In 1789, the speed of transportation and communication was measured by horses and sailing vessels. These two very important factors heavily influenced the structure of the newly created federal judiciary and contributed to the circuit system that exists today.

By the time the Evarts Act was enacted in 1891, transportation and communication had improved. Railroads and steam ships supplanted buggies and schooners. The telegraph had speed but extremely limited capacity for extensive communications and the telephone was in its infancy. Despite these improvements, the federal circuit system and the newly created courts of appeals could not escape the grasp of regionalization and the associated problems of fragmentation.

The past 100 years, however,  have brought about dramatic technological advances. We now measure transportation by the hour rather than by days or weeks. Communications are now instantaneous through electronic means, and, unlike the telegraph, suitable for voluminous transmissions.

These profound changes should widen our horizons and free our thinking about the proper structure of the federal intermediate appellate courts. The constraints of transportation and communication so influential in 1789 and 1891 simply do not exist in this era of jet air travel, E-Mail, Fax,  Interactive-Video transmission, electronic document filing, and automated legal research.

It matters little today whether judges fly from their home chambers in Chicago for argument in Denver or New York or New Orleans or Jacksonville. The speed at which their draft opinions are transmitted via e-mail is the same no matter how far the destination is from their point of origin. Consequently, the regionalization that is so characteristic of, and so ingrained in, the current thinking on court structure should be drastically minimized, if not abandoned altogether.

There is another, more troublesome feature of the current system that improvements in travel and communication have not overcome. Soon after the Evarts Act was enacted, the Circuit Courts of Appeals began to consider themselves as separate entities maintaining what might be called "diplomatic relationships" with their counterparts. Judges were encouraged in this approach by Mast Foos & Co. v. Stover Mfg. Co., 177 U.S. 485 (1990), an unfortunate decision handed down by the Supreme Court in 1900.

In that case, the Supreme Court refused to apply issue preclusion when presented with two different circuit decisions involving the same patent.

The Supreme Court rested its ruling on comity concerns -- concerns that persuade rather than compel. In retrospect, it is clear that the Mast Foos Court missed the opportunity to expound the principle that the various Circuit Courts of Appeals were part of a national system and as such required to apply federal law on a uniform basis.

Thus began the balkanization of federal law and the establishment of the regional judicial fiefdoms so prevalent in the Courts of Appeals today.

The drafters of the Evarts Act envisioned the Supreme Court as the arbiter of intercircuit conflicts, and at the time, that was neither an unreasonable nor unrealistic assessment. Congress failed, however,  to anticipate the explosion of litigation that has occurred in the federal courts.

Even as late as 1924, the Supreme Court managed to review some 10% of the cases filed in the Courts of Appeals. By 1990, however, that figure had fallen to less than 1%, and is even lower today. Now, only a small number of the ever increasing disparate decisions by the courts of appeals can be reconciled by the Supreme Court. Practically speaking, this means that litigants in various parts of the country are governed by federal case law that differs from circuit to circuit.

To combat conflicts in precedents, the "law of the circuit" concept took root some years ago and is in full bloom today. This practice requires appellate panels within a particular court of appeals to follow the precedent set by earlier opinions of that same court. A similar obligation is imposed on the district courts within that circuit. This "law of the circuit" approach was a step in the right direction, but paradoxically, it seems to have strengthened and even seemingly legitimatized the precedential independence of the various courts of appeals. The goal of national uniformity in the interpretation of federal law has been lost in the process of encouraging circuit uniformity.

Litigants are understandably dissatisfied by the uncertainty this development promotes. Worse yet,  this balkanization has fostered the birth of agency non-acquiescence. Today, executive agencies complain that their efforts to apply policy on a national level are frustrated by inconsistent rulings among the various courts of appeals. Agencies handle this difficulty by refusing to follow decisions with which they disagree. Thus, we have the unseemly spectacle of government agencies openly defying the courts of appeals.

In one notable instance, the United States Postal Service presented the same issue to twenty different courts, and eight courts of appeals, losing in every instance until the Supreme Court finally ended the travesty by handing the agency the ultimate defeat.

As has no doubt been pointed out to this Commission, the difference in size among the various courts of appeals has long been a concern. Over the years, we have tinkered with the system and applied band-aid solutions, perpetuating these size disparities and balkanization without devising a comprehensive plan for the entire system of intermediate courts.

Unfortunately, I believe that is still the majority approach. It is time for us to address the basics of appellate structure. It is time to challenge and discard the outmoded mind-sets of judges, lawyers, academics and congressional members that limit our response to present day problems based upon historical factors no longer relevant today.

Because of time constraints, I can only briefly touch on some of the fundamental defects that plague the present system. A more detailed discussion of the serious problems created by intercircuit conflicts and the hodge-podge nature of the present structure can be found in my law review article in the St. Louis Law Review, copies of which I believe have been furnished to you.

The following five basic concepts may serve as guideposts to prepare a plan for the future of our federal intermediate court system:

First, the Supreme Court should retain the ultimate power to declare law in the federal system.

Second, the intermediate court structure should provide uniform national interpretation and application of federal law. Construction of state law should be a secondary consideration and should not be allowed to impair the primary goal of federal law uniformity.

Third, deliberate creation of conflicts among the federal appellate courts should be discouraged. Inadvertent conflicts should be resolved by an internal unit of the intermediate court. Aberrant rulings of various components should be subject to review, not only by the Supreme Court, but by an internal body within the intermediate appellate court itself.

Fourth, while three-judge panels are a reasonably sized unit for initial appellate review, these panels should operate within a larger unit. Under such a system, the pre-filing circulation of precedential opinions would reduce the likelihood of aberrant decisions and secure the benefits of broader consideration. My experience suggests that nine would be the optimum size for such a unit, which for clarity's sake, I will refer to as a "division."

Judges should be appointed to a specific division and expected to serve there during their career on the court. Each division would have a chief judge and a clerk. It would be assigned appeals from designated districts and administrative agencies. In some instances, a division could have its cases assigned based on subject matter jurisdiction similar to the approach used in the present Court of Appeals for the Federal Circuit.

Because travel and communications no longer pose burdensome limitations, the various divisions would be assigned appeals on a basis that would equalize work load. Although geographical contiguity or state borders could still be given some consideration, the workload factor would become the dominant concern. Because state law issues and diversity cases should be subordinate to achieving the goal of uniformity and coherency in federal decisional law, the allocation of appeals from existing districts to a particular division need not be shackled by state boundaries. Thus, for example,  depending on the case load, one division could be allocated appeals from the Middle and Eastern Districts of Pennsylvania, while another division could be assigned appeals from the Western District of Pennsylvania, the Southern District of Ohio, and the Northern District of West Virginia.

Fifth, the unified Court of Appeals would be presided over by a Chief Judge, having general administrative supervision of the unified court. He or she would also be the presiding judge of an entity we could label the Central Division. That body would resolve divisional conflicts and review asserted aberrant decisions of the various divisions.

Judges of the Central Division would be drawn from the various divisions and seniority and experience would be considered. They could be selected by the chief judge of the court in consultation with the various divisions. The members of the Central Division could serve in that capacity for short or lengthy terms. All judges of the unified court would receive the same salary and hold the same rank.

This, or course, is only the general outline of what would be an improved and more efficient structure designed not only to cope with today's caseload but the larger ones anticipated for the future. This is not the occasion to elaborate fully because I seek only to offer the generalized concept of a unified court of appeals for consideration. I am not a completely impractical visionary and I am acutely aware that my concept of a unified court elicits gasps and incredulity when presented to any given group of circuit judges.

An appropriate study of the unified court plan requires time and resources beyond those allocated to this Commission. From my work as Chairman of the Federal Courts Study Committee, I understand the time constraints placed on Commission members and the limited resources that are available. I am also attuned to political realities. It took twenty-five years of wrangling before Congress reached the compromise that became the Evarts Act. Indeed, the proposal to split the Ninth Circuit has already caused substantial controversy.

To have the unified court concept adopted on a nationwide basis at this time is, to put it mildly, somewhat unlikely.

But the unified court concept can help resolve the Ninth Circuit problem. I see a splendid opportunity to test the practicalities of a unified court organization. Because of its size, the Ninth, unlike the other circuits, offers a laboratory for innovation and improvisation.

Presently, the Court of Appeals for the Ninth Circuit has 28 active judges and it appears that more may be authorized. Because the Circuit covers a wide area, there is widespread disagreement over whether it should be split geographically. I believe, however,  that a politically reasonable compromise that would offer tremendous promise for future structural reform of the entire federal intermediate appellate system can be achieved on a basis other than simple geographic surgery.

My proposal is that the Ninth Circuit not be split, but rather, that it be reorganized into permanent divisions of no more than nine active judges apiece. These divisions would function in panels of three judges with provision for divisional en banc action to ensure uniformity. Pre-filing circulation of precedential opinions would be required as a means of quality control, as well as to promote uniformity.

Each division would be bound to follow precedent set by the other divisions. In the event of aberrant decisions or inadvertent divisional conflicts, a circuit en banc whose members would be drawn from all divisions would be convened.

Each division would be assigned appeals from designated districts within the Circuit. This allocation, I recognize, would be politically sensitive, but insofar as possible, equalization of work load should be the prime criterion. State borders should not necessarily dictate case allocation. For example, one division could be assigned only cases from certain districts in California. Other divisions could handle cases from a particular California district plus several districts from other states. Overall, flexibility would be the prime characteristic of the case allocation procedure.

As much as possible, presently commissioned judges could be assigned to a specific division in accordance with their personal preferences.

Future appointments would then be made to specific divisions.

To implement this proposal, it would not be necessary to construct new courthouses or clerks' offices. Present buildings, offices and equipment could be shared by the divisions. For example, even if a division was assigned appeals from outside the Northern District of California, it could nevertheless hear arguments in the courthouse in San Francisco if it desired. Similarly, although appeals to a given division would be filed with its clerk, the physical facilities could be shared with other divisions.

The divisions' physical structures need not correspond with the geographical area generating its appeals. Convenience to court personnel and litigants should be emphasized more than geographical boundaries. For example, it may be that the most convenient place to hear arguments in certain cases would be outside the territory allocated to a particular division. If so, the flexibility inherent in the unified court model would allow the division or division panel to schedule sittings in the locations best suited to the needs of the courts and litigants.

I have sketched for you today a general outline of the unified court concept and I disclaim any assertions of originality. I offer it as a possible answer to the structural problems inherent in the federal intermediate appellate court system.

As a first step toward that goal, a restructure of the Court of Appeals for the Ninth Circuit offers a unique opportunity to test the feasibility of a true national appellate system. In this way, the answer to the current problem would not be another temporary move, but a step toward a more comprehensive nationwide resolution in the future. As dissenting Justice Brandeis wrote over sixty-vie years ago in New State Ice Co. v. Liebman, 285 U.S. 262, 311 (1952): "To stay experimentation in things social and economic is a grave responsibility."

So too it is in judicial administration.