Ten—
A Court Too Large?

So while it is inevitable that the northern part of the circuit will eventually be separated from the southern part, the time is not yet here.
Judge William Denman, 1937

Splitting the Ninth Circuit is an old idea whose time has not yet arrived.
Chief Judge Alfred T. Goodwin, 1990

The New Deal transformed the work of the Ninth Circuit. In prior decades issues of railroad consolidation, Chinese immigration, mining and natural resource development, and criminal appeals under war-related and Prohibition statutes had constituted a large segment of the court's docket. By the late 1930s and early 1940s, appeals raising issues under Roosevelt's New Deal programs increasingly occupied the attention, not just of the Ninth Circuit, but of every federal circuit court of appeals. At the same time that the nature of its docket was changing, the Ninth Circuit as an institution was undergoing revision. The seven new judges seated by 1937—Wilbur, Garrecht, Denman, Mathews, Haney, Stephens, and Healy—each undoubtedly brought a new perspective to the court, but the changes touched on a deeper concern: whether, within the existing framework of court rules and procedures, a court of seven members with a wide range of ideological and jurisprudential viewpoints could provide justice across the circuit's massive geographical jurisdiction.

The issue of the Ninth Circuit's ideal size was a source of controversy even before the Evarts Act of 1891 had created the circuit courts of appeals. A year before that law's enactment, San Francisco attorney Frank M. Stone had urged the Senate Judiciary Committee to reconsider its configuration of the Ninth Circuit. Such a large circuit, Stone wrote, "would be more than any one such court of appeals . . . could possibly

attend to without the business running behind, and the calendar becoming clogged, if such circuit judges attended to nothing but the appeals, and sat as a court of appeals." To emphasize his point, he enclosed a large map of the West with the proposed circuit boundaries marked. These concerns, however, had more to do with fears of administrative unwieldiness than with regional variations in the substance of the cases appealed to the court or fears that judges appointed from one part of the Ninth Circuit would be insensitive to the social and economic concerns of litigants in other parts of the circuit.^[1]

In the first two decades of the twentieth century, for example, a southern California circuit judge, Erskine Ross, had more often than not succeeded in protecting Pacific Northwest lumber interests over the determined opposition of his Oregon colleague, William Gilbert.^[2] But concern over efficient judicial administration persisted and, as the court's first half-century came to a close, sparked a contentious debate among Ninth Circuit judges over dividing the circuit.

I. Docket Pressures

Individual lawyers had been expressing concern for some time, but concerted debate over the ideal size of the Ninth Circuit did not begin in earnest until the 1930s. At the time, the burden of handling cases arising from the largest geographical land mass under the jurisdiction of any United States court, save the Supreme Court itself, stemmed primarily from the vast travel distances required of judges and litigants. Through 1939, when it ranked behind only the Second Circuit, the Ninth Circuit had stood comfortably amid the other circuits in the number of new cases filed each year, but in fact all the circuit courts of appeals had experienced significant growth in annual docketings. The Ninth Circuit itself reported an increase from 158 appeals filed in 1920 to 363 in 1935. The Eighth Circuit had exceeded the Ninth in total number of filings between 1900 and 1929, but in 1929 Congress divided the Eighth, and by the following year the new Eighth had five judges and the new Tenth had four. In comparison, the Ninth was severely shorthanded, with only three judges.^[3]

During the 1920s a number of factors combined to increase the court's workload. One was the economic and population growth of the West. The clerk of the court, Paul O'Brien, reported that part of the court's burgeoning caseload was attributable to a 40 percent increase in immigration into the circuit between 1920 and 1930. From 1900 to 1930, the

population in the Ninth Circuit's states grew from 3,204,752 to 10,131,325. The increase in the number of judges was not concomitant. Although the official number of authorized circuit judgeships remained at three in 1900 and 1930 (William Hunt's assignment from the moribund Commerce Court added a fourth seat between 1911 and 1928), the number of district judges in the circuit rose from eleven to twenty-eight.^[4]

In addition to the natural growth in population and economic activity and the large number of Prohibition-related cases during this decade, another significant influence on the federal appellate judiciary's workload came from the Judiciary Act of 1925. The movement to reform the Supreme Court's jurisdiction in the mid-1920s had to overcome many of the old misgivings that had accompanied the creation of the circuit courts of appeals in 1891. In the late 1880s many lawyers believed that litigants had a right to Supreme Court review to vindicate their federal claims. Lawyers initially resisted having the circuit courts of appeals effectively serve as courts of last resort when the Supreme Court denied certiorari, even when the cases raised no important national or constitutional issues. By the 1920s, however, distrust of the appellate court system had diminished, thanks to the effectiveness and efficiency of the circuit courts of appeals in handling a large number of cases that until the Evarts Act would have required Supreme Court review.^[5]

Unburdened by the bar's previous resistance to conferring greater authority on the appellate tribunals to dispose of cases, Chief Justice Taft and his brethren promoted a bill that aimed to meet the Supreme Court's needs. The central goal of the "Judges' Bill," as it was appropriately nicknamed, was to preserve for Supreme Court review only those cases of national significance. In 1924, nearly one-third of the Court's docket consisted of mandatory appeals from district courts. A smaller but no less important source of work for the Court consisted of cases requiring mandatory review of certain decisions by circuit courts of appeals. From 1916 to 1925 the Supreme Court considered an average of forty such cases each year. The Judges' Bill proposed to end the Court's nondiscretionary review of these latter cases. Virtually all cases disposed of by a circuit court of appeals would henceforth be reviewed only upon a writ of certiorari by the Supreme Court. In this manner the bill proposed to magnify the importance of the appellate courts' work. It would also increase the number of situations in which these courts would review district court decisions. Because the Supreme Court had more discretion under the 1925 law to decide which cases to hear, the circuit courts of

appeals became in practical terms the tribunals of last resort for all but a small percentage of federal cases. The reform enabled the Supreme Court to regulate its caseload with relative ease, but the appellate courts had no such luxury. They became completely captive to the litigation tendencies of their regions; as litigation increased, so too did the dockets of the lower federal courts.^[6]

By affecting review of cases from dependencies of the United States, the 1925 act had a significant direct effect on the Ninth Circuit. In the years leading up to the Judges' Bill Congress had enacted a wide array of laws that accorded miscellaneous review in the Supreme Court to cases arising in Hawaii, Alaska, Puerto Rico, the Philippines, the Panama Canal Zone, the Virgin Islands, and the United States Court in China. The 1925 act withdrew these varied avenues of review by the Supreme Court and distributed them to the appropriate circuit courts of appeals. For the Ninth Circuit, this reform involved accession of cases from Hawaii, Alaska, and the China Court.^[7] And whereas economic, demographic, and jurisdictional influences theoretically would affect the circuits more or less equally, the Ninth Circuit was in practice hit particularly hard by the increase in cases. The longevity of Gilbert, Ross, and Morrow, which enabled the court to build a fine reputation in its early decades, gave way to the short tenures of judges under whom the court's docket fell increasingly in arrears and its reputation for quality work declined. The failed nomination of Wallace McCamant, the short terms of Frank Dietrich and William Sawtelle (three years each), and the untimely death of Frank Rudkin threw the court into a decade-long disarray from 1925 to 1935.

By the end of 1930, the state of affairs in the Ninth Circuit had reached crisis proportions: Dietrich had died in October, and Gilbert, senior judge of the court, was seriously ill. Shorthanded as it was, the court struggled to dispatch an ever-increasing caseload. The number of cases docketed in the Ninth Circuit totaled 307 in 1930, 327 in 1931, 378 in 1932, and 322 in 1933. In this same period, the number of cases pending at the end of the term grew from 99 in 1930 to 160 by 1933. The shortage of judges became particularly acute in 1934, when at the close of the fiscal year 220 cases were still awaiting decision. The appointment of new judges did not help ease the load until long after the end of 1935, when 257 cases remained undecided. In May of 1932, therefore, the California State Bar passed a resolution urging Congress to create a fourth judgeship on the Ninth Circuit.^[8]

Legislation to increase the size of the court was not, however, immediately forthcoming. Congress did enact legislation in 1933 to reauthorize the Ninth Circuit seat that had lapsed with Judge William B. Gilbert's death. The death of Gilbert on April 27, 1931, was followed almost at once by Rudkin's death, on May 3, 1931. Three years later William H. Sawtelle died, leaving Wilbur and Garrecht alone to constitute the court, with two positions remaining to be filled. The senior judge, Wilbur, had only four years' experience; Garrecht, only one. With the number of cases in arrears steadily mounting, President Roosevelt attempted to fill these vacancies, but to no avail. His nomination of Judge Frank H. Norcross, of Carson City, Nevada, was never acted upon by the Senate. The judge who eventually filled this seat, which had once been held by Gilbert, was William Denman, who did not join the court until March of 1935.^[9]

As the court limped along with this shortage in judicial personnel, it adopted expedient means of handling cases, the most controversial of which was the "one-judge decision." It is unclear when the court started this practice, but as the number of pending cases mounted, the court informally began to assign a case to a particular judge before oral argument. After the hearing, but before any conference was conducted, that judge prepared a complete and final opinion of the court without knowing what the other two members of the panel thought of the case. In his first year on the court William Denman prepared sixty-five such opinions. As he confessed to the Senate Judiciary Committee during hearings on more judgeships for the court, "I take no pride in them, I pray they are right, for in 90 percent of them we never had a conference. The opinion came back O.K.'d by the other judges, with whom the cases had not been discussed. That has been the practice of the court for six or seven years." The Ninth Circuit's senior judge, Curtis D. Wilbur, believed that the step was necessary if they were to attempt to keep up with the workload.^[10]

A wide gulf existed between the Ninth Circuit's elimination of post-argument conferences and the Supreme Court's practice of full conferences on every case it heard. Wilbur maintained that application of the Supreme Court procedure, which was also generally followed in the Tenth Circuit, would cut the output of the Ninth Circuit roughly in half. Denman flatly rejected this view, but his colleague from Washington, Francis Garrecht, was ambivalent. Garrecht agreed with Denman that full conferences would promote better justice, but he also acknowledged the partial validity of Wilbur's contention that the court could not keep abreast of its work if it had a full conference after each case was heard.^[11]

It is difficult to gauge the merit of Wilbur's view, given the modern practice of brief, post-argument conferences. The "one-judge decisions" somewhat speeded the handling of cases. The average time between submission of the appeal and the rendering of an opinion between 1926 and 1935 was 123 days, whereas during the first three decades of the court's history the average time of disposition fluctuated between 145 and 168 days. When the court reached its full complement of judges during the New Deal era, the Ninth Circuit judges averaged 70 days to render an opinion. Not only did the addition of more judges speed the actual decision-making process, the court's reversal rate in the Supreme Court also dropped. In 1942, when the court was at its full strength of seven judges, Denman wrote to Bert Haney regarding the difference in reversal rate between the "no conference" period and the "conference" period, which began again after the authorization of two additional judges in 1937. Taking two-year periods before and after, Denman reported that in the former, the court decided 416 cases and was reversed 21 times. In the latter, the court decided 876 cases and was reversed in 26; the reversal rate had dropped from 5 percent to 2.9 percent. This decline could not be attributed to the Supreme Court's general acceptance after 1937 of the constitutionality of the New Deal, since only a small fraction of the reversals involved such issues (3 of 21), with a like ratio in the latter period. Although these statistics did not compel specific conclusions, Denman pointed out that "in several conferences with the justices of the Supreme Court there has been comment upon the improvement in the character of the opinions of this court since we have had seven judges."^[12]

II. "Denman's Judgeships"

By January, 1935, with two vacancies still remaining, Wilbur and Garrecht undoubtedly could not even begin to envision the seven-member court that would exist by 1937. Perhaps because he was still smarting from the skirmish over the Norcross nomination, Roosevelt had tarried in sending new names to the Senate. He finally nominated William Denman of San Francisco and Clifton Mathews of Phoenix. From the start, Denman worked tirelessly to improve judicial administration and add more judges to the Ninth Circuit. His third day in chambers had barely ended when he began his lobbying effort for more circuit judgeships. Citing the vast growth in docketings, Denman wrote to Roosevelt that four judges "cannot possibly discharge this volume of business.

Already the court is heavily in arrears and by the October term will be 300 cases behind its docketings." He pointedly spelled out the political implications: "The Bar is bitterly resentful, which is not helpful to the Administration. The resentment is justifiable, though, of course, its objective is not."^[13]

By the summer of 1935, Roosevelt was able to respond that he was "in the throes of the problem and the bill will be signed in a week." Roosevelt made good on his promise, and Bert E. Haney of Oregon brought the court's strength up to five in September of 1935. Such lobbying success, especially by a new judge, would satisfy most people, but Denman was not yet content. In the face of what he later would describe bluntly as the "defeatist" attitude of Garrecht and Wilbur, Denman continued to push for more judgeships. Having failed to persuade his colleagues that it was in their interest to contribute funds, Denman traveled to Washington at his own expense to pressure Congress and the president.^[14]

In March, 1936, Denman began to write Roosevelt with relentless regularity, and Roosevelt approved of his efforts. After one series of letters, Roosevelt replied that Denman should "[k]eep up the good work and keep my material up to date. I am getting all ready for a real speech on the subject. Legislation at this session is, I fear, out of the question but it is a fine objective and a fine platform." Denman employed numerous tactics to get the president's attention. In one letter he appealed to the president's sense of public accountability: once the Roosevelt administration had "adequately staffed the courts," he said, and "given them abundant funds under their own administration, just as with the Congress—you and the Congress can properly call them to account in the performance of their 'coordinate' constitutional function." Denman also raised the party issue, pointing out that 186 district, circuit, and Supreme Court jurists had been appointed since Wilson left office; only 14 of the 159 selected by Republican presidents were Democrats. (Twenty-seven judges had claimed no party affiliation, and Denman was quick to suggest that these "should be checked up by someone.") Denman strongly hinted that the president should lay the blame for the courts' backlogs on the Republican domination of the courts.^[15]

Denman did not confine his lobbying to the president. He also wrote to Attorney General Homer Cummings, sending carbon copies to Roosevelt. He persuaded Senior Judge Curtis Wilbur to propound publicly the need for between twelve and fifteen new circuit court judgeships, an increase systemwide of 30 percent. Although Denman knew Wilbur

did not expect him to succeed, he was not above trumpeting the senior judge's support in his letters to Roosevelt. He also convinced Garrecht to write a supportive letter to Representative Hatton W. Sumners, who chaired the House Judiciary Committee. And, lest Roosevelt uncharacteristically fear Republican opposition to the administration's quest for an expanded judiciary, Denman took the opportunity of Chief Justice Charles Evans Hughes's proposal for more judges to argue that this freed the president to pursue reform. In this election year, Denman had no doubt that Roosevelt would retain his office. He urged the president to think ahead: "[W]hat must be done in your next term! The need of the litigant for more judges will then be at its peak."^[16]

This correspondence subsided briefly while Denman addressed the pile of opinions to be written from his first full term on the court. In letters to Wilbur and Haney at the end of June, 1936, Denman expressed his pleasure in distributing sixty-nine opinions during his first fiscal year, about forty of them during the second six months. To Haney he explained, "The reason I pressed through the number I have written is because I do not want some of the Congressmen to say, when I am lobbying for more judges at the next session, 'this fellow Denman is trying to shunt off to somebody else his own work, because he can't keep up with the procession.'" As soon as Roosevelt triumphed over Alf Landon in the 1936 election, Denman renewed his lobbying, prefacing his congratulatory missive with the curt statement, "The New Deal needs more Federal Judges." The letter surveyed a number of national and international problems before coming to the main point: "Hence, in my congratulations, I repeat it as a credo. More Judges." It is unclear whether Denman's "Dear Franklin" letter had any direct impact on the president. Roosevelt responded at intervals only with a brief note of encouragement. But even if these letters were limited in their effect, despite Denman's clever attempts to get them on the president's desk, the judge's testimony before Congress provided powerful evidence of the need for more judges.^[17]

Testifying before a subcommittee of the Senate Judiciary Committee on February 24, 1937, Denman used both quantitative and qualitative arguments to establish the severe need for more Ninth Circuit judges. He noted, not only that the number of appeals to the court had doubled during the past fifteen years, but also that the cases had become increasingly complex. One outcome of the shortage was that southern California and Arizona, which provided 37 percent of the appeals to the Ninth Circuit, could not have a session of the court in Los Angeles.

Denman added that, in part because the state of California alone produced 59 percent of the Ninth Circuit's appeals, the California Bar Association firmly backed reform efforts. In 1936, according to Denman, it had passed a resolution calling on Congress to relieve the congestion by appointing more judges.^[18]

The problem had reached alarming proportions. One measure of the Ninth Circuit's size was the enormous number of federal districts and territorial supreme courts over which it had appellate jurisdiction. In 1937 the Ninth Circuit heard appeals from twenty-eight federal district judges and the Hawaii Supreme Court: according to Denman, this was double the average number of district judges of the other nine circuits and one-third more than the circuit next in number of districts. In addition, the court had jurisdiction over appeals from the United States Court for China. Not only did the court preside over a large number of subordinate tribunals, the states composing the Ninth Circuit enjoyed the highest rate of population increase, and this growing population engaged in a vast array of social and economic activities that bred litigation. The range of issues requiring adjudication covered the spectrum of western development: agriculture, reclamation, gold dredging, lumber, Native American issues, Hawaiian feudal law, petroleum, stock raising, irrigation, immigration, shipping, commercial fishing, electrical power, manufacturing of all kinds, and community property. The number of judges on the court had not begun to keep pace with the litigation produced by these social and economic concerns. The near-doubling of the caseload between 1922 and 1937 with the addition of only one judgeship meant that each jurist had to take on two-thirds more work.^[19]

The hearing at which Denman gave this testimony concerned a bill proposed by Senator Homer T. Bone of Washington to add two judges to the court. Bone was later to be a key player in the struggle to divide the circuit, a policy alternative Denman in 1937 believed to be inevitable. "We need these two additional judges now," he testified. "Before long, as I will later discuss, you will have to divide the circuit and have still more judges." By 1941, when a serious proposal to split the circuit had to be confronted, Denman changed his view. Nevertheless, the Ninth Circuit's geographical size suggested two competing options: one, to increase the number of judges on the court; the other, to divide the circuit, as Congress had done with the Eighth. The latter alternative, however, invariably required the authorization of more judgeships to ensure that each of the new circuits would have at least three active judges. The predominant issue, therefore, was whether administrative need justified

division. Denman did not think so. In 1937, when the threat of division was low, he estimated the number of appeals from Washington, Oregon, Montana, Idaho, and Alaska as 111 per year, which he did not deem significant enough to warrant a split. Such numbers could not be padded easily by alternate configurations, since a division of California—the easiest way to balance the caseloads of the two proposed circuits—would lead inevitably to conflicting interpretations of state law.^[20]

Just as quantitative data supported the need for more judges, so too did qualitative factors. The court sat predominantly in San Francisco, although once per year it held a session of approximately ten days that was divided between Seattle and Portland. Addition of two more judges would enable the court to fix regular panels to sit in Los Angeles, San Francisco, Portland, and Seattle, with Senior Judge Wilbur "sitting less frequently than the rest, watching between the two groups to see that there was no conflict of opinion." The addition would also rid the court of the "unsatisfactory" situation of needing to use district judges to fill out panels. The occasional use of district judges had not obviated the problem of delay, but Denman carefully couched his criticism in administrative terms. The district judge who sat by designation, he testified, "has to hurry back to his crowed district. [He] has no library in his temporary chambers. He has no law clerk, nothing but a stenographer. He is not equipped for that work. Our work as a supreme court is intensive and requires permanent resident judges always ready for conference."^[21]

Denman's testimony signaled a significant change in the practice of the federal courts. Ever since two district judges had conducted the very first argument session of the Ninth Circuit, district judges had participated frequently in circuit court of appeals panels. In the first decade of the twentieth century, when the court of Gilbert, Ross, and Morrow was at the height of its powers, the use of district judges to complete appellate panels ebbed somewhat, but it never ceased. As the caseload of the court increased after World War I and the original corps of judges began to retire, district judges again contributed mightily to the Ninth Circuit. By the early 1930s, when the growth in filings combined with the shortage of judges to produce a crisis of serious proportions, district judges assisted Wilbur, Sawtelle (until his death in 1934), and Garrecht between 1931 and 1935. The addition of five new members to the court between 1935 and 1937 then subtly changed the nature of judicial functions. Whereas in the Ninth Circuit's first two decades, circuit and district judges routinely performed both trial and appellate duties, by the mid-1930s they had become more specialized. Circuit judges heard virtually

all appeals in the Ninth Circuit without the assistance of district judges. And whereas Gilbert, Ross, and Morrow had sat frequently as trial judges until the end of 1911, when the old circuit courts were abolished, the circuit judges of the 1930s and 1940s almost never tried cases.^[22]

If Denman's effort to secure two additional Ninth Circuit judgeships contributed to the transformation in judicial functions, he did not highlight that fact for Congress. Instead he presented the legislation as being mandated simply by administrative need. Denman approached his lobbying with an uncommon zeal undampened by his colleagues' lack of enthusiasm. Wilbur and Garrecht apparently doubted whether Congress would authorize the additional seats. They must have viewed Denman's lobbying as a waste of time and money. In a letter to Wilbur, Garrecht conceded that "all the judges here (four) agree that the volume of the court's business warrants an increase in the number of judges, but we are not agreed as to the method that we should follow in order to secure results." Denman attributed this position to "a defeatist feeling that they will not get attention and will not get relief." Indeed, Denman held Wilbur partly responsible for the court's urgent need for two more judges. In a typically direct letter to his colleagues on the court, Denman complained that "our Senior not only made no recommendation for relief to the Judicial Conference but actually joined in the Conference report (1934) that all was well with the Circuit Courts of Appeal." Denman's determination to see this reform through reflected a spirit that could easily cause offense but whose aims were generally admirable: "If I fail this time here I shall come on again and again, if necessary, if, as now, I can borrow, or, later have the money for the expense."^[23]

Denman did not fail. Congress authorized the appointment of two additional Ninth Circuit judges in April of 1937. In a letter he wrote but did not send, Wilbur gave Denman sole credit for the bill's passage: "I think you realize that none of the judges here believed that Congress would pass legislation at this time for additional judges in this circuit and we all realize that it is wholly due to your efforts that you have gotten so far along with it. As you know, my view did not accord with yours in regard to the presentation of the matter to Congress but that question need not be discussed further at this time." Wilbur may well have decided not to send the letter because a later paragraph chided Denman's "continued absence" as an "embarrassment in regard to some of the cases under submission." With the authorization of two more judges, any backlog caused by Denman's lobbying was quickly dissipated. In two and one-half years, the Ninth Circuit had grown from

two sitting judges to seven, becoming for the moment the largest circuit court of appeals.^[24]

III. Proposals to Divide the Ninth Circuit

The court benefited immediately from the added judgeships, which were assumed by Albert Lee Stephens and William Hearly. In 1937, 332 cases were docketed, with 310 decided and 228 pending. Two years later, only 150 cases were pending, even though more were docketed in the meantime. Put another way, the court terminated 363 cases in 1938 and 352 in 1939. The slight drop in filing in 1938 and 1939, with 290 and 347 docketed respectively in those years, contributed to this increased efficiency.^[25] A second benefit of the additional judgeships was less quantifiable. The court now was more representative of its geographical area: Francis Garrecht of Spokane, Bert Haney of Portland, and William Healy of Boise came from the northwestern quadrant of the circuit; Curtis Wilbur and William Denman provided the San Francisco perspective; and Clifton Mathews of Phoenix and Albert Lee Stephens of Los Angeles represented the southern tier. Of the non-San Francisco judges, only Healy and Mathews relocated to the court's Bay Area headquarters. With this geographical spread, the court was well positioned to handle hearings in a number of different locations without unduly disrupting the judges' lives.

Despite being well represented on the Ninth Circuit, litigants in the Pacific Northwest nevertheless felt aggrieved. A single annual session was shared between Portland and Seattle, and even that session was far from ideal. In 1938, the Seattle Bar Association formally petitioned Senior Judge Wilbur to assist in expediting completion of the new federal courthouse in Seattle. Such facilities would enable the court to sit there more often (without relying on temporary, inadequate chambers) with the sessions held "in rooms more appropriate, the existing courtrooms being so noisy that counsel and witnesses are heard with difficulty."^[26] Completion of the new courthouse was to play an important role in heading off plans to split the Ninth Circuit, a proposal that Oregon United States Attorney Carl Donaugh had advanced in 1937.

Details of Donaugh's proposal are sketchy. Apparently at some point in 1937 he wrote to Attorney General Homer Cummings advocating division of the Ninth Circuit. When an assistant attorney general sent the Ninth Circuit judges a copy of Donaught's letter, they protested hotly. In an unsigned report that bears all the indicia of Denman's authorship, the

court launched a full broadside at the U.S. attorney's proposal, particularly for his failure to notify the judges and his use of erroneous data that inflated the need for a new circuit. This proposal threatened to undo the benefits of the added judgeships. According to the court's statement, each judge in the new Eleventh Circuit would have only 34.4 opinions each year to write, whereas the remaining four in the Ninth Circuit would have 77.5 each.^[27]

As proponents of both sides of the issue collected data, devised arguments, and drafted legislation, builders were busy at work on court-houses in Los Angeles and Seattle. Completion of these new facilities in 1938 and 1939, respectively, would arguably have the greatest impact on the future success of proposals to split the circuit. These buildings ultimately enabled the judges to "ride circuit" more effectively in the fashion envisioned in the 1789 Judiciary Act for Supreme Court justices. The Seattle and Los Angeles courthouses would provide the judges with temporary chambers, library support, and courtrooms. Instead of placing the travel burdens and costs on most litigants, the government and its officers would contribute a greater share to the expense and hardship associated with travel to Ninth Circuit hearings. Although some attorneys in the Pacific Northwest were particularly restive in advocating creation of a new circuit, not surprisingly, perhaps, others opposed such efforts because of the attractions of spending a few days in San Francisco at clients' expense.^[28]

The strongest effort to divide the court was made in 1940 by Judge Bert E. Haney. Apparently at the behest of some members of the Oregon bar, Haney issued a pamphlet from his chambers in October of 1940 advocating splitting the Ninth Circuit into two: one to consist of California, Nevada, Arizona, and Hawaii; the other to comprise Oregon, Washington, Idaho, Montana, and Alaska. Perhaps to insulate himself from criticism, Haney commissioned the pamphlet from his law clerk, Vernon J. Veron, and published it under the clerk's name.^[29] This pamphlet served as the seed for bills proposed in Congress the following year.

Veron offered four reasons for dividing the circuit, each of which was to arouse rancorous debate. He first contended that the Ninth Circuit covered too large an area, one twice the size of the next biggest—the Tenth—and twenty-five times greater than the smallest—the First. The principal difficulty of the circuit's vast size was the travel cost and burden on litigants from the Pacific Northwest, 80 percent of whom had to journey as far as 1,800 miles round trip to argue in San Francisco before the court. Veron derived this fact from an estimate that only one-fifth of

the cases that originated in the Pacific Northwest were heard in the annual sitting in Seattle and Portland.^[30] The solution seemed obvious: to base a new circuit in Seattle, so litigants would not have far to travel.

Veron next maintained that the present seven-member court was unwieldy, since any two judges could theoretically bind the court to a particular view of the law even though the other five disagreed with it. Such a situation had already arisen in the case of Lang v. Commissioner . In Lang , a case raising an issue of estate tax determination, the panel of Denman, Mathews, and Healy disagreed with the controlling precedent established by an earlier case that had been decided by a split decision with Haney and Garrecht in the majority and Wilbur in dissent. Because Lang required application of this unpopular precedent, Denman's panel certified the question to the Supreme Court, out of a concern that a minority of two could bind the other five members of the court.^[31] When the case returned to the Ninth Circuit, Denman explained that the 1891 statute establishing the circuit courts of appeals did not authorize sittings of panels greater than three judges.

Any other interpretation of the [Evarts Act] would lead to grave difficulties of administration. If, because there are four or six circuit judges in a circuit, a four or six-judge court were deemed created by Congress, then each judge may well have the right and the duty to demand his place in the hearing on each appeal. This would lead to the embarrassment of an evenly divided court. More important still, it would increase so greatly the amount of work of each individual judge that it would cause again the arrearages which the addition al judges have been created to remove.^[32]

If Denman's fears appear quaint to modern observers accustomed to large en banc panels, the concerns of adequately administering the federal courts of appeals as they grow in size nonetheless seem to endure. Administrative concerns that in succeeding decades would become routine trifles seemingly presented fundamental questions to Judge Haney and his law clerk in 1940: "[W]ho should have authority to designate who shall and who shall not sit in any particular case? Who should have authority to determine when any particular judge should sit? Could all judges sit in one case at the same time?"^[33]

A third reason for division cited by Veron was that three-judge courts impaired the Ninth Circuit's efficiency. Such courts were authorized in 1910 during the relative infancy of the circuit courts of appeals, when Congress was reluctant to vest full power in the appellate courts to pass on constitutional issues. Accordingly, Congress required a three-judge court, composed of one circuit judge and two district judges, to rule on

interlocutory injunctions against enforcement of state statutes by state officers. Decisions made by this panel were appealable directly to the Supreme Court. In subsequent decades Congress expanded the courts' jurisdiction to include injunctions of administrative orders by an agency acting under a state statute, hearings on permanent as well as interlocutory injunctions, and challenges to the constitutionality of congressional legislation.^[34]

As the situations calling for three-judge courts multiplied, so did the amount of time spent by judges traveling to hear these suits. Veron calculated that for each three-hour hearing attended by a three-judge court in the far-flung Ninth Circuit, the circuit judge had to travel between three and six days between the hearing site and San Francisco. These cases—an average of four per year—required over sixteen travel days annually. The solution proposed by Veron was, not to restrict the situations in which such courts were needed or to abolish them altogether, as Congress finally decided in 1976 for all cases except those concerning congressional district apportionment, but to split the circuit to decrease the distances traveled by circuit judges from a Seattle headquarters to the hearing site.^[35]

Although the three-judge cases must have been an irritant to the circuit judges, the marginal utility of Veron's solution hardly justified dividing the circuit. He thus offered a fourth rationale: the increasing amount of litigation in the states that would compose the contemplated Eleventh Circuit. Unable to show a significant growth in the number of cases originating in the Pacific Northwest, however, Veron clung to the tenuous thesis that the region's people "are more litigious than others." He calculated that the proposed circuit would outrank eight others in per capita litigation. Even if the proposed new circuit's population would be lower than that of every other circuit, Vernon reasoned, "[i]t follows . . . that the per capita litigation displaces total population as a test." In any event, the base level of population and amount of litigation were "sufficient" to justify a split.^[36]

Driving the proposal seemed to be disgruntlement among the bar and its clientele over the amount of travel required to litigate appeals in the Ninth Circuit. As the Seattle Post-Intelligencer editorialized, "Three members of the present ninth circuit were appointed from the Pacific Northwest and retain residences in its states. It is certainly more reasonable to ask them to establish permanent quarters in Seattle or in Portland than to require lawyers from this region to journey to San Francisco whenever a case is carried into the circuit courts." The Oregon

State Bar Association echoed these concerns, focusing on the "unduly large cost of litigation" for parties involved in Ninth Circuit appeals. At its meeting in October, 1940, the Oregon Bar passed a resolution calling for the division of the Ninth Circuit. The State Bar of Washington and the Board of Governors of the Montana Bar approved similar resolutions, but the Alaska Bar opposed division.^[37]

Although the Ninth Circuit judges may have heard of these actions by the various state bars, they apparently did not yet know of Veron's pamphlet or of Haney's role in promoting division. Indeed, in June, 1941, at the Ninth Circuit Conference—the statutory venue for advising Congress on the ways and means of "improving the administration of justice" in each circuit—Haney had not said a word about splitting the circuit. Instead, he had circulated his pamphlet to lawyers and newspapers in the Pacific Northwest without sharing it with his colleagues. As Denman later complained, "Here is our friend sitting with us all these days, never telling us what is going on, and we find this situation." If Haney suffered any angst in deciding whether to distribute Veron's pamphlet among his colleagues, he apparently had none in sending it to Senator Homer T. Bone of Washington, who translated its policy directives into legislation.^[38]

On July 28, 1941, Bone proposed a bill to divide the Ninth Circuit. The new Ninth Circuit would consist of the federal districts in California, Nevada, Hawaii, and Arizona. The new Eleventh Circuit would comprise the districts of Alaska, Idaho, Montana, Oregon, and Washington. The bill authorized the Ninth Circuit to sit in San Francisco and Los Angeles and the Eleventh Circuit to hear cases in Portland and Seattle. The bill assigned the existing Ninth Circuit judges to these respective circuits based on their current residency. Accordingly, Garrecht, Haney, and Healy would sit on the new Eleventh Circuit, and Wilbur, Denman, Mathews, and Stephens would remain Ninth Circuit judges. A similar bill was proposed in the House of Representatives by Warren Magnuson, also of Washington.^[39]

IV. Judicial Opposition to Division

With the introduction of these bills, the debate over splitting the circuit left the realm of theory and entered a more overtly political phase. The San Francisco Recorder joined the discussion by observing that the number of cases docketed in the proposed new Ninth Circuit was double that of the proposed Eleventh Circuit. According to the newspaper,

the three Eleventh Circuit judges would be responsible for approximately twenty-one opinions each, whereas the Ninth Circuit jurists would have to write fifty opinions each to keep the court current.^[40] If the Bone bill passed, the new Ninth Circuit would experience a crunch like the one that had required authorization of more judges during the 1930s.

On August 12, 1941, the court issued an announcement that it hoped would derail the momentum building toward division. The judges published a new internal operating rule requiring an additional sitting in Seattle and Portland. Henceforth the court would conduct sittings in September and in March or April in those cities. The court thus hoped to remove "the most powerful argument" for dividing the circuit, the expense of traveling from the Pacific Northwest to San Francisco for hearings. It could no longer be accused of giving insufficient attention to the needs of that region's litigants. Significantly, every judge signed this internal operating rule amendment except Haney, who decried it as a rank political ploy: "Believing it to be apparent that the Court was being induced to officially enter the field of politics, I did not attend the meeting in San Francisco, and declined to sign the order mentioned."^[41]

A week later, on August 20, 1941, Denman issued a lengthy pamphlet attacking Haney's proposal to split the circuit. With the attention to detail and analysis of relevant statistics for which he had undoubtedly become renowned on Capitol Hill, Denman proceeded to rebut the arguments in Veron's pamphlet point by point, exposing the "errors and omissions of fact which must have had a compelling influence on the members of the bar and the newspapers to whom they were given."^[42] Denman began his dissertation by elucidating those features of the Ninth Circuit's administration of justice that made it different from other circuits:

Unique among the circuits, (1) we have practically abolished terms of court by making the last day of the term the day before the beginning of the next; (2) the court sits in fifty of the fifty-two weeks of the year—the seven judges arranging their vacations so that this is possible. . . . (3) Unlike in some of the understaffed circuits, [it] is the fact that practically every case decided on the merits is upon written opinion, thus making certain for litigants the disciplined reasoning of its judges. . . . With the seven judges we now have the necessary judicial energy (4) to study our briefs before hearing; (5) to grant freely extended time to counsel for argument; (6) to give a true conference between bench and counsel in the course of the presentation of the appeal; (7) thereafter to have, before decision, a conference (and often conferences) in chambers of three fully prepared judges; (8) to assign the case for opinion

writing only after reaching a decision based upon such consideration; and (9) for the careful study by the other two judges of the final form of the opinion before it is handed down.^[43]

From a high of 257 cases pending in 1935, the Ninth Circuit had steadily reduced its arrearage to 150 in 1939, two years after Stephens and Healy joined the court. By the end of the fiscal year, in June of 1941, the court had decided all but fourteen cases, and by the end of July only five were left. Having addressed its longstanding backlog of cases, the court could no longer be accused of being inefficient.^[44]

With this preface, Denman began to dissect Veron's arguments. He observed that the principal concern articulated by advocates of the split—travel burdens—was no longer relevant. Adoption of the new rule meant that henceforth all Pacific Northwest litigants could argue their appeals in Portland or Seattle. Haney's complaint that this rule was a "political ploy" suggested that perhaps unarticulated intangibles, such as the Pacific Northwest's historic disdain of California domination, might be operating. Local bias was difficult to combat or rebut, and no one in this era was unseemly enough to suggest that prejudice justified reform. Haney's position on this issue was somewhat curious. It might be expected that he would approve the new rule both as a concession to the bars of the Northwest and as personally advantageous, but instead he opposed holding additional sittings outside San Francisco. He wrote that the value of such sessions was "too disproportionate to the cost in money and drain on the strength of the Judges."^[45] He thus opposed excessive travel for both litigants and judges without proposing any alternative to division.

In the other judges' minds, the alleged need for a new circuit was based on decreasing litigants' travel burdens. Once that problem had been redressed, the question remained whether a division could be justified as an administrative matter. Caseload statistics and comparisons to other circuits became the focal point of the debate. Denman alleged that filings in the proposed Eleventh Circuit had actually decreased during the preceding three years. Dividing the last twelve years into four segments, Denman contended that the average number of appeals filed in these states had declined from 117 in 1930–1932 to 90 in the 1939–1941 period. Based on the last triennium, the Pacific Northwest Circuit's caseload would be minuscule, only 78 percent of the "greatly overstaffed" First Circuit, which ranked last among the circuits. Indeed, the "admitted excess of judge power [in the First Circuit] has led to repeated proposals to transfer to it two of the states of the adjoining" Second Circuit.^[46]

Veron also had attempted to justify creation of a new circuit by contending that the number of cases generated in the region was equivalent to the average number filed during the entire Ninth Circuit's first two decades. This kind of appeal to history struck Denman as somewhat foolish. Not only did it ignore basic developments in transportation and communication that increased judicial efficiency, it also failed to acknowledge the tremendous amount of trial work that had been done by Ninth Circuit judges in the old circuit courts. According to Denman, the court's minutes from 1905 recorded that William Morrow had presided over 628 trials in circuit court and heard the appeals of another 122 cases in the circuit court of appeals.^[47] By omitting mention of this double function, Veron greatly underestimated the judicial work of the circuit judges during the first two decades under the Evarts Act.

As a corollary to this caseload argument, Veron had calculated, by dividing the 250 working days each year by the 100 average cases likely to be heard in the new Eleventh Circuit, that each judge would have two and one-half days to consider a case and write an opinion on it. Denman made short work of this argument by noting that full judicial work would be required on fewer than 100 cases. Veron's analysis erred by anticipating that every case docketed mandated a full hearing and written opinion. Of these 100 cases, over 25 percent were disposed of by stipulation or dismissed without needing any expenditure of judicial effort. The proposed Eleventh Circuit would therefore actually have approximately 71 cases that required a hearing and an opinion. Denman did not even mention that the time necessary to dispose of a case was significantly less than two and one-half days when the judge was not responsible for writing the opinion. He also ridiculed the proposal by observing that if an eleventh circuit could be justified on the basis of 71 cases per year, the United States ought to have thirty-three circuit courts of appeals rather than just the ten in existence. Such a ratio of workload to personnel would require doubling the size of the judiciary. "At the risk of mauling the obvious," Denman added sardonically, "it may be remarked that such an addition of judicial power would cost $20,000 per judge or for 50 about $1 million per annum."^[48]

If the caseload generated in the new circuit failed to justify division, there still remained the issue of the vast geographical area covered by the Ninth Circuit. Including the territories of Alaska and Hawaii, the Ninth Circuit embraced an area of 1,372,461 square miles. The next largest circuit, the Tenth, was only 41 percent as large. "All this seems irrelevant," Denman commented, "once we consider that with the sessions

now provided in Portland and Seattle no litigant need travel any farther to have his case heard" in the proposed Eleventh Circuit "than he does in the present [N]inth."^[49] By adjusting its internal rules to accommodate litigants' travel burdens, therefore, all the Ninth Circuit judges except Haney hoped to avert a division of the circuit.

The only arguments that remained arose from the alleged administrative difficulties encountered in the large circuit. In a typically drawn three-judge appellate panel, if two judges agreed on the law and the third dissented, the decision of two judges would bind four other judges who had not sat on the appeal and who might disagree with the panel majority. This problem, Denman suggested, did not arise in the Court of Appeal of Great Britain, where panels of two or three heard appeals. That court had twenty-five judges eligible to hear appeals, and Denman knew of no complaints about British justice arising from a pair of judges binding the entire Court of Appeal. Moreover, in his six years as circuit judge Denman could remember only one case in 1,200 in which a majority of the Ninth Circuit judges had formally complained about a decision by a two-member panel majority. In a sense, the San Francisco judge was being disingenuous in denying the existence of the problem that Veron had raised. If the New Deal cases were any guide, a majority on the Ninth Circuit had seemingly suffered in collective silence as the court's two most conservative judges—Wilbur and Mathews—were often able to hamstring the president's recovery program. Of course, one solution to this problem, which Congress had authorized and the court employed, was to certify the question to the Supreme Court.^[50]

Another alternative, not discussed by Denman and Haney at the time, emerged in December, 1941, when the Supreme Court sanctioned en banc panels for circuit courts of appeals, thereby enabling all members of the court to sit in review of a case. In Textile Mills Securities Corporation v. Commissioner , the Supreme Court upheld the Third Circuit's use of an amendment to its internal rules by providing the option of en banc sittings in certain contexts. In Lang v. Commissioner , the Ninth Circuit had read the Judicial Code as precluding this option. The Supreme Court's decision to side with the Third Circuit's statutory interpretation by authorizing en banc sittings of the circuit courts of appeals would in time prove to be a major innovation in the federal court system. It enabled Congress to add circuit judgeships without fear that pockets of powerful minorities could bind majority members of the court through deviant panel decisions. When the Ninth Circuit heard its first appeal en banc in 1942, the judges of the court had little reason to know that this

experiment would provide a source of stability and continuity in the circuit for more than three decades and that, as the numbers of judges grew greater still in the late 1970s, the procedures adopted for a limited en banc would themselves supply ammunition to both sides in the ongoing debates over dividing the circuit.^[51]

The Supreme Court's ratification of en banc hearings would directly address the Haney-Veron concern over intracircuit conflicts of opinion. The Ninth Circuit's adoption of the rule requiring an extra Northwest sitting had already ostensibly mitigated the issue of the burdens of travel. But the release of Denman's rebuttal pamphlet aroused media attention, bringing the debate between Haney and his San Francisco colleague to public view. On August 29, 1941, "Denman Criticizes Judge Haney" ran as a headline in the San Francisco Chronicle . The paper cited a letter Denman wrote to Senator Frederick Van Nuys, the chairman of the Senate Judiciary Committee, attacking the Haney-Veron analysis. A war of words ensued. The press surely placed the dispute in a more personally antagonistic light than the correspondence between Denman and Haney reveals. Nevertheless, the Chronicle reported that long-time court watchers "here do not recall a time" when one of the judges "took such a strong stand against a colleague as has Judge Denman." In his long legal career Denman rarely, if ever, backed down from a fight, and this controversy was no exception. Haney could be equally stubborn. Despite the adverse publicity in local newspapers, the two jurists continued to battle over the preservation of the existing Ninth Circuit.^[52]

Meanwhile, the editorial positions of newspapers around the circuit suggested conflicting regional viewpoints. The San Francisco Chronicle wrote that Denman "took up the cudgel" against Haney's proposal. The Portland Oregonian quoted Haney as advocating a new Pacific Northwest circuit "'despite the illusory statements of my colleague, Judge William Denman, San Francisco.'" The Seattle Post-Intelligencer merely noted that the judges' disagreement over the proposal "was accentuated in Portland yesterday." All three newspapers, however, reported the statements issued by both chambers. Denman accused Haney of making "substantial errors" and being "misinformed" in his pamphlet advocating a new circuit. Haney was certainly entitled to advocate the "destruction of the historic Ninth Circuit," Denman conceded, but the Oregon judge should not be applauded for publishing erroneous statistics.^[53]

The row escalated when the California State Bar held its annual meeting at Yosemite National Park in September of 1941. The Bar considered a resolution to disapprove the Bone and Magnuson bills "as

legislation destructive of the service now rendered by the United States Court of Appeals for the Ninth Circuit, and creating an unnecessary Eleventh Circuit at an unwarranted cost to federal taxpayers." Denman and Haney received invitations to debate the issue at the meeting, but Haney politely declined. He was sitting in Portland beginning September 15 and feared that he would be unable to get to Yosemite in time. Haney did respond, however, to Denman's pamphlet in a letter to the State Bar. Denman's charge of statistical inaccuracy, particularly the allegation that Veron had overestimated the caseload in the Northwest, had stung. Haney contended that the numbers differed because Veron had analyzed a ten-year average given by the clerk of the Ninth Circuit, whereas Denman had broken down a twelve-year period into four three-year chunks. For reasons that no one seemed able to explain, the most recent three-year period showed a marked decrease in the caseload originating in the Pacific Northwestern states. Haney accused Denman of minimizing that region's litigation by emphasizing this statistically anomalous period when the ten-year average was 100.8 cases per year. This figure comported with the annual average in the twelve-year period chosen by Denman, which was 100.4 cases per year.^[54]

Denman was present at the California State Bar Association meeting and spoke immediately after Haney's letter was read. He began with a trenchant point: "[T]his is not merely one judge disagreeing with another judge as to the destruction of the Circuit. It is six judges of the Circuit disagreeing with another judge. The other five will express their views before the Senate and the House Committees, that is if this bill ever comes up for consideration." Nor was Denman persuaded that the new rule requiring an additional sitting in the Northwest was a "political question," as Haney had asserted. To Denman, this rule demonstrated the court's accountability to Congress and the public for passage of the two-judge legislation in 1937. That assistance merited action by the court to improve service to litigants. Such an internal reform by the court should not be considered "politics," Denman continued, "unless you take politics in the broadest sense of the term and include the sometimes exceptional performance of the promises you give to legislators when they grant you legislation."^[55]

Denman also adroitly handled Haney's contention that the new circuit would hear one hundred cases per year. Regardless of whether the average was ninety or one hundred docketings per year, a large number required no judicial action. Denman repeated the point he made in his pamphlet that approximately 25 percent of all docketings were dismissed

or settled by stipulation without requiring any judicial action. The delegates to the California State Bar meeting were undoubtedly predisposed to resist division, but Denman's arguments did help to solidify this sentiment. A motion carried that the bar opposed the Bone and Magnuson bills. In language highly reminiscent of Denman's August pamphlet, the resolution proclaimed the unfair effects of division on the southern states remaining in the Ninth Circuit. The court "will be required either to accumulate heavy arrearages," it concluded, "or decide its cases without giving its litigants the full appellate process."^[56]

The debates over splitting the Ninth Circuit publicly signaled for the first time the court's awareness of its own politics. The disagreements between Ross and Gilbert over natural resource cases earlier in the century had seemed almost as if they were occurring in an academic setting by contrast with the highly visible quarrel between Denman and Haney over dividing the Ninth Circuit. By the early 1940s, the judges' conceptions of the court had changed, as institutionally it had been subjected to numerous influences through its first half-century. No longer was the court a somewhat monastic institution in which virtually every decision about court administration was kept private. The Denman-Haney debates announced that the judges knew they were accountable to the public in dispatching their work in a timely and just manner.