ALVIN B. RUBIN, Circuit Judge:
The right to counsel, safeguarded by the constitutional guarantee of due process of law, includes the right to choose the lawyer who will provide that representation. The defendants in each of these cases employed the brother-in-law of the judge before whom the case was pending. That judge then assigned the case to another judge, who found that in each case, the defendant had engaged the judge’s relative as a strategem in order to disqualify the judge and that the employment was a sham. We hold that the judge to whom each case was initially assigned was required to disqualify himself as soon as he became aware that his brother-in-law had been enrolled as counsel. He, therefore, should not have taken any further action in either case, and the reassignment order was improper. Accordingly, we remand the cases for reassignment in an appropriate manner.
, Because the motion to disqualify the lawyer will be pending on remand, and because it has been fully briefed, we address the merits to provide guidance to the district judge. We consider a litigant’s right to be represented by counsel of his choice, the lawyer’s professional obligation expressed in the Code of Professional Responsibility, and the need to preserve the integrity and efficiency of the judicial system. We conclude that, if the district court should find in either case that the sole or primary motive for retaining the relative of the original judge was to disqualify that judge, the lawyer must be disqualified.
I. Facts
A. No. 82-2330, McCain v. Texas Power & Light Co.1
In 1975, Tilmon McCuin filed a class action in the Sherman Division of the Eastern District of Texas, alleging that Texas Power discriminates against black employees. The case was automatically assigned to Chief Judge William Wayne Justice, who at that time was the only judge in the Eastern District assigned to the Sherman Division. The Eastern District has several other divisions: Beaumont, Tyler, Marshall, Paris, [1258]*1258Texarkana, and Lufkin. Because Texas Power does business in forty-eight counties in northeastern Texas, it might have been sued on this claim in any division of the Eastern District. The district court intimated that the plaintiffs may have chosen to sue in the Sherman Division because cases in that division were assigned only to Chief Judge Justice and because they may have thought he would be more favorably disposed to their cause than other judges.2
The case proceeded slowly. In 1978, the plaintiffs filed their sixth amended complaint alleging for the first time that Texas Power’s employment practices also discriminate against females. At this stage, the Equal Employment Opportunity Commission was permitted to intervene. The plaintiffs had moved for class certification in 1975, but no certification hearing was held until 1980, by which time a considerable amount of discovery had taken place. Chief Judge Justice granted the motion to certify a class of plaintiffs in September, 1981.
Two months later, in November 1981, more than six years after suit had been filed, Texas Power requested that the Clerk add as co-counsel an additional lawyer, J. Mike Rowan, who is the brother of Chief Judge Justice’s wife. The plaintiffs then moved to disqualify Mr. Rowan as counsel and for the “limited recusal” of Chief Judge Justice for the purpose of ruling on its motion. Chief Judge Justice signed an order assigning the entire case to Judge Robert Parker “for such action as may be required.”
The local rules contain no provision for reassignment of cases if a judge is disqualified. Two judges, including Judge Parker, sit in the Beaumont Division. Chief Judge Justice and Judge William M. Steger are now, and were in 1981, assigned to both the Tyler and Sherman Divisions.
After a hearing, Judge Parker found that Texas Power had “intentionally created a conflict under the [disqualification] statute, and now seeks to benefit from the statute’s protection.” 538 F.Supp. at 314. He noted that, although Texas Power had employed Rowan on a retainer for almost a decade, the case had been pending for six years before it sought to associate him in the litigation. He added: “[Rowan] has no reputation for expertise in class actions or employment discrimination law, and virtually none in litigation. His primary qualification as a trial lawyer arose with the enactment of amended [28 U.S.C.] section 455, and consists of his ability, up until now, to assist litigants in removing themselves from Judge Justice’s purview.” Id. He noted that, in a case tried before him on the same day as the hearing in this case, Chief Judge Justice had recused himself after Rowan was associated and Rowan had thereafter made no appearance in the proceeding. He took judicial notice that the City of Tyler had associated Rowan in the Dow case, which is companion to McCuin in this appeal, and concluded: “The practice is fast becoming epidemic.” Id. He characterized employment of Rowan as a “ploy” and added, “[I]t suffices to remark that it is most certainly a sham.” Id.
He then concluded: “A litigant should not be permitted to utilize a disqualification issue as part of his trial strategy.” Id. (quoting Potashnick v. Port City Construction Co., 609 F.2d 1101, 1115 (5th Cir.), cert. denied, 449 U.S. 820, 101 S.Ct. 78, 66 L.Ed.2d 22 (1980)). He disqualified Rowan, ordered the removal of any documents on file that are “the product of [his] handiwork,” ruled inadmissible at trial “any discovery or other work that Mr. Rowan has prepared for the case,” and required Rowan to return any fees Texas Power had paid to him. He also ordered the ease “returned to Judge Justice’s court for further proceedings.”
Texas Power challenges these findings of fact as clearly erroneous. See Fed.R.Civ.P. 52(a). Because we conclude that Chief Judge Justice should have taken no action [1259]*1259in the case, even to allot it to another judge, after Rowan enrolled as counsel, we do not assess the findings.
B. No. 83-2115, Dow v. City of Tyler
Dow and 18 other plaintiffs each filed separate suits against the City of Tyler on February 23, 1982. Each suit alleged that the city had violated the plaintiff’s civil rights during a lengthy undercover drug operation carried out in 1978 and 1979. All of the suits were filed in the Tyler Division. The cases were, in accordance with local rule, assigned by random draw. Twelve were assigned to Chief Judge Justice and seven were assigned to Judge Steger.
Represented by attorney Charles H. Clark, the City filed its answer in all 19 cases on March 9, 1982. Two weeks later the City associated Rowan as additional counsel in all of the cases. The City later filed a motion to consolidate all of the cases with the first case filed. Because the first of the 19 suits had been assigned to Judge Steger, all of the cases would be heard in his court if the motion were granted. Unlike many other districts, however, the Eastern District has no rule that provides for the assignment of all related cases to a single judge. The plaintiffs opposed the consolidation motion and it has not yet been decided.
Free access — add to your briefcase to read the full text and ask questions with AI
ALVIN B. RUBIN, Circuit Judge:
The right to counsel, safeguarded by the constitutional guarantee of due process of law, includes the right to choose the lawyer who will provide that representation. The defendants in each of these cases employed the brother-in-law of the judge before whom the case was pending. That judge then assigned the case to another judge, who found that in each case, the defendant had engaged the judge’s relative as a strategem in order to disqualify the judge and that the employment was a sham. We hold that the judge to whom each case was initially assigned was required to disqualify himself as soon as he became aware that his brother-in-law had been enrolled as counsel. He, therefore, should not have taken any further action in either case, and the reassignment order was improper. Accordingly, we remand the cases for reassignment in an appropriate manner.
, Because the motion to disqualify the lawyer will be pending on remand, and because it has been fully briefed, we address the merits to provide guidance to the district judge. We consider a litigant’s right to be represented by counsel of his choice, the lawyer’s professional obligation expressed in the Code of Professional Responsibility, and the need to preserve the integrity and efficiency of the judicial system. We conclude that, if the district court should find in either case that the sole or primary motive for retaining the relative of the original judge was to disqualify that judge, the lawyer must be disqualified.
I. Facts
A. No. 82-2330, McCain v. Texas Power & Light Co.1
In 1975, Tilmon McCuin filed a class action in the Sherman Division of the Eastern District of Texas, alleging that Texas Power discriminates against black employees. The case was automatically assigned to Chief Judge William Wayne Justice, who at that time was the only judge in the Eastern District assigned to the Sherman Division. The Eastern District has several other divisions: Beaumont, Tyler, Marshall, Paris, [1258]*1258Texarkana, and Lufkin. Because Texas Power does business in forty-eight counties in northeastern Texas, it might have been sued on this claim in any division of the Eastern District. The district court intimated that the plaintiffs may have chosen to sue in the Sherman Division because cases in that division were assigned only to Chief Judge Justice and because they may have thought he would be more favorably disposed to their cause than other judges.2
The case proceeded slowly. In 1978, the plaintiffs filed their sixth amended complaint alleging for the first time that Texas Power’s employment practices also discriminate against females. At this stage, the Equal Employment Opportunity Commission was permitted to intervene. The plaintiffs had moved for class certification in 1975, but no certification hearing was held until 1980, by which time a considerable amount of discovery had taken place. Chief Judge Justice granted the motion to certify a class of plaintiffs in September, 1981.
Two months later, in November 1981, more than six years after suit had been filed, Texas Power requested that the Clerk add as co-counsel an additional lawyer, J. Mike Rowan, who is the brother of Chief Judge Justice’s wife. The plaintiffs then moved to disqualify Mr. Rowan as counsel and for the “limited recusal” of Chief Judge Justice for the purpose of ruling on its motion. Chief Judge Justice signed an order assigning the entire case to Judge Robert Parker “for such action as may be required.”
The local rules contain no provision for reassignment of cases if a judge is disqualified. Two judges, including Judge Parker, sit in the Beaumont Division. Chief Judge Justice and Judge William M. Steger are now, and were in 1981, assigned to both the Tyler and Sherman Divisions.
After a hearing, Judge Parker found that Texas Power had “intentionally created a conflict under the [disqualification] statute, and now seeks to benefit from the statute’s protection.” 538 F.Supp. at 314. He noted that, although Texas Power had employed Rowan on a retainer for almost a decade, the case had been pending for six years before it sought to associate him in the litigation. He added: “[Rowan] has no reputation for expertise in class actions or employment discrimination law, and virtually none in litigation. His primary qualification as a trial lawyer arose with the enactment of amended [28 U.S.C.] section 455, and consists of his ability, up until now, to assist litigants in removing themselves from Judge Justice’s purview.” Id. He noted that, in a case tried before him on the same day as the hearing in this case, Chief Judge Justice had recused himself after Rowan was associated and Rowan had thereafter made no appearance in the proceeding. He took judicial notice that the City of Tyler had associated Rowan in the Dow case, which is companion to McCuin in this appeal, and concluded: “The practice is fast becoming epidemic.” Id. He characterized employment of Rowan as a “ploy” and added, “[I]t suffices to remark that it is most certainly a sham.” Id.
He then concluded: “A litigant should not be permitted to utilize a disqualification issue as part of his trial strategy.” Id. (quoting Potashnick v. Port City Construction Co., 609 F.2d 1101, 1115 (5th Cir.), cert. denied, 449 U.S. 820, 101 S.Ct. 78, 66 L.Ed.2d 22 (1980)). He disqualified Rowan, ordered the removal of any documents on file that are “the product of [his] handiwork,” ruled inadmissible at trial “any discovery or other work that Mr. Rowan has prepared for the case,” and required Rowan to return any fees Texas Power had paid to him. He also ordered the ease “returned to Judge Justice’s court for further proceedings.”
Texas Power challenges these findings of fact as clearly erroneous. See Fed.R.Civ.P. 52(a). Because we conclude that Chief Judge Justice should have taken no action [1259]*1259in the case, even to allot it to another judge, after Rowan enrolled as counsel, we do not assess the findings.
B. No. 83-2115, Dow v. City of Tyler
Dow and 18 other plaintiffs each filed separate suits against the City of Tyler on February 23, 1982. Each suit alleged that the city had violated the plaintiff’s civil rights during a lengthy undercover drug operation carried out in 1978 and 1979. All of the suits were filed in the Tyler Division. The cases were, in accordance with local rule, assigned by random draw. Twelve were assigned to Chief Judge Justice and seven were assigned to Judge Steger.
Represented by attorney Charles H. Clark, the City filed its answer in all 19 cases on March 9, 1982. Two weeks later the City associated Rowan as additional counsel in all of the cases. The City later filed a motion to consolidate all of the cases with the first case filed. Because the first of the 19 suits had been assigned to Judge Steger, all of the cases would be heard in his court if the motion were granted. Unlike many other districts, however, the Eastern District has no rule that provides for the assignment of all related cases to a single judge. The plaintiffs opposed the consolidation motion and it has not yet been decided. The plaintiffs filed a motion to disqualify Rowan and Chief Judge Justice referred all of the cases assigned to him to Judge Parker “for such action as may be required.” After holding an evidentiary hearing, Judge Parker granted the motion for the same reasons he gave in McCuin.
II. Judicial Disqualification
Before Congress amended the judicial disqualification statute in 1974, a judge’s kinship to a lawyer appearing before him did not automatically require the judge’s recusal.3 The statute provided simply that the judge was to disqualify himself if the relationship would “render it improper, in his opinion, for him to sit. .. .”4 In a report recommending a change in the statute, a House of Representatives Committee criticized this standard because it “made the judge himself the sole decider of ... the relationships which would be improper and lead to disqualification.”5 Congress, therefore, amended the statute to mandate the judge’s disqualification in cases in which a person related within the third degree to the judge or the judge’s spouse is acting as a lawyer or has a substantial interest in the outcome of the proceedings.6
The legislative history of the amendment to § 455 indicates that Congress’ purpose was to adopt a categorical rule that would be self-enforcing.7 In Po[1260]*1260tashnick, we noted that § 455 makes it the judge’s duty to disqualify himself as soon as he is aware that the grounds exist. The mandate is made inexorable by an express prohibition against waiver.8 The statute applies whenever the judge’s relative acts as a lawyer “in the proceeding,” whether or not he is enrolled as counsel. In United States ex rel. Weinberger v. Equifax, Inc., 557 F.2d 456, 463 (5th Cir.1977), cert. denied, 434 U.S. 1035, 98 S.Ct. 768, 54 L.Ed.2d 782 (1978), we held that one “acts as a lawyer” within the meaning of § 455(b)(5)(h) when he actually participates in the case. There is evidence in the record that Rowan has participated in discovery proceedings in both cases. Rowan’s participation in these cases, therefore, required Chief Judge Justice’s recusal.
Even if, however, Rowan had not actually participated in pretrial discovery, we would be constrained to hold that he began participating in the case when he enrolled as counsel of record. The judge may not wait to see what part the lawyer-relative will play. It would defeat the bright-line purpose of the statute for a judge to continue to preside over a case in which one of his relatives is counsel of record until the moment when that lawyer chooses to change his role from passive to active participation, putting the case under a Damoclean sword that will fall whenever the judge’s relative chooses to cut the hair by emerging from passivity.
This conclusion is reinforced by another section of the statute requiring the judge’s disqualification when a relative has an interest in the case that “could be substantially affected by [its] outcome.” 28 U.S.C. § 455(b)(5)(iii). A fortiori, counsel of record in a case must have such an interest, else why would he enroll? We have held that, when a partner in a law firm that is counsel in a case is related to a judge within the third degree, the partner automatically has an interest that could be substantially affected by the outcome of the proceedings. Potashnick, 609 F.2d at 1113. Hence Rowan, who enrolled as an individual, has a “substantial interest” in the outcome of these proceedings.
In Dow, the City of Tyler argues that, because Rowan’s enrollment as counsel automatically disqualified Chief Judge Justice, he had no alternative but to recuse himself and allow the case to be reallotted in accordance with local practice. Accordingly, the City argues, his order referring the case to Judge Parker should be set aside. Texas Power does not make that argument in its McCuin brief. It did, however, argue in the district court that “the only orders [sic] which [Chief Judge] Justice may properly enter in this case is an order of recusal reassigning this case to another trial judge.” We consider the issue in both cases because disqualification cannot be waived, the issue is solely one of law,9 and it would be unfair to reach different results in these two cases so intimately related.
The business of a court having more than one judge is divided among the judges [1261]*1261as provided by the rules and orders of the court. 28 U.S.C. § 137 (1976). The local rules of court for the Eastern District of Texas contain no provision for the reassignment of cases in which a judge is disqualified. Section 137 provides that, in the absence of such rules and orders, it is the duty of the chief judge of the district to assign cases, but patently that judge has no authority to sign such an order, or any other judicial order, pertaining to a proceeding in which he is disqualified. To permit a disqualified chief judge to select the judge who will handle a case in which the chief judge is disabled would violate the congressional command that the disqualified judge be removed from all participation in the case. It might, in addition, create suspicion that the disqualified judge will select a successor whose views are consonant with his. Congress intended the statutory antisepsis to be absolute in order to avoid any bacterium of impugnment.
We, therefore, must vacate Chief Judge Justice’s order assigning the cases to Judge Parker and remand the cases to the district court. Because Chief Judge Justice is disqualified to conduct any further proceedings in both of the cases, they shall be reallotted in accordance with local practice, if one exists; if not, they shall be referred to the senior active judge of the district as Acting Chief Judge for reallotment.10
Because the merits of the cases have been fully briefed and because no other courts have broken this ground to guide the district judge on remand, we set forth the controlling principles.11
III. Is Judge-Shopping Reprobated?
Forum-shopping is sanctioned by our judicial system. It is as American as the Constitution, peremptory challenges to jurors, and our dual system of state and federal courts. The extension in Article III of federal judicial power to “controversies between citizens of different states,” implemented by statute continuously since 1789,12 permits a plaintiff who might sue in a state court to select a federal forum for the claim. The statutory provision for removal to federal courts of such diversity cases filed in state court permits the defendant to opt for a federal forum.13 Virtually all causes of action created by federal law may be asserted in either a state or a federal court.14 Many claims that may be asserted in the courts of one state may also be asserted in the courts of another. Not only may a litigant frequently select among several jurisdictions, he may, within a jurisdiction, lay venue in more than one court.15
The existence of these choices not only permits but indeed invites counsel in an adversary system, seeking to serve his client’s interests, to select the forum that he considers most receptive to his cause. The motive of the suitor in making this choice is ordinarily of no moment: a court may be selected because its docket moves rapidly, its discovery procedures are liberal, its jurors are generous, the rules of law applied are more favorable, or the judge who pre[1262]*1262sides in that forum is thought more likely to rule in the litigant’s favor.
Even in a particular forum, other tactical measures are available to determine who will be the trier of fact. For most causes of action, the plaintiff in a federal court may seek either a jury trial or a bench trial.16 In these cases, if the plaintiff waives trial by jury, the defendant may veto the bench trial by demanding a jury trial.17 Finally, the very existence of the peremptory challenge in jury cases bespeaks regard for some latitude in selection of the trier-of-fact on the basis of getting rid of those whom the lawyer considers unfavorably predisposed.
In a perfect judicial system forum-shopping would be paradoxical. The same results would obtain in every forum and after every type of trial. But the actual litigation process is not a laboratory in which the same result is obtained after every test. In some situations, such as when a statute of limitations is involved, the choice of forum may determine the rule of law that will be applied. Even when legal rules are identical, justice can be obtained only through human beings, and neither judges nor jurors are fungible. In recognition both of this and of the nature of the adversary, client-serving process, we tolerate a certain amount of manipulation without inquiry into motive.
Some states expressly permit judge-shopping, allowing the parties peremptory challenges to the judge.18 The possibility of permitting this maneuver in federal courts has been broached to the Congress,19 but never adopted. In federal court, the parties clearly have no right to a “judge of their choice.”20
Litigants do have a right to be represented by counsel and this ordinarily implies a right to lawyers of their choice.21 The right to counsel does not, however, entail absolute freedom of choice. Counsel must be a member of the bar and must be admitted to practice before the court in which he appears. He must not have a conflict of interest with another party.22 His employment must not entail disclosure of confidential information.23 The choice is never completely unfettered.
Subject to these general limitations, the right to counsel in criminal cases is expressly guaranteed by the sixth amendment; the right to counsel in civil cases is no less fundamental and springs from both statutory authority and from the constitutional right to due process of law.24 Therefore, disqualification of counsel “is an extreme remedy that will not be imposed [1263]*1263lightly.” Duncan v. Merrill Lynch, Pierce, Fenner & Smith, 646 F.2d 1020, 1025 n. 6 (5th Cir.), cert. denied, 454 U.S. 895, 102 S.Ct. 394, 70 L.Ed.2d 211 (1981).
The right to counsel of one’s choice may be overridden when “compelling reasons exist.”25 The right should be balanced in cases in which it is challenged against the right to “untainted prosecution of the lawsuit” and society’s need to maintain the highest ethical standards of professional responsibility.26 It cannot be exercised without thought also to the needs of effective administration of justice. “[T]he ultimate decision on [delaying a trial for the appointment of separate counsel] must remain with the trial judge; otherwise unscrupulous defense attorneys might abuse their ‘authority,’ presumably for purposes of delay or obstruction of the orderly conduct of the trial. The State has an obvious interest in avoiding such abuses.... When an untimely motion for separate counsel is made for dilatory purposes, our holding does not impair the trial court’s ability to deal with counsel who resort to such tactics.” Holloway v. Arkansas, 435 U.S. 475, 486-87, 98 S.Ct. 1173, 1180, 55 L.Ed.2d 426, 436 (1978) (footnote omitted).
Thus, in a case cited by the Court in Holloway, the Second Circuit affirmed the conviction of a defendant who sought a delay to obtain new counsel when his lawyer became disabled. The court refused to reverse the conviction because the defendant had had a month’s notice of the problem, his lawyer was available for consultation with a new lawyer, and the defense was uncomplicated. United States v. Dardi, 330 F.2d 316 (2d Cir.), cert. denied, 379 U.S. 845, 85 S.Ct. 50, 13 L.Ed.2d 50 (1964). The Dardi court stated: “While the right to counsel is absolute, its exercise must be ‘subject to the necessities of sound judicial administration.’ ” 330 F.2d at 355 (quoting United States v. Arlen, 252 F.2d 491, 494 (2d Cir.1958)). We have long held that a motion by the defendant to substitute counsel on the eve of a trial and then to delay the trial need not be granted even if the reason is an alleged conflict of interest and the defendant’s right to counsel with undivided loyalty.27
When a lawyer enrolls to represent two defendants in a criminal case, the district judge must personally inform both clients of the possibility of a conflict of interest and, even if an actual conflict is not apparent, ascertain whether there is a possibility that a conflict will develop and whether the effective and fair administration of justice would be adversely affected by continued joint representation.28 This inquiry includes the warning that, if a conflict develops at a time when enrollment of new counsel is impractical, the defendant may be held to have waived the right then to seek separate counsel of his choice.
Similar needs for effective administration exist in civil cases. If a party who is offended by a judge’s ruling could disqualify the judge by employing new counsel after the case had been pending before that judge for years, he could force on other litigants and the courts the penalty of starting over before a new judge. Both effective judicial administration and economy of litigation costs require that a case be assigned as soon as possible after it is filed to a single judge who will become fully familiar with the issues and exercise effective [1264]*1264control over the pleadings, discovery, pretrial, and trial. Every one of the 94 federal judicial districts is now on an individual calendar system, assigning each case to a judge soon after its filing. If several cases are filed in different courts, the multidistrict litigation statute permits their consolidation and assignment to a single judge.29 If, after seeing who the judge is or weighing his rulings for a period of years, a litigant could in effect veto the allotment and obtain a new judge by the simple expedient of finding one of the judge’s relatives who is willing to act as counsel, it would become possible for any party to disrupt preparation for, or, indeed, the trial itself.
The drafters of § 455 warned that “each judge must be alert to avoid the possibility that those who would [seek his disqualification] are in fact seeking to avoid the consequences of his expected adverse decision.”30 In light of Congress’ intent and the needs of judicial efficiency, we hold that counsel may not be chosen solely or primarily for the purpose of disqualifying the judge. The district court threatened with such maneuvers need not confine itself to grievance proceedings against errant counsel. “A motion to disqualify counsel is a proper method for a party-litigant to bring the issues of conflict of interest or a breach of ethical duties to the attention of the court.”31 Indeed “a District Court is obliged to take measures against unethical conduct occurring in connection with any proceeding before it.” Woods, 537 F.2d at 810.
IV. The Attorney’s Duty Under the Code of Professional Responsibility: Canon 9
The needs of judicial administration are not our sole consideration. Lawyers are members of a learned profession asserting high ethical standards. The lawyer’s exclusive right to practice is afforded because of the ethical standards of the profession as well as its members’ technical knowledge and specialized skill. An ethical code is not a garment that lawyers may don and doff at pleasure. A lawyer must qualify to practice by satisfying an examining committee of his good moral character and by passing an examination that includes testing for knowledge of ethical precepts as well as substantive rules. Lawyers are not permitted to do everything for a client that he would stoop to do himself had he but their knowledge. A lawyer must not make “any statement or suggestion ... that he can or would circumvent” procedures by which legal matters can be presented in an impartial manner.32 A lawyer should not, therefore, lend himself to a contrivance by which his services are sought not for his ability but solely because his relationship with a judge enables the litigant who employs him to exercise a de facto peremptory challenge to the judge.33
Canon 4 of the Code of Professional Responsibility deals with confidentiality of information gained from clients and conflicts of interest; Canon 5 deals with the lawyer’s duty of loyalty to the client. Considering these as strictures on a lawyer’s conduct [1265]*1265and not merely hortatory appeals to his conscience, federal courts have ordered lawyers disqualified in situations involving conduct proscribed by the Code.34
Other ethical precepts also limit the lawyer’s conduct. Canon 9 of the Code of Professional Responsibility provides: “A Lawyer Should Avoid Even the Appearance of Professional Impropriety.” The purpose of this injunction is to preserve public confidence in the bar and in the legal process.35 Although, like the other Canons, it expresses a norm, not a mandate, we have held that counsel may be disqualified to appear in a matter if his conduct violates the prescription. A two-step analysis determines whether a lawyer is disqualified under Canon 9. First, the district court should inquire whether there is “at least a reasonable possibility that some specifically identifiable impropriety” has occurred. Woods, 537 F.2d at 813. This aspect of the analysis focuses solely on past improprieties.36 Second, the court should consider whether the likelihood of public suspicion of, or obloquy regarding, such an impropriety is sufficiently strong to outweigh the interest of the party being represented by counsel of its choice. Woods, 537 F.2d at 813 n. 12. Disqualification under Canon 9 is required when both conditions are satisfied.
A lawyer’s acceptance of employment solely or primarily for the purpose of disqualifying a judge creates the impression that, for a fee, the lawyer is available for sheer manipulation of the judicial system. It thus creates the appearance of professional impropriety. Moreover, sanctioning such conduct brings the judicial system itself into disrepute. To tolerate such gamesmanship would tarnish the concept of impartial justice. To permit a litigant to blackball a judge merely by invoking a talismanic “right to counsel of my choice” would contribute to skepticism about and-mistrust of our judicial system.
“[Jjudges do not choose their cases, and litigants do not choose their judges. We all operate on a blind draw system. Sometimes, both litigants and judges are disappointed by the luck of the draw. But the possibility of such disappointment is a risk judges and litigants alike must assume....”37
We recognize that a litigant’s motives for selecting a lawyer are not ordinarily subject to judicial scrutiny and that, by permitting inquiry into these motives, we open the door to a host of problems. But a contrary ruling would permit unscrupulous litigants and lawyers to thwart our system of judicial administration. Hence, we make no attempt to predict all the problems that may arise or to provide a compendium of answers. The general rule of law is clear: a lawyer may not enter a case for the primary purpose of forcing the presiding judge’s recusal.
We REMAND both cases to the district court for reassignment. In the interests of judicial economy, both cases should be assigned to a single judge. That judge shall rule on the motions to disqualify Mr. Rowan in accordance with the principles we have set forth. If the motion is denied, the district judge shall either continue with the assignment or reassign the cases to another nondisqualified judge, as efficient administration and the business of the court require. If Mr. Rowan is disqualified in either case, the judge may determine whether that case should be reassigned to Chief Judge Justice.38 Regardless of the out[1266]*1266come, we do not intimate that disciplinary action against Mr. Rowan and counsel who associated him is unwarranted nor does this opinion preclude such action. If an ethical violation has occurred, disqualification of counsel is not the sole sanction available.
REMANDED.