of a number of judges sufficient to: maintain traditional access to, and excellence of, federal appellate justice; preserve judicial collegiality and the consistency, coherence, and quality of circuit precedent; and facilitate effective court administration and governance. An appellate "court," in this special sense, is not merely an administrative entity. Nor should it consist of a large group of strangers -- like a jury venire -- who are essentially unknown to one another. Rather, a "court" is a cohesive group of individuals who are familiar with one another's ways of thinking, reacting, persuading, and being persuaded. The court becomes an institution -- an incorporeal body of precedent and tradition, of shared experiences and collegial feelings, whose members possess a common devotion to mastering circuit law, maintaining its coherence and consistency (thus assuring its predictability), and adjudicating cases in like manner. (Emphasis added.)

In my view, the salient principle underlying the emphasized language is that, when a circuit gets so large that an individual judge cannot truly know the law of his or her circuit, and I do not mean by having a staff attorney or law clerk look it up when the occasion occurs, the circuit is too large and must be split. I note in this regard that, in the Third Circuit, we follow the practice of pre-filing circulation of all published opinions. In this way the judges read them before they are filed, and by e-mail correspondence frequently point out problems with the circulating opinion in terms of the consistency and coherence of circuit law. This process often heads off problems, better caught before the opinion is filed.⁽³⁾ Most importantly, in this way, a judge can "master" circuit law. (He or she could, of course, read all the slip opinions after they are filed, but there is no pressure to do so at that time.) That said, I cannot imagine a judge in a circuit as large as the Ninth, with its staggering volume of opinions, being able to do what we in the Third Circuit do, or, through whatever device, other than superhuman brainpower, being able to "master" circuit law. If this assumption is correct, the Ninth Circuit, according to my rough rule of thumb, needs to be split.

You will doubtless be getting a great deal of advice as to whether and how to split the Ninth Circuit. In my view, the landscape in the area has changed by reason of the recent adoption by the California Supreme Court of a certification procedure. This development should make more acceptable the view that Judge O'Scannlain and others have advanced (myself included, see 46 Am. U.L. Rev. 263, 318 (1996)) that California can be divided into two circuits, with the provision of an en banc panel of the two circuits should there be differing predictions (not resolved by the certification procedure) about what California law is.

The Long Range Plan endorses the notion of regional circuits. See Recommendation 16. Since Alaska and Arizona can in no reasonable sense be deemed to be part of the same region, some of the proposals for realignment of the Ninth Circuit do not, as I see it, pass muster. Indeed Arizona, Idaho and Montana seem to be more a part of the region that comprises the Tenth Circuit, and I would hope that consideration be given to realignment in that direction. Most of the other circuits are pretty much regional in character, but query whether South Dakota and Arkansas in the Eighth Circuit, and Michigan and Tennessee in the Sixth Circuit meet the "regional" standard.

Notwithstanding my "once every hundred years" comment, I would never suggest that the Commission engage in a redrawing exercise willy nilly. I acknowledge the value of long working institutional arrangements and the importance of the ability of lawyers and judges to rely on precedent. And so a strong case needs to be made before disrupting existing circuit alignment. That is the gist of the commentary to Recommendation 17 of the Long Range Plan. But Recommendation 18 of the Long Range Plan is also important, providing that "to the extent practicable, workload should be equalized among judges of the courts of appeals nationally." This precept may commend other realignments, which are, of course, within the scope of your charter, and to the extent that it does, the regionalism principle should, I think, give sway.⁽⁴⁾ If this is done, to repeat, the republic will stand.

Certain small circuits are cohesive and administer a high quality of justice. On the other hand, query whether, given the hydraulic pressures projected for the federal appellate courts circa 2020, we can afford to have a circuit as small as either the First (six judges), D.C. or Federal remain intact for the next century. While these circuits currently work well (and "ain't broke"), I fear that they represent a luxury we will not be able to afford in 2020. If we end up having to split the larger circuits, resulting in too many small circuits, we will thereby tax the ability of the United States Supreme Court to be the sole arbiter of conflicting precedents among the circuits (see Recommendation 19 commentary).⁽⁵⁾

That commentary adverts to the possibility of creating new appellate structures (e.g., an intercircuit tribunal or a new tier of federal courts), and of allowing the Supreme Court to refer cases presenting conflicts to a court of appeals not involved in the conflict. The Committee on Long Range Planning considered and rejected each of these alternatives, and I hope that the Commission will do likewise, notwithstanding that the breadth of its charter encompasses their consideration. The commentary also references the possibility of "jumbo circuits." My previous comments set forth my position as to why they are undesirable. Judge Newman has far more eloquently demonstrated that this is so, see Jon O. Newman, 1,000 judges -- the limit for an effective federal judiciary, 76 Judicature 188 (1993). I add that, as a federal appellate judge with some 16 years experience, I do not view the proposals to increase appellate productivity by increasing the size of the staff attorney's office or by stepped-up "screening" procedures with anything but disfavor. I believe that most federal appellate judges share this view.

The Commission's charge "to report to the President and the Congress its recommendations for such changes in circuit boundaries or structure as may be appropriate for the expeditious and effective disposition of the caseload of the Federal Courts of Appeals, consistent with fundamental concepts of fairness and due process" strikes me as one that may counsel certain realignments. I note in this regard that the experience following the division of the Fifth Circuit demonstrates that the problems of sorting out precedent following a circuit split are fully manageable. While reconfiguration makes for a more complicated problem with respect to precedent, I am confident that it also could be managed.

Respectfully yours,

Edward R. Becker

1. Background information on the development of the Plan reposes in the Administrative Office of the United States Courts, whose Long Range Planning Office provided staff support to the Long Range Planning Committee. The files of that office contain the reports, research and analyses considered by the Committee, as well as public comments and hearing testimony on proposed Plan recommendations.

2. The report states that in 2020 annual appeals could approach 335,000 (in 1995 they numbered about 50,000). And see Thomas Baker, 24 Pepp.L.Rev. 859 (1997) (Professor Baker makes 2020 the judicial equivalent of Orwell's "1984").

3. This pre-filing circulation is not intended as a quality control device, though I note that it frequently works that way to the great benefit of our judicial work product. Again, it is better to have flaws corrected before the opinion is filed.

4. Realignment of the judicial circuits will, of necessity, impact on the district courts as well. While the Commission will no doubt need to explore that issue carefully should it recommend realignments, I will not attempt to deal with that issue in this letter.

5. There are a number of possible realignments with respect to the small circuits. At the very least the First Circuit could absorb Vermont. Broader possibilities also exist. The D.C. Circuit workload is so much smaller than the other circuits that perhaps it should be merged with the Fourth or the Federal Circuit. The latter court is also not as busy as the other circuits, and the Long Range Plan is quite tentative in its support for such a specialized circuit, even one that is a facet of history (and political compromise). ("Nevertheless, in most instances the well-known dangers of judicial specialization outweigh any such benefits." Long Range Plan at 43.) Moreover, the current jurisdiction of the Federal Circuit is a hodge-podge (patent cases, federal personnel appeals, cases from the Court of Veterans Appeals, etc.). I acknowledge that the suggestion that the D.C. Circuit be merged is dramatic. Its administrative law caseload is celebrated, but all the circuits have those cases, if in smaller numbers, and the D.C. Circuit also has many drug and sentencing cases, just like the other circuits.