UNITED STATES of America, Plaintiff-Appellee, v. Eric D. PEARSON, Defendant-Appellant, 203 F.3d 1243

HENRY, Circuit Judge. A jury convicted Eric Pearson of the following offenses: (1) conspiring to obstruct commerce by robbery (in violation of the Hobbs Act, 18 U.S.C. § 1951); (2) obstructing commerce by robbery (also in violation of the Hobbs Act); and (3) carrying or using a firearm in connection with a murder (in violation of 18 U.S.C. §§ 924(c)(1) and (j)(l)). The district court imposed concurrent sentences of 240 months for each Hobbs Act violation and life in prison for the § 924 violation. Mr.

Pearson now appeals his convictions and sentences on thirteen grounds, arguing that: (1) the district court erroneously denied his motion for random reassignment of his case; (2) the jury selection system in the Wichita-Hutchinson division of the District of Kansas violated his Sixth Amendment and statutory rights;

(3) Congress lacked the constitutional authority to enact 18 U.S.C. § 1951; (4) the convictions under 18 U.S.C. § 1951, as well as under § 924(c)(1) and (j)(l), violated the Double Jeopardy Clause; (5) the district court erroneously refused to suppress his statements to the police after his arrest; (6) the evidence was insufficient to support the jury’s finding that he committed felony murder as defined by 18 U.S.C. § 1111(

a) and as required for his conviction under 18 U.S.C. § 924(c)(1) and (j)(l); (7) the district court erroneously instructed the jury on the elements of felony murder; (8) the district court erroneously refused to submit his requested lesser included offense instructions to the jury; (9) the district court erred in admitting evidence of his character; (10) the jury instructions improperly omitted an element necessary to prove a violation of 18 U.S.C. § 1951;' (11) the district court violated his due process rights when it denied his motion for a new trial after one of the government’s witnesses recanted; (12) the district court misapplied § 2A1.1 of the Sentencing Guidelines; and (13) the district court was biased against him in violation of his due process rights.

For the reasons set forth below, we affirm Mr. Pearson’s convictions and sentences. I. BACKGROUND We begin by summarizing the events leading up to Mr. Pearson’s arrest and the evidence presented at trial. Then, because a central issue in the case concerns the judicial assignment procedures adopted by the District of Kansas; we discuss the manner in which the case was assigned to United States District Judge Monti L. Be-lot. A. The Robbery and Murder at Mr. Goodcents Two men, one wielding a handgun, robbed Mr. Goodcents Subs & Pastas (“Mr.

Goodcents”) in Wichita, Kansas, at approximately 9:55 p.m. on Monday, February 17, 1997. As the men emptied the cash register and safe, the handgun accidentally discharged, killing Amie Montgomery, the nineteen-year-old shift supervisor who was on duty. The robbers fled with roughly $2,500. Based upon interviews with various informants, the Federal Bureau of Investigation (“FBI”) arrested Eric Pearson (“Mr. Pearson”) and several others in relation to the events at Mr. Goodcents. According to the FBI agents who interrogated him, Mr. Pearson confessed to being involved in the Mr.

Goodcents robbery and implicated his cousin, Dominic Pearson (“Dominic”), and their friend Courtney Martin. • In February 1997, the United States Attorney’s Office filed three separate in-formations in the United States District Court for the District of Kansas charging Eric and Dominic Pearson and Courtney Martin with violating 18 U.S.C. §§§ 2, 924, and 1951 for their roles in the killing of Amie Montgomery and the robbery of Mr. Goodcents. Three days later, a grand jury returned three separate superseding indictments charging the Pearsons and Mr. Martin with those offenses.

On March 12, 1997, the grand jury issued a consolidated superseding indictment adding two other defendants, Deborah Meyer and Gracie Ginyard, another cousin of Eric Pearson. The superseding indictment also added several new counts under 18 U.S.C. §§§ 2, 924, and 1951 against Eric and Dominic Pearson. On April 17, 1997, the government filed a second superseding indictment against the same defendants. Mr. Martin, Ms. Ginyard, and Ms. Meyer pleaded guilty and agreed to testify at Mr. Pearson’s trial. Ms. Meyer, who had been an assistant manager at Mr. Good-cents, stated that Mr.

Pearson, her boyfriend at the time, had .discussed with her various plans to take money from Mr. Goodcents. She testified that she also met with Mr. Pearson and his cousin Dominic and talked about where the restaurant kept its money and when would be the best time to rob it. She stated that before the date of the robbery, the Pearsons left her house intending to rob Mr. Goodcents, but later told her they could not complete the crime because there were police around the restaurant. However, according to Ms. Meyer, Mr. Pearson continued to plan to take money from Mr. Goodcents until February 17, 1997. Finally, Ms.

Meyer testified that Mr. Pearson was not employed during the time she knew him but that he supported himself as a “hustler” and “a pimp [who] just had women.” Rec. vol. VII, at 132. Upon objection from defense counsel, the court instructed the jury to disregard Ms. Meyer’s reference to Mr. Pearson as a “pimp.” However, the court overruled the objection as to the term “hustler.” Ms. Ginyard testified that she had heard Mr. Pearson talking about robbing Mr. Goodcents in January. She further testified that, in February, he approached her while she and Ms.

Meyer were working at the restaurant and asked her to help him stage a robbery of Ms. Meyer when she went to deposit the store’s receipts. Ms. Ginyard stated that she agreed to the plan, but then changed her mind after speaking with Ms. Meyer. According to Ms. Gin-yard, Ms. Meyer subsequently told her that Mr. Pearson was going to rob the store on the evening of February 16, when Ms. Ginyard was working and could ensure that no one would be hurt. However, she noted that Mr. Pearson did not rob the restaurant that night. During her testimony, Ms. Ginyard also identified the murder weapon as Mr. Pearson’s gun. Mr.

Martin testified that, on February 17, he and Dominic had robbed the restaurant while Mr. Pearson waited in the car. Mr. Martin claimed he was at his girlfriend’s house prior to the robbery and that Dominic arrived around 8:30 p.m. Mr. Martin stated that, prior to going to Mr. Pearson’s residence, they drove around, smoking marijuana. According to Mr. Martin, when they arrived at Mr. Pearson’s residence around 9:00 p.m., Mr. Pearson and Dominic conversed in another room before discussing the robbery with him. Mr. Martin claimed that Mr. Pearson assured him that the robbery would go well because Ms.

Meyer had given Mr. Pearson detailed information about how to commit the crime. Mr. Pearson also explained that he could not enter the restaurant because the employees might recognize him. Mr. Martin stated that Mr. Pearson supplied the gun and the clothing for Mr. Martin and Dominic to wear, drove them to Mr. Goodcents, waited in the car, and then drove them away after the crime. Mr. Martin admitted that he was holding the gun when it fired, killing Ms. Montgomery. He also testified that the three of them had split the robbery proceeds.

The government also presented several witnesses who were not involved in the crime. FBI Special Agent Charles Pritch-ett testified that, during an interrogation just after his arrest, Mr. Pearson confessed to robbing Mr. Goodcents with Dominic and Mr. Martin. According to Special Agent Pritchett, Mr. Pearson told him that he drove Dominic and Mr. Martin to Mr. Goodcents, parked on the north side of the store, waited while Dominic and Mr. Martin robbed the store, and then split the proceeds with them. Monie Dyer, a former girlfriend of Mr. Pearson’s, testified that on February 20th, Mr. Pearson paged her.

She stated that she drove Mr. Pearson to his house, where he gave her a rifle wrapped in a blanket, which she took to her garage. She added that, as she was taking the rifle out of the trunk of her car, she noticed something falling out of the blanket. When she looked in her trunk, she saw a handgun and a rifle clip. She claimed that on the following day she disposed of the handgun in a dumpster after she learned of Mr. Pearson’s arrest and after Bruce Dikes (Mr. Pearson’s cousin) told her to dispose of the gun “for [her] own good.” Id. vol. VIII, at 372-78.

After receiving an anonymous phone call, which they later determined was made by Ms. Dyer, the police retrieved a handgun from the same dumpster. Forensic tests on the gun showed that it had fired the bullet that killed Ms. Montgomery. On cross-examination, Ms. Dyer testified that she was not sure how the handgun got into her trunk and that someone who borrowed her car on the twentieth must have put the gun in the trunk. When asked repeatedly who had her car on the twentieth, she responded: I just couldn’t get to my car. I was riding with a friend and my car was at my house.

Matter of fact, I don’t know where my car was that night. But I couldn’t, I couldn’t get to it at that time. I was riding with someone else. I was way across town. I don’t know don’t know exactly. I don’t think [I let any men borrow my car]. People were telling me that [a man named Bernard] was the one that put the gun in my car.... No, [Bernard] didn’t use my car. Id. at 353-54, 358, 359, 368-69. Angela Starks, an employee of Mr. Go-odcents who was present during the robbery, testified to the details of the crime. She stated that she knew Mr. Pearson and would have recognized him had he robbed the store.

She admitted that she did not see Mr. Pearson rob the store and did not see any cars or movement during the robbery. Steve Peterson, the owner of Mr. Good-cents, testified that the restaurant is part of a national chain and purchases goods from both Kansas and out-of-state companies for sale to customers. According to Mr. Peterson, business suffered after the robbery and killing, and the stolen money would have been used to purchase goods produced outside of Kansas. Mr. Peterson stated that he knew Mr. Pearson, because he had taken Mr. Pearson and Ms. Meyer to dinner, and would have recognized him if Mr.

Pearson had robbed the store while Mr. Peterson had been working. After the close of the government’s case, Mr. Pearson called Kenneth Hawkins as an alibi witness. Mr. Hawkins testified that Mr. Pearson was at his house on the night of the robbery from before sundown (around 6:10 p.m.) until 3:00 a.m. According to Mr. Hawkins, Mr. Pearson only left once during the evening, after 10:00 p.m., and was gone for no more than twenty minutes. On cross-examination, Mr. Hawkins admitted that when questioned by FBI agents during the investigation, he did not tell the FBI that Mr.

Pearson was with him on the night of the robbery. He explained his failure to disclose this information by noting that the FBI agents never asked him specifically if he knew where Mr. Pearson was during the Mr. Good-eents robbery. Mr. Hawkins further explained that he was the only person working in his store when the FBI agents came to interview him, and he was busy. After hearing this testimony, the government called rebuttal witnesses, including Kaleb Fowler, who worked at Mr. Go-odcents and knew Mr. Pearson, and Shannon Miller, a former girlfriend of Mr. Pearson’s. Mr. Fowler testified that he called Ms.

Meyer’s house between 7:15 and 7:30 p.m., after dark, on the night of the robbery and that Mr. Pearson, whose voice he recognized, answered the telephone. In her testimony, Ms. Miller identified a letter sent to her in Mr. Pearson’s handwriting. The letter was signed with Mr. Pearson’s nickname, Ace, and sent from the Harvey County jail, where Mr. Pearson was being held. The letter read, in pertinent part: [T]hese feds have been trying to get hell and dirty on me. Telling lies, saying that I’ve said things that I haven’t. Really trying to screw me.

I was sitting here thinking about that night and I remembered at 10:00 I was talking to you on the phone. First we talked on my cell phone. Then I called back on the studio phone so it would be cheaper. That’s when we talked about me coming over. It was 10:00. This is very important to my case. Would you be willing to let them know you were talking to me? Please, baby, it’s important. If they ask how you know it was 10:00 say the news came on while we were talking. Don’t mention us talking about me coming over because I let them know I was at the studio until 2:45 a.m. But we did talk from about 9:50 until 10:10.

I’m gonna call you tonight.... If you’re down when I call I’ll say “are you all good.” You just say yes. Then I’ll know. Id. vol. IX, at 690-91. Ms. Miller averred that Mr. Pearson called her to ensure that she would testify as he asked in the letter, but she refused: “I told him that I didn’t appreciate ... him involving me in this whole ordeal, that I would not lie for him, I would not perjure myself.” Id. at 692. At the conclusion of all the evidence, the jury convicted Mr. Pearson on all counts. At the sentencing hearing the judge overruled Mr.

Pearson’s objections to the pre-sentence report and imposed concurrent 240 month sentences for the Hobbs Act violations and a life sentence for the § 924 violation. During the trial, the judge made several remarks about Mr. Pearson’s character. After a bench conference, the court reporter told the judge that she saw Mr. Pearson making threatening gestures to -a witness. In discussing the court reporter’s observation with the attorneys, the judge referred to Mr. Pearson as “a punk, first class, ... a manipulator ... [who] runs whores ..'. [and] lives off ... women.” Id. vol. VII, at 276-77.

During the sentencing hearing, the district court referred to Mr. Pearson as “a predator, a manipulator who preys on women” and as someone who “has [n]ever done anything decent in his life” and “has nothing going for him.” Id. vol. X, at BB-SS. Observing that Mr. Pearson smirked through the trial as if it were “entertainment,” the court also called Mr. Pearson “repulsive” and a “poster boy for a life sentence in a federal penitentiary.” Id. at 35. Finally, reacting to a disruption in the audience during Mr. Pearson’s sentencing, the district court stated, “Another one of your girlfriends, I assume.

A lot of stupid people - around here.” Id. at 36. When Mr. Pearson’s counsel spoke in his defense, the court, recalling trial testimony, responded: What redeeming qualities are there about someone whose claim to fame is impregnating three different women [and] not supporting his children? [Wjhere do you suppose those children are going to be in 15 or 20 years? Or 25? With mothers who apparently are prostitutes and a father who is spending the rest of his life in a federal penitentiary. Id. at 39-40; see also Rec. vol. VII at 189, 193 (testimony from Ms. Meyer noting that Mr.

Pearson had three children with three different women). B. The Assignment of the Case to Judge Belot When the government filed three separate informations against Eric Pearson, Dominic Pearson, and Courtney Martin in February 1997, the court clerk’s office assigned each case to a different United States District Judge: Eric Pearson’s case was assigned to United States District Judge John Thomas Marten; Dominic Pearson’s case was assigned to Judge Monti L. Belot; and Mr. Martin’s case was assigned to United States District Judge Frank G. Theis.

However, when the grand jury returned the consolidated superseding indictment against Eric Pearson, Dominic Pearson, Mr. Martin, Deborah Meyer, and Gracie Ginyard on March 12, 1997, the court clerk’s office assigned the consolidated case to Judge Belot. Judges Marten and Theis then dismissed the previous indictments against Eric Pearson and Courtney Martin. After the consolidated case was assigned to Judge Belot, the government filed a motion to sever the trials of Dominic Pearson and Courtney Martin from those of Eric Pearson, Ms. Meyer, and Ms. Gin-yard.

The government based its motion on the fact that Eric Pearson, Ms. Meyer, and Ms. Ginyard had each made statements that implicated others and therefore “[i]t would be easier to deal with those people first and deal with others that we don’t have statements from second.” Rec. vol. Ill, doc. 68, at 2. Judge Belot granted the motion to sever, but all of the cases remained assigned to him. Subsequently, Mr.

Pearson filed a motion seeking random reassignment to a new judge, contending that the government had purposely filed the charges against the defendants in such a manner that the consolidated case would be assigned to Judge Belot. Mr. Pearson explained the method employed by the government as follows: This case assignment was not a random assignment. a. Where formerly the order of the accused persons had listed Eric D. Pearson as first on the Information and first on the individual Indictment numbered 97-10026-01; b. On the Superseding Indictment, Eric Pearson was listed second.

Inexplicably, and for the first time in this caSe, Mr. Dominic Pearson was listed first. c. Upon information and belief, the normal practice of the U.S. Attorney is to list the defendant thought the most culpable as the first, or lead, defendant. d. The Clerk’s office did not place the Superseding Indictment in the random assignment pool, because there were already judicial assignments on the related, technically superseded, individual Indictments. e. Instead, the Clerk’s office continued a prior assignment. f.

There were three prior assignments available to the Clerk ... [i.e., the cases pending before Judges Marten, Belot, and Theis]. g. Instead of selecting among the three randomly, or choosing the allegedly most culpable defendant, or choosing to assign the Superseding Indictment to the Judge who already had the first filed Indictment (United States v. Eric D. Pearson, Case No. 97-10025), the clerk based the judicial assignment on how the Superseding Indictment was pled. Rec. vol. I doc. 77, at 2-3. Mr.

Pearson maintained that the government’s motive for “judge-shopping” was a series of rulings that Judge Belot had made in two similar murder cases, rulings that allegedly were “in large part favorable to the government.” Id. at 3. Although Mr. Pearson maintained that some of these rulings “were not directly dictated by higher court precedent,” id., he also acknowledged that “[s]ome are, it is true, dictated by precedent.” Id. at 5. Judge Belot held a hearing on the motion. Mr.

Pearson introduced testimony from Bonnie Stinson, an employee in the clerk’s office in the Wichita branch of United States District Court for the District of Kansas. The purpose of her testimony was to show that the assignment system was susceptible to manipulation. Ms. Stinson testified that, as a general rule, cases involving superseding indictments were not randomly reassigned to a new judge. Instead, the United States Attorney’s office labeled the superseding indictment with the same case number that had been previously assigned to. the case involving the defendant listed first in the superseding indictment.

The new case would then be assigned to the judge handling the prior case. However, in this instance, the government did not include a case number on the superseding indictment. Ms. Stinson testified that, when the clerk’s office telephoned the U.S. Attorney’s office to inquire why the number had been left blank, the U.S. Attorney’s office responded “that it would be up to . [the clerk’s office] to decide what case number would be assigned, that [the U.S. Attorney’s office] did not want to make that decision.” Id. vol. XI, at 6.

On the superseding indictment, the court clerk’s office then filled in the number originally assigned to the prior case involving only Dominic Pearson (who was the first defendant listed in the consolidated superseding indictment). The selection of that case number had the effect of placing all defendants before Judge Be-lot. In response to Mr. Pearson’s motion for random reassignment, the government denied the allegation that it purposefully sought to have the case assigned to Judge Belot. It stressed that it had left the case number on the superseding indictment blank “so that the U.S.

District Court Clerk’s Office would make the assignment in the manner it or the Court deemed appropriate,” id. vol. II, doc.-129, at 2, but it offered no explanation as to why it changed the order in which it listed the defendants’ names. The government did, however, state that “if this Court wishes to make another judicial assignment herein, the United States would not object.” Id. at 3. After hearing Ms. Stinson’s testimony about the case assignment procedures in the District of Kansas, Judge Belot denied the motion for random reassignment from the bench, stating that it was “much ado about nothing.” Id. vol.

XI, doc. 263, at 14. He reasoned that if the government “had wanted it assigned to me they could just have put [the number for Dominic Pearson’s original case] on the indictment and they didn’t do it.” Id. at 15. He explained his view of the record as follows: The Court:.... I’ve been around here off and on for 25 years and I’ll guarantee you that I know the judges in this district better than Mr. Henry [one of Mr. Pearson’s attorneys] does. And I also know that there aren’t any of them that are more likely to favor the defense or favor the government.

And I do not like the implication on the record that I somehow favor the government. Mr. Gradert: Well, Your Honor, we’re not trying to make that implication. In fact, in Mr. Henry’s defense, Mr. Henry did not prepare this motion. And this motion was prepared by someone other than Mr. Henry and myself; however, we all discussed. The Court: Well, who was it prepared by? Mr. Gradert: It was prepared by Mr. Dedmon, Your Honor. The Court: That’s another one that I would think — Mr. Dedmon has never appeared in my court. Knows absolutely nothing about me.

Yet the implication of this is that I will not be fair to a defendant. That’s the entire implication of this, that I favor the government. Mr. Gradert: Your Honor, it’s not so much that you will favor one or the other; but it’s the Government’s perception frequently that you might be that way and that was their purpose for— The Court: There is no evidence on the record in this case. You had an opportunity. You could have called the United States Attorney. You could have called any of the assistants to establish that if you could. Mr.

Gradert: Your Honor, I chose not to call them because their response indicated that they, that they selected to do this in the manner they did; and frankly, I’m not too sure that they would, that they would give me a response that I would want to hear with regard to— The Court: Well, that really isn’t the issue, is it, Mr. Gradert? If their response is truthful it really doesn’t make any difference whether it’s a response that you want to hear, is it? Id. at 16-17. II. DISCUSSION A. Alleged Manipulation of the Judicial Assignment System On appeal, Mr.

Pearson first argues that the government violated the Due Process Clause of the Fifth Amendment by improperly manipulating the judicial assignment system so that his case was assigned to Judge Belot. Mr. Pearson’s argument raises not only constitutional issues, but also significant questions regarding the fair administration of justice. Accordingly, we will consider his challenge not only under the Due Process Clause but also, pursuant to our supervisory authority over the district courts under 28 U.S.C. § 2106. That

section codifies this court’s power to order such relief as is “just under the circumstances.” Id. 1. Due Process Challenge In his motion for random reassignment, Mr. Pearson argued that there was a significant difference between the judge sought by the government — Judge Belot— and the other judges who could have been assigned to the case. He noted that Judge Belot had recently issued rulings in two capital cases that were “in large part favorable to the government and were not directly dictated by higher court precedent.” Rec. vol. I, doc. 77 at 3; see id. at 3-5 (discussing United States v.

Chanthadara, 928 F.Supp. 1055 (D.Kan.1996), and United States v. Nguyen, 928 F.Supp. 1525 (D.Kan.1996), aff'd, 155 F.3d 1219 (10th Cir.1998)). On appeal, Mr. Pearson does not invoke these decisions, and he does not suggest that Judge Belot ruled any differently on issues in this case than the other judges in the district would have had they been assigned to it. Instead, he argues that, even absent a claim that he suffered actual prejudice from the assignment of the case to the judge preferred by the government, the government violated his due process rights by manipulating the case assignment system.

He thus characterizes the government’s alleged judge-shopping as a structural error, a fundamental “defect[ ] in the constitution of the trial mechanism, which def[ies] analysis by ‘harmless-error’ standards.” See Aplt’s Reply Br. at 2 (quoting Arizona v. Fulminante, 499 U.S. 279, 309, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991)). a. Due Process and Judicial Assignment The Supreme Court has recognized that “judges are not fungible.” Laird v. Ta tum, 409 U.S. 824, 834, 93 S.Ct. 7, 34 L.Ed.2d 50 (1972) (Rehnquist, J.) (quoting Chandler v.

Judicial Council of the Tenth Circuit of the United States, 398 U.S. 74, 137, 90 S.Ct. 1648, 26 L.Ed.2d 100 (1970) (Douglas, J., dissenting)): [T]hey cover the constitutional spectrum; and a particular judge’s emphasis may make a world of difference when it comes to rulings on evidence, the temper of the courtroom, the tolerance for a proffered defense, and the like.

Lawyers recognize this when they talk about “shopping” for a judge; Senators recognize this, when they are asked to give their “advice and consent” to judicial appointments; laymen recognize this when they appraise the quality and image of the judiciary in their own community. Id. at 834-35, 93 S.Ct. 7. More recently, the Seventh Circuit has offered a similar observation: [T]he exercise of discretion is shaped by a judge’s values and intuitions, which in turn are shaped by the judge’s background and experiences.

Among a group of six American judges, even of the same court in the same county, there is likely to be considerable, and relevant, diversity in background and experience. Former prosecutors may have a different bent from former defense lawyers, former lawyers for tort plaintiffs a different bent from former lawyers for insurance companies. Tyson v. Trigg, 50 F.3d 436, 439 (7th Cir.1995) (hereafter Tyson II). In spite of these significant differences between judges, there is scant authority discussing the requirements (if any) imposed by the Due Process Clause on the judicial assignment phase of adjudication.

Congress has granted broad discretion to the federal district courts in the assignment of cases to particular judges. See 28 U.S.C § 137 (“The business of a court having more than one judge shall be divided among the judges as provided by the rules and orders of the court.”). In light of this discretion, a number of courts have concluded that “a defendant does not have a right to have his case heard by a particular judge,” see Sinito v. United States, 750 F.2d 512, 515 (6th Cir.1984), that “a defendant has no right to any particular procedure for the selection of the judge,” Cruz v.

Abbate, 812 F.2d 571, 574 (9th Cir.1987), and that he or she.does not enjoy “the right to have [the] judge selected by a random draw.” Sinito, 750 F.2d at 515; see also Board of School Directors of City of Milwaukee v. Wisconsin, 102 F.R.D. 596, 598 (E.D.Wis.1984) (“Even a criminal defendant has no due process rights in the assignment of his case.”); United States v. Keane, 375 F.Supp. 1201, 1204 (N.D.Ill.1974) (concluding that “a defendant has no vested right to have his case tried before any particular judge, nor does he have the right to determine the manner in which his case is assigned to a judge”).

This circuit has considered the judicial assignment phase of adjudication on only a few occasions. Recently, in United States v. Diaz, 189 F.3d 1239, 1243-45 (10th Cir.1999), we concluded that 28 U.S.C. § 137 vests the district courts with broad discretion in the assignment of cases to particular judges. We rejected a defendant’s due process challenge to a rotating assignment system in which different judges were assigned to various phases of the same case.

See Diaz, 189 F.3d at 1243 (stating that the defendant’s argument was “undermined by 28 U.S.C. § 137, which vests the district court with broad discretion in assigning court business to individual judges”). Previously, in Martinez v. Winner, 771 F.2d 424, 434 (10th Cir.1985), vacated as moot, 800 F.2d 230 (10th Cir.1986), we concluded that a district judge accused of violating the Due Process Clause by improperly assigning a case to himself was entitled to absolute immunity.

We observed, “Although it is an ‘administrative’ act, in the sense that it does not concern the decision who shall win a case, the assignment of cases is still a judicial function in the sense that it directly concerns the case-deciding process.” Id. Although these decisions both concern the district court’s assignment of cases to particular judges, neither Diaz nor Martinez addresses the due process limitations, if any, on prosecutorial involvement in the assignment process. Accordingly, we turn to a sister circuit for helpful analysis.

In a case involving an allegation of improper manipulation of the case assignment system by a judge rather than a prosecutor, the Ninth Circuit concluded: While a defendant has no right to any particular procedure for the selection of the judge — that being a matter of judicial administration committed to the sound discretion of the court — -he is entitled to have that decision made in a manner free from bias or the desire to influence the outcome of the proceedings. Cruz, 812 F.2d at 574; see also Schweiker v.

McClure, 456 U.S. 188, 195, 102 S.Ct. 1665, 72 L.Ed.2d 1 (1982) (noting that the Supreme Court “repeatedly has recognized [that] due process demands impartiality on the part of those who function in judicial or quasi-judicial capacities”). In our view, if the assignment of a case to an individual judge should not be based on “the desire to influence the outcome of the proceedings,” then allowing a prosecutor to perform that task raises substantial due process concerns.

In particular, although the Due Process Clause imposes strict neutrality requirements on officials performing judicial or quasi-judicial functions, those requirements “are not applicable to those acting in a prosecutorial or plaintiff-like capacity.” Marshall v. Jerrico, Inc., 446 U.S. 238, 248, 100 S.Ct. 1610, 64 L.Ed.2d 182 (1980). “In an adversary system, [prosecutors] are necessarily permitted to be zealous in their enforcement of the law.” Id.

When prosecutorial rather than judicial functions are involved, “the constitutional interests in accurate finding of facts and application of law, and in preserving a fair and open process for decision, are not to the same degree implicated.” Id. In light of the role that prosecutors play as advocates, two state courts have concluded that judicial assignment systems allowing prosecutors to select the judge assigned to a particular case violate due process. In State v.

Simpson, 551 So.2d 1303 (La.1989) (per curiam), the defendant filed an application for a supervisory writ seeking reassignment of his case to another judge. Noting that the prosecutor and the defense attorney had stipulated that in the Louisiana district at issue, the prosecution was allowed to select the judge who presided over criminal cases, the Louisiana Supreme Court granted the writ.

The court reasoned: To meet due process requirements, capital and other felony cases must be allotted for trial to the various divisions of the court, or to judges assigned criminal court duty, on a random or rotating basis or under some other procedure adopted by the court which does not vest the district attorney with power to choose the judge to whom a particular case is assigned. 551 So.2d at 1304. The Simpson court based this conclusion on decisions holding that “[d]ue process of law requires fundamental fairness, i.e., a fair trial in a fair tribunal.” Id. (citing Turner v.

Louisiana, 379 U.S. 466, 85 S.Ct. 546, 13 L.Ed.2d 424 (1965); State v. Mejia, 250 La. 518, 197 So.2d 73 (La.1967)). The court noted decisions from other jurisdictions concluding that “courts may utilize different methods of assigning criminal cases to judges,” but observed that these decisions “do not stand for the proposition that the prosecutor may assign cases to the judge of his choice.” Id. at 1304 n. 3. In an earlier decision, a New York state court took a similar approach. In McDonald v.

Goldstein, 191 Mise. 863, 83 N.Y.S.2d 620 (N.Y.Sup.Ct.1948), the court rejected a district attorney’s challenge to an order divesting his office of its long-accepted authority to select judges for criminal cases. See. id. at 622 (noting that “[t]he District Attorney for some time past has selected the judge in each case by moving indictments for trial directly to the several parts of the court”). The court based its ruling on general principles of judicial independence, noting that judges should be free from outside control, especially by any of the litigants.

See id. at 625 (“It is the people’s prerogative, not the District Attorney’s to say who will preside over the County Court of Kings County.”). In contrast to Simpson and McDonald, most federal courts that have addressed the issue of prosecutorial involvement in judicial assignments have not found due process violations. In Tyson v.

Trigg, 50 F.3d 436, 439-42 (7th Cir.1995) (“Tyson II”), the most recent and thorough of these federal decisions, the Seventh Circuit rejected an argument raised in a habe-as corpus proceeding that the case assignment system in an Indiana state court violated the defendant’s due process rights. The system in question allowed the prosecutor to select one of six grand juries to which a proposed indictment would be presented. Each grand jury was assigned to a specific judge, and thus, by selecting the grand jury, prosecutors could choose the judge to which the case would be assigned.

The habeas petitioner in Tyson II did not argue that the assigned judge was prejudiced against him. Instead, he asserted “to allow the prosecutor to pick the judge so greatly stacks the deck against the defendant as to make the trial unfair — so unfair as to deny due process of law.” Id. at 439. The Seventh Circuit rejected that argument. First, it noted a lack of precedent holding that prosecutorial steering could constitute a due process violation warranting the reversal of a conviction.

Additionally, it concluded that the fact that the prosecutor might gain a certain advantage over the defendant in being allowed to select the judge did not render the trial fundamentally unfair. See id. at 440-41. It reasoned that the American system of criminal procedure is not balanced equally between the prosecution and the defense at every stage, but rather represents “an aggregate of imbalances.” Id. at 440. Thus, prosecutors have certain advantages in the investigative stage and in impeaching witnesses, while the rules on burdens of proof favor defendants. See id.

Absent any allegation that the judge selected by the prosecutor was actually biased against the defendant, the imbalance caused by the Indiana system was not so egregious as to affect the fairness of the trial. Several other federal courts have held that, in order to establish a due process violation for prosecutorial judge-shopping, a defendant must demonstrate that he has been actually prejudiced by the assignment of a particular judge to his case. For example, in United States v.

Gallo, 763 F.2d 1504, 1532 (6th Cir.1985), the Sixth Circuit rejected the defendant’s argument that he was entitled to a new trial because the prosecutors had engaged in a pattern of steering significant criminal cases to the judges of their choice. See id. at 1532. The court relied on its earlier decision in Sinito v. United States, 750 F.2d 512 (6th Cir.1984), in which it had held that due process concerns were not implicated by a clerical error resulting in the assignment of a case to a particular judge. See Gallo, 763 F.2d at 1532.

The Sinito panel had concluded that “ ‘a defendant does, not have the right to have his case heard by a particular judge,’ ” does not “ ‘have a right to have his judge selected by a random draw,’ ” and “ ‘is not denied due process as a result of the error unless he can point to some resulting prejudice.’” Gallo, 763 F.2d at 1532 (quoting Sinito, 750 F.2d at 515). The Gallo panel found this reasoning dispositive, rejecting the defendant’s argument because he had not alleged that he was prejudiced by the prosecutor’s alleged steering of cases. 763 F.2d at 1532.

Several other decisions have similarly required a showing of prejudice. See, e.g., United States v. Erwin, 155 F.3d at 815, 825 (6th Cir.1998); United States v. Osum, 943 F.2d 1394, 1401 (5th Cir.1991). Although all of these decisions offer helpful and relevant analysis, they differ from the instant case in several important respects. The Seventh Circuit’s Tyson II decision rejects a claim of prosecutorial steering, but some of its discussion applies only to habeas corpus proceedings and not to direct appeals.

See Tyson II, 50 F.3d at 439-40 (noting the absence of precedent on the issue and observing that new rules of constitutional law may not be applied in habeas proceedings). Moreover, Tyson II does not address the situation in which a prosecutor succeeds in having a case assigned to a particular judge due to some perceived advantage that judge will afford the government. See id. at 441-42. Cases like Gallo and Erwin, which require a showing of prejudice in order to establish a judge-shopping claim, do not discuss how such a showing may be made.

Although these cases suggest that prosecu-torial judge-shopping may violate the Due Process Clause, they do not contain substantial analysis in support of that proposition, and they do not set forth a standard for determining what kinds of prosecutorial judge-shopping are constitutionally prohibited. These cases also do not address the type of conduct at issue here: the alleged manipulation of a case assignment system that appears to afford the prosecutor discretion in selecting the judge in certain instances.

Finally, the state court cases — Simpson and McDonald — although containing sweeping language about the impropriety of allowing prosecutors to select judges, address the judge-shopping issue before the defendant was convicted. Thus, they too do not address the situation that confronts us — the alleged manipulation of the case assignment system in an individual case and the contention that a conviction should be overturned because of that manipulation.

In the absence of guidance from the Supreme Court on the due process limitations, if any, on prosecutorial steering, we are therefore presented with an issue of first impression in this circuit. Upon review of the record, we conclude that the resolution of the constitutional question presented by Mr. Pearson’s allegation of prosecutorial steering is not necessary to the disposition of this appeal. See United States v.

Cusumano, 83 F.3d 1247, 1250 (10th Cir.1996) (en banc) (“The Supreme Court has long endorsed, if not always adhered to, the notion that federal courts should address constitutional questions only when necessary to a resolution of the case or controversy before it. This is a ‘fundamental rule of judicial restraint.’ ”) (quoting Three Affiliated Tribes v. Wold Engineering, 467 U.S. 138, 157, 104 S.Ct. 2267, 81 L.Ed.2d 113 (1984)). Instead, we will assume, without deciding, that the Due Process Clause of the Fifth Amendment entitles Mr. Pearson to an impartial method of judicial assignment.

We will further assume, without deciding, that the prosecution deprived Mr. Pearson of that right by manipulating the system so that his case was assigned to Judge Belot. Affording Mr. Pearson the benefit of those two favorable assumptions, we nevertheless conclude that the assumed due process violation does not warrant reversal of his convictions. b. The Question of Structural Error Mr.

Pearson characterizes the prosecution’s alleged manipulation of the case assignment system as a structural error — a “defect[] in the constitution of the trial mechanism, which def[ies] analysis by harmless error standards.” Arizona v. Fulminante, 499 U.S. 279, 309, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991). Such errors affect “[t]he entire conduct of the trial from beginning to end” and deprive the defendant of “basic protections,” without which “ ‘a criminal trial cannot reliably serve its function as a vehicle for determination of guilt or innocence.’ ” Id. at 310, 111 S.Ct. 1246 (quoting Rose v.

Clark, 478 U.S. 670, 577-78, 106 S.Ct. 3101, 92 L.Ed.2d 460 (1986)). “Errors of this type are so intrinsically harmful as to require automatic reversal ... without regard to their effect on the outcome.” Neder v. United States, 527 U.S. 1, -, 119 S.Ct. 1827, 1833, 144 L.Ed.2d 35 (1999). If a criminal proceeding includes such an error, the resulting punishment “may [not] be regarded as fundamentally fair.” Rose, 478 U.S. at 577-78, 106 S.Ct. 3101. Supreme Court decisions have found structural error “only in a ‘very limited class of cases,’ ” Neder, 527 U.S. at -, 119 S.Ct. at 1833. (quoting Johnson v.

United States, 520 U.S. 461, 468, 117 S.Ct. 1544, 137 L.Ed.2d 718 (1997)), including thosé involving: (1) the total deprivation of the right to counsel at trial, see Gideon v. Wainwright, 372 U.S. 335, 344-45, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963); (2) a biased presiding judge, see Tumey v. Ohio, 273 U.S. 510, 523, 47 S.Ct. 437, 71 L.Ed. 749 (1927); (3) the systematic exclusion of members of the defendant’s own race from a grand jury, see Vasquez v. Hillery, 474 U.S. 254, 262-63, 106 S.Ct. 617, 88 L.Ed.2d 598 (1986); (4) the denial of the right to self-representation at trial, see McKaskle v.

Wiggins, 465 U.S. 168, 174, 104 S.Ct. 944, 79 L.Ed.2d 122 (1984); (5) the denial of the right to a public trial, see Waller v. Georgia, 467 U.S. 39, 49-50, 104 S.Ct. 2210, 81 L.Ed.2d 31 (1984); (6) the denial of the right to have a district judge (rather than a magistrate judge) preside over jury selection, see Gomez v. United States, 490 U.S. 858, 876, 109 S.Ct. 2237, 104 L.Ed.2d 923 (1989); and (7) a defective reasonable doubt instruction, see Sullivan v. Louisiana, 508 U.S. 275, 281, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993).

In Rose, the Supreme Court described these kinds of errors as the exception rather than the rule. “[I]f the defendant had counsel and was tried by an impartial adjudicator, there is a strong presumption that any other errors that may have occurred are subject to harmless-error analysis.” Rose, 478 U.S. at 579, 106 S.Ct. 3101.

In a helpful review of the concept of structural error, the Second Circuit has noted, “We do not understand [the Supreme Court’s] list of examples of violations that have been held exempt from harmless error review to mean that any violation of the same constitutional right is a ‘structural defect,’ regardless whether the error is significant or trivial.” Yarborough v. Keane, 101 F.3d 894, 897 (2d Cir.1996).

Conversely, the fact that the Supreme Court has applied harmless error analysis to one level of violation of a particular right does not necessarily mean that an egregious violation of that same right may never constitute structural error. See id. Thus, the determination of whether an error is structural depends on not only the right violated, but also the “nature, context, and significance of the violation.” Id. For example, the total deprivation of the right to counsel constitutes structural error, while the denial of the right to counsel at a preliminary hearing is subject to harmless error review.

See id. (contrasting Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963) with Coleman v. Alabama, 399 U.S. 1, 90 S.Ct. 1999, 26 L.Ed.2d 387 (1970)). Similarly, although the unjustified exclusion of a defendant from the entire trial would constitute structural error, a defendant’s absence when the judge engaged in two conversations with a juror has been subjected to harmless error analysis. See id. (citing Rushen v. Spain, 464 U.S. 114, 117 n. 2, 120-21, 104 S.Ct. 453, 78 L.Ed.2d 267 (1983)).

In our view, the due process violation alleged here — one that resulted from allowing the prosecutor to select the judge — should be placed on an analogous continuum.

Although the District of Kansas’s case assignment system arguably allowed the prosecutor to perform a kind of quasi-judicial function, there are important distinctions between allowing a prosecutor perform judicial functions after the judge has been selected and the case proceeds to trial and final decision (i.e., functions like ruling on motions and objections, issuing findings of fact and conclusions of law, and sentencing defendants) and allowing the prosecutor to perform the quasi-judicial function of selecting the judge.

Although a prosecutor’s performance of such post-selection judicial functions necessarily deprives the defendant of an impartial adjudicator, we conclude for several reasons that prosecutorial involvement in the selection of the judge does not necessarily result in the same degree of deprivation. First, a prosecutor’s choice of judges is limited.

Even if a case assignment system allows the prosecutor to select the judge, the prosecutor must still choose from a group who have undergone the process of selection and appointment, who have sworn to uphold the law and defend the Constitution, and whose conduct can be scrutinized through appellate review. There is “a presumption of honesty and integrity in those serving as adjudicators,” Withrow v. Larkin, 421 U.S. 85, 47, 95 S.Ct. 1456, 43 L.Ed.2d 712 (1975), and, as a result, we cannot presume that a federal judge selected by the prosecutor will be his agent or henchman.

Additionally, we note that a prosecutor may want the case assigned to a particular judge for a variety of reasons, some of which may not involve any disadvantage to the defendant at all: a prosecutor may simply make a random selection or he or she may seek out the most intelligent or the most experienced judge, or the one most familiar with a particular area of law. Moreover, a defendant who must proceed to trial before a judge selected by the prosecutor is not without remedies. If the judge appears biased, a defendant may file a motion for recusal. See 28 U.S.C. § 455; Nichols v.

Alley, 71 F.3d at 347, 351 (10th Cir.1995). If the judge denies that request, the defendant may challenge that decision prior to trial by filing a petition for a writ of mandamus or prohibition with this court. See Nichols, 71 F.3d at 350. In certain instances, by invoking this court’s supervisory powers, a defendant may also, prior to trial, challenge the case assignment procedure itself. See Utah-Idaho Sugar Co. v. Ritter, 461 F.2d 1100, 1104 (10th Cir.1972) (granting a petition for a writ of prohibition and mandamus barring a judge from assigning a case to himself and requiring reassignment).

Additionally, a defendant may contest the government’s prosecution of the case by filing pretrial motions, making objections at trial, and introducing evidence; he may also challenge the trial judge’s rulings on appeal. For all these reasons, a defendant in a case in which the prosecutor has selected the judge does not necessarily receive a trial that “cannot reliably serve its function as a vehicle for determination of guilt or innocence” and consequently renders any resulting punishment fundamentally unfair. Rose, 478 U.S. at 577-78, 106 S.Ct. 3101.

Finally, we have unearthed no decision finding structural error in analogous circumstances. As stated above, the Seventh Circuit has expressly rejected the argument that allowing a prosecutor to select the judge constitutes structural error. See Tyson II, 50 F.3d at 442 (characterizing structural error as involving a “denial of the most fundamental constituents of due process”). The Louisiana courts, although following Simpson’s holding that the prosecutor’s selection of the judge violates due process, have nevertheless applied harmless error analysis. See Jonathan L.

En-tin, The Sign of “the Four”: Judicial Assignment and the Rule of Law, 68 Miss. L.J. 369 (1998) (“Even in post-Simpson Louisiana, the courts have rejected numerous claims on harmless error grounds because the aggrieved party could not show how the defective assignment prejudiced the case.”); State v. Huls, 676 So.2d 160, 167-68 (La.Ct.App.1996) (applying harmless error analysis and affirming conviction even though case assignment system violated due process); State v. Romero, 552 So.2d 45, 49 (La.Ct.App.1989) (same). Accordingly, even assuming that the Due Process Clause entitles Mr.

Pearson to a neutral method of selecting a judge and that the prosecution deprived Mr. Pearson of that right by manipulating the judicial assignment system here, this assumed error is not structural. We therefore turn to the question of harmless error. c. Harmless Error Typically, when an error occurs at trial, w;e inquire on direct appeal whether the error ‘“substantially influenced’ the outcome of the trial, or whether we are left in ‘grave doubt’ as to whether it had such an effect.” United States v. Snow, 82 F.3d 935, 940 (10th Cir.1996) (quoting United States v. Tome, 61 F.3d 1446, 1455 (10th Cir.1995)).

When constitutional error is involved, “we must be persuaded that the error was harmless beyond a reasonable doubt.” Id. (citing Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)). In a variety of circumstances, courts have applied this kind of harmless error analysis to violations of due process. See Rushen v.

Spain, 464 U.S. 114, 118-19, 104 S.Ct. 453, 78 L.Ed.2d 267 (1983) (applying harmless error analysis to a due process violation involving ex parte contacts with jurors and noting that “ ‘[cjases involving [such constitutional] deprivations are [therefore] subject to the general rule that remedies should be tailored to the injury suffered ... and should not unnecessarily infringe on competing interests’ ” (quoting United States v. Morrison, 449 U.S. 361, 364, 101 S.Ct. 665, 66 L.Ed.2d 564 (1981) (alterations in original))).

Here, upon a thorough review of the record, we conclude that the government has established beyond a reasonable doubt that the alleged due process violation arising out of the judicial assignment was harmless. As we have noted, Mr. Pearson has not alleged that he was actually prejudiced by the assignment. Moreover, our independent review has uncovered no evidence that Judge Belot decided any substantive issues in a manner more favorable to the government than the other judges in the district would have decided those issues. Finally, as we conclude below, Mr.

Pearson has not made a sufficient showing of bias on the part of Judge Belot under the applicable statutes. In

summary, Mr. Pearson’s allegations of an improper manipulation of the case assignment system raise substantial due process concerns. However, even if we accept Mr. Pearson’s contentions as to the prosecution’s motivation in reordering the defendants’ names on the superseding indictment so that the case would be assigned to Judge Belot, the assumed due process violation arising out of that conduct is not structural error and is harmless beyond a reasonable doubt. As a result, Mr. Pearson’s due process challenge does not warrant a new trial. 2.

Review of the case assignment system pursuant to the court’s supervisory powers The allegations that Mr. Pearson directs at the government raise concerns beyond the requirements of the Due Process Clause. In addition to the due process problems we have discussed, our review of prior decisions and scholarly commentary reveals four related problems with the practice of allowing prosecutors to steer cases to particular judges.

First, the practice arguably affords the government an unfair advantage in litigating the case. “[I]f a litigant can choose which of [a group of] judges shall preside at the trial, that party may be able to obtain a subtle advantage over the other by selecting a judge more likely to resolve close questions in that party’s favor, even if the trial is to be a jury trial so that the judge will not make the ultimate decision.” Tyson II, 50 F.3d at 439. Second, prosecutorial maneuvering in an attempt to steer a case to a particular judge may involve an abuse of the judicial system.

Thus, prosecutors may file successive cases in multiple districts in search of a judge whom they think will treat their arguments more favorably. See Erwin, 155 F.3d at 825 (finding the filing of successive cases “strongly suggestive of judge shopping”).

In such an instance, when the choice of forum is made not for its connection with the dispute or for convenience to the parties and witnesses but rather as a means of obtaining a particular judge, a prosecutor’s conduct may violate ethical rules prohibiting the filing of actions that “ ‘delay[,] ... harass or maliciously injure another.’ ” Note, Forum Shopping Reconsidered, 103 Harvard L.Rev. 1677, 1690 (1990) (quoting Model Code of Professional Responsibility, DR 7-102(A)(l) (1981)) (alteration in original); see generally id.

Third, a system that allows prosecutorial judge-shopping arguably lacks “the appearance of impartiality that is required to obtain the confidence of the public and the accused in the system.” See Tyson v. State, 619 N.E.2d 276, 300 (Ind.App.1993) (“Tyson I ”); see also Tyson II, 50 F.3d at 441 (concluding that “[t]he practice [in the Indiana state courts] of allowing the prosecutor to choose the grand jury and hence the trial judge is certainly unsightly ...; it does lack the appearance of impartiality”).

If a judge is selected by a prosecutor rather than by a neutral procedure, then one might reasonably question the decisions made by the selected judge. Although the judge’s decisions might well be justified by the facts and the applicable law, the suspicion arises that the real reasons for the decisions may lie in some unspoken understanding or shared values that led the prosecutor to select the particular judge to handle the case.

Prosecuto-rial steering thus “exposes the tension between the ideal of the rule of law and the reality of a system created and administered by human beings.” Forum Shopping Reconsidered, supra, at 1686. Fourth, the practice, if undertaken on a broad scale, arguably threatens the independence of the judiciary.

If a judge receives case assignments not through some neutral system, but rather because of prosecutors’ opinion that he or she is more favorably disposed to the government’s arguments than another judge in the same district, then a judge’s caseload might be based in part on prosecutors’ evaluations of judicial performance. Under that scenario; judges meeting with the prosecutors’ approval might have future cases assigned to them whereas judges whom prosecutors dislike might not receive future assignments.

In rendering judges’ workloads dependent on advocates’ assessment of their decisions, a widespread tolerance of prosecutorial steering might tempt judges to base their decisions in a given case on the effect of those decisions on their future assignments. See McDonald, 83 N.Y.S.2d at 626 (criticizing a county system allowing the district attorney’s office to select judges for particular cases and concluding “[t]hat a judge should ever be burdened with the thought that his assignments depended on the district attorney’s appraisal of his court work is unthinkable in American jurisprudence”).

In light of these serious concerns about the practice of prosecutorial steering, we believe that Mr. Pearson’s allegations warrant the exercise of our supervisory power over the district court. See 28 U.S.C. § 2106 (providing that the court may order such relief as “may be just under the circumstances”). We therefore consider whether, aside from the due process claim raised by Mr. Pearson, his allegations of an improper manipulation of the case assignment process warrant reversal of his conviction.

In conducting this analysis, we will apply the standard that we have adopted for reviewing certain violations of the recusal statute, 28 U.S.C. § 455, when there is no indication that the tribunal is actually biased. As in this case, these recusal cases view the promotion of “public confidence in .the integrity of the judicial process” as an important policy. Liljeberg v. Health Servs. Acquisition Corp., 486 U.S. 847, 858 n. 7, 108 S.Ct. 2194, 100 L.Ed.2d 855 (1988). As with prosecutorial steering, the elimination, if possible, of even the appearance of impropriety is desirable. See id. at 860, 108 S.Ct. 2194.

In Harris v. Champion, 15 F.3d 1538 (10th Cir.1994), we concluded that a judge should have recused himself from a case pursuant to § 455 because his participation in the case created an appearance of impropriety and because he was related to a party.

We then considered the following factors in determining whether the error was harmless: (1) “the risk of injustice to the parties in the particular case”; (2) “the risk that the denial of relief will produce injustice in other cases”; and (3) “the risk of undermining the public’s confidence in the judicial process.” Harris, 15 F.3d at 1571-72 (quoting Liljeberg, 486 U.S. at 864, 108 S.Ct. 2194). Because that approach addresses both the individual and systemic interests at issue in this instance of alleged prosecutorial steering, we apply it to Mr. Pearson’s claim.

Having already concluded that the alleged selection of Judge Belot did not affect this result in this case, we proceed to the second factor — the risk of injustice in other cases. Here, it is significant that the District of Kansas’s practice operates to allow prosecutors to select judges in only a narrow set of circumstances: when a superseding indictment is filed after a series of related cases have been assigned to different judges. There is no indication that the alleged manipulation of the case assignment system will recur in a large number of cases. Cf.

Tyson II, 50 F.3d at 438 (observing that a law of general application that “provided that the U.S. Attorney in each district shall designate the federal judge to preside in criminal cases ... would raise profound issues under the due process clause”). Here, Ms. Stinson (an employee of the clerk’s office in the District of Kansas) testified that, in most instances, the district follows a random assignment system. In our view, such an assignment system will usually protect against the most egregious forms of prosecutorial steering.

Moreover, the specific practice at issue here appears to be remediable in other cases. In particular, the record suggests no reason why the District of Kansas could not employ a case assignment system that does not allow the prosecutor to select the judge in instances in which the government files several separate but related cases and then obtains a consolidated superseding indictment. We strongly urge the District of Kansas to adopt such a system, and we note that, even under the current system, defendants in other cases may still seek reassignment or recusal if the facts warrant such relief.

Thus, this factor too, suggests that the error in the assignment of the case was harmless. The third factor — the risk of undermining the public’s confidence in the judicial process — presents a closer question. Unfortunately, the cases applying this harmless error test do not set forth a clear standard for determining when a particular violation undermines the public’s confidence in the judicial process. Liljeberg, the Supreme Court decision from which this standard is derived, indicates that the determination must be made on a case-by-case basis. See 486 U.S. at 862-87, 108 S.Ct. 2210.

In our view, several considerations support a finding of harmlessness here. First, in contrast to other instances of prosecutorial steering, Mr. Pearson has not alleged that the government has violated particular ethical rules or criminal laws. See United States v. August, 745 F.2d 400, 401-04 (6th Cir.1984) (affirming conviction for interfering with the due administration of justice in violation of 18 U.S.C. §§ 371 and 1503 by circumventing the case assignment system in a bankruptcy court); State v.

Jurek, 52 Ohio App.3d 30, 556 N.E.2d 1191, 1198 (Ohio Ct.App.1989) (affirming conviction for bribing bond commissioners in order to steer criminal cases to particular judges). Even if we assume that the prosecution manipulated the judicial assignment system by reordering the defendants’ names on the style of the indictment, as Mr. Person suggests, we are not presented with such egregious prosecutorial conduct that reversal of his conviction is necessary in order to protect the public’s confidence in the judicial system.

Second, we must acknowledge that a certain type of judge-shopping inheres in our federalist system. See McCuin v. Texas Power & Light Co., 714 F.2d 1255, 1261 (5th Cir.1983). In the case before us, for example, we note that although robbery and murder charges outside federal or Indian land are typically filed in state courts, the provisions of the Hobbs Act, 18 U.S.C. § 1951, authorized the United States Attorney’s office to bring charges against Mr. Pearson in federal court.

The decision to proceed in federal court (rather than to allow the state prosecutors to try the case in state court) could have been made for a variety of reasons. See McCuin, 714 F.2d at 1261. However, if the prosecution here admitted that it chose the federal forum because it thought that the federal judges in Kansas were more favorable to the government than the state judges, we do not believe that such'a form of judge-shopping would undermine the public’s confidence in the integrity of the proceedings to the extent that reversal of the conviction would be warranted. Cf. United States v.

Andersen, 940 F.2d 593, 596 (10th Cir.1991) (holding that a defendant’s due process rights are not violated by the federal government’s decision to prosecute under a federal, rather than state, statute, notwithstanding the harsher penalties).

This limited tolerance for certain kinds of judge-shopping (in contrast, for example, to our judicial system’s complete intolerance of biased judges) echoes Judge Posner’s observation in Tyson II that our system of criminal procedure “far from being balanced every step of the way, is an aggregate of imbalances,” 50 F.3d at 440, some of which favor the prosecution and some of which favor the defendant. Most importantly, as further discussed in later sections of this opinion, Mr.

Pearson “was tried before an impartial judge, under the correct standard of proof and with the assistance of counsel.” Neder, 527 U.S. at -, 119 S.Ct. at 1834. “[A] fairly selected, impartial jury was instructed to consider all of the evidence and argument....” Id. Moreover, the allegedly improperly selected judge committed no reversible legal errors and was not the trier of fact. See United States v.

Jordan, 49 F.3d 152, 159 (5th Cir.1995) (affirming conviction but remanding for resentencing when the judge erred in failing to recuse herself and distinguishing between the judge’s role in ruling on legal questions at trial and her role as a factfinder at sentencing).

Although public confidence in our judicial system may be undermined to some degree by a prosecutor’s manipulation of the case assignment system, we believe that it would also be undermined by requiring a second trial of a defendant who was convicted by a properly selected and instructed jury that considered the evidence and found the defendant guilty beyond a reasonable doubt in a trial conducted by a judge who committed no reversible errors.

In this instance, we are confident that the District of Kansas can adopt an assignment system that prevents prosecutors from selecting judges on the basis of the order of the defendants’ names in superseding indictments. Our confidence in the adoption of such a system and our assessment of the fairness of the trial that Mr Pearson actually received convince us that, in the exercise of our supervisory powers, we need not vacate Mr. Pearson’s conviction. 3.

Summary In

summary, we have assumed, without deciding, that the Due Process Clause of the Fifth Amendment entitles Mr. Pearson to an impartial method of assigning his case to a particular judge. We have fur: ther assumed, without deciding, that the prosecution here deprived Mr. Pearson of that right by manipulating the system so that his case was assigned to Judge Belot. Viewing Mr. Pearson’s arguments in light of these assumptions, we conclude that these assumed errors are not structural and are harmless beyond a reasonable doubt.

We also conclude that these errors are harmless pursuant to our supervisory power over the district court. Thus, Mr. Pearson’s allegations that the prosecution manipulated the case assignment system do not warrant a new trial. B.The Jury Selection System In the district court proceedings, Mr. Pearson contended that, the jury selection system in the District of Kansas violated his constitutional right to a trial before a fair cross-section of the community by improperly excluding minorities. On appeal, he argues that he has established a prima facie violation of the fair-cross

section requirement. According to Mr. Pearson, the government should now be required to show that “attaining a fair cross-section of the community would be incompatible with a significant government interest.” Aplt’s Br. at 20. We review the factual decisions underlying the district court’s ruling for clear error and the legal determination of whether a prima facie violation of the fair cross-section requirement has occurred de novo. See United States v. Shinault, 147 F.3d 1266, 1271 (10th Cir.1998).

In Shinault, we considered and rejected a Sixth Amendment and Jury Selection and Service Act challenge to the selection system in the Wichita-Hutchinson division of the District of Kansas. See id. at 1270-73. Rather than restate Shinault’s comprehensive analysis, we simply rule that it forecloses this portion of Mr. Pearson’s appeal, which challenges the same jury selection system on the same grounds. C. Congress’s authority to enact 18 U.S.C. § 1951 Mr. Pearson moved to dismiss his indictment, claiming that Congress lacked the constitutional authority to enact § 1951.

The district court denied his motion. “We review challenges to the constitutionality of a statute de novo.” United States v. Bolton, 68 F.3d 396, 398 (10th Cir.1995). We have previously concluded that § 1951 “regulates activities that in aggregate have a substantial effect on interstate commerce.” Id. at 399. Therefore, § 1951 “represents a permissible exercise of the authority granted to Congress under the Commerce Clause.” Id. Because § 1951 is within Congress’s Commerce Clause authority, Mr. Pearson’s constitutional challenge to his convictions fails. D. Double Jeopardy Clause Mr.

Pearson contends that his convictions under 18 U.S.C. §§ 1951 and 924(c)(1) and (j)(

l) violate the Double Jeopardy Clause because they require proof of the same elements. We review Mr. Pearson’s Double Jeopardy Clause challenge de novo. See United States v. Cordoba, 71 F.3d 1543, 1545 (10th Cir.1995). In light of our holdings in United States v. Overstreet, 40 F.3d 1090, 1094-95 (10th Cir.1994), and United States v. Lanzi 933 F.2d 824, 826 (10th Cir.1991), we are not persuaded by Mr. Pearson’s argument. A person may be prosecuted for more than one crime based on the same conduct (1) if each crime requires proof of a fact that the other does not, see Blockburger v.

United States, 284 U.S. 299, 304, 52 S.Ct. 180, 76 L.Ed. 306 (1932), or (2) if Congress has clearly expressed its intent to impose cumulative punishment for the same conduct under different statutory provisions, see Garrett v. United States, 471 U.S. 773, 778, 105 S.Ct. 2407, 85 L.Ed.2d 764 (1985). In Overstreet and Lanzi, we held that, through the plain language and legislative history of § 924(c)(1), Congress clearly expressed its intent that § 924(c)(l)’s punishment be cumulative with the punishment for the underlying violent crime. See Overstreet, 40 F.3d at 1094-95 (holding that convictions under §§ 924(c)(

i) and 2119 did not violate the Double Jeopardy Clause); Lanzi, 933 F.2d at 826 (holding that convictions under §§ 924(c)(1) and 2113(

a) and (

d) did not violate the Double Jeopardy Clause). Although the statute at issue in Mr.

Pearson’s case, § 1951, concerns a different violent crime (robbery) than was at issue in Overstreet (carjacking) and Lanzi (armed bank robbery), Congress’s clear intent to provide multiple punishments to defendants who commit violent crimes while using or carrying a firearm is consistently evident in the plain language of § 924(c)(1): [Whoever], during and in relation to any crime of violence or drug trafficking crime ... uses or carries a firearm, ... shall, in addition to the punishment provided for such crime of violence or drug trafficking crime, be sentenced to imprisonment of [five years to life imprisonment, depending on the circumstances]. 18 U.S.C. § 924(c)(l)(A)(I) (emphasis added).

We are not persuaded by Mr. Pearson’s argument that United States v. Dixon, 509 U.S. 688, 113 S.Ct. 2849, 125 L.Ed.2d 556 (1993), overruled Garrett and Hunter and established Blockberger’s same elements test as the sole test for whether multiple punishments violate the Double Jeopardy Clause. Dixon simply did not consider Congressional intent to impose multiple punishments, and the few citations in Dixon to Garrett and Hunter do not cast those cases into doubt.

See Dixon, 509 U.S. at 709 n. 14, 710 n.15, 724, 735, 113 S.Ct. 2849 (White, J., concurring and dissenting in part); id. at 746, 113 S.Ct. 2849 (Souter J., concurring and dissenting in part); see also United States v. Gonzales, 40 F.3d 735, 737-38 (5th Cir.1994). As the rule of Garrett and Hunter is to examine legislative intent to see if that intent clearly favors imposing multiple punishments for the same conduct, those cases are in accord with Blockberger’s principles and have not been implicitly overruled by Dixon. We reject Mr.

Pearson’s Double Jeopardy Clause challenge to his convictions under §§ 1951 and 924(c)(1) and (j). E. Mr. Pearson’s statements to the police after his arrest Mr. Pearson unsuccessfully moved the trial court to suppress the statements he made to the FBI after his arrest. On appeal, Mr.

Pearson argues that his confession was fruit of the poisonous tree of his arrest, which he claims was unlawful because the FBI did not have probable cause and did not obtain an arrest warrant from á neutral and detached magistrate, despite the fact that the agents had time to do so. “On appeal from a denial of a motion to suppress, we view the evidence in a light most favorable to the government and accept the district court’s findings of-historical fact unless clearly erroneous.” United States v. Lewis, 71 F.3d 358, 360 (10th Cir.1995).

A court will find probable cause to arrest when facts and circumstances from a reasonably trustworthy source are within the officer’s knowledge and sufficiently warrant a person of reasonable caution to believe a crime has been or is being committed by the person to be arrested. See United States v. Morgan, 936 F.2d 1561, 1568 (10th Cir.1991). The agents had probable cause to arrest Mr. Pearson. As the trial court noted, before the FBI arrested Mr. Pearson, it heard statements from four people who independently, and without knowledge of each others’’ statements, implicated Mr. Pearson in planning the robbery. Mr.

Pearson’s attempts to belittle the significance of these statements is not convincing and not completely forthright: [The witnesses] simply [gave] statements informing the F.B.I. that during the month preceding the robbery, [Mr.] Pearson had tried to obtain the restaurant’s deposits without force, but with the help of Deborah Meyer and Gracie Ginyard, both employees of Mr. Good-cent’s.... Based, upon this information of a possible conspiracy to embezzle monies from the Mr. Goodcent’s restaurant [the FBI arrested Mr. Pearson], Aplt’s Opening Br. at 25-26 (citations omitted).

The witnesses’ statements did not “simply” concern a plan “to obtain” the restaurant’s deposits “without force” or “to embezzle monies.” According to Special Agent Daniel Jablonski, Wandra Ginyard (Gracie Ginyard’s sister) told the FBI that she had heard Mr. Pearson at a party “discuss[ing] robbing the Mr. Goodcents restaurant.” Rec. vol. V, at 5. Monique Gasper corroborated her statements when shé informed the agents that at a party Mr. Pearson “joined in or was leading the discussions about ... how easy it would be to rob Mr. Goodcents.” Id. at 8. Gracie Ginyard further corroborated Wanda’s and Ms.

Gasper’s statements when “[s]he advised that she was present during that gathering in which ... the robbery of the Mr. Goodcents store was talked about.” Id. at 9. Additionally, Gracie Ginyard told the agents that she agreed and then declined to accompany Ms. Meyer when Mr. Pearson was going to take the deposits. Finally, Ms. Meyer told the FBI that Mr. Pearson had not only asked her to allow him to take Mr. Goodcents deposits from her, with and without Gracie Ginyard’s help, but also discussed with her plans to rob Mr.

Goodcents, including telling her that he was going to rob the store on the night of the sixteenth and seventeenth. These four independent statements amounted to probable cause to arrest Mr. Pearson in connection with the Mr. Good-cents robbery, and Mr. Pearson’s statements, to the FBI should not have been suppressed on the grounds that they were the fruit of an arrest made without probable cause. Nor can Mr. Pearson’s statements be suppressed because his arrest was warrantless.

His argument that the FBI was required to obtain a magistrate-approved arrest warrant because the agents had time to do so ignores long-standing and unequivocal Supreme Court precedent: Law enforcement officers may find it wise to seek arrest warrants where practicable to do so, and their judgments about probable cause may be more readily accepted where backed by a warrant issued by a magistrate.

But we decline to transform this judicial preference into a constitutional rule when the judgment of the Nation and Congress has for so long been to authorize war-rantless public arrests on probable cause rather than to encumber criminal prosecutions with endless litigation with respect to the existence of exigent circumstances, whether it was practicable to get a warrant, whether the suspect was about to flee, and the like. United States v. Watson, 423 U.S. 411, 423-24, 96 S.Ct. 820, 46 L.Ed.2d 598 (1976) (citations omitted). Having already determined that there was probable cause to support Mr.

Pearson’s arrest, we reaffirm that the agents were not required to obtain an arrest warrant before taking Mr. Pearson into custody, even though they had time to do so. The district court did not err in refusing to suppress Mr. Pearson’s comments to the FBI. F. Sufficiency of the Evidence and Jury Instructions for § 1111(

a) Felony Murder In order to convict Mr. Pearson of a violation of § 924(j)(l), the jury had to find that, “in the course of a violation of [§ 924](c),” he murdered Ms. Montgomery in violation of 18 U.S.C. § 1111(a). See 18 U.S.C. § 924(j)(l). Mr. Pearson contends, in related arguments, that the evidence was insufficient to convict him of § 1111(a), causing his § 924(j)(

l) conviction to be invalid and that the jury was erroneously instructed as to the elements of a § 1111(

a) violation. Both of Mr. Pearson’s objections are based on his mistaken contention that § 1111(

a) requires the government to independently establish malice aforethought to commit the killing, rather than simply to prove that the killing occurred in the course of an underlying felony.

Section 1111(

a) defines murder to include felony murder. To prove “malice aforethought” in felony murder cases, the prosecution need only show commission of the specified felony. See Montoya v. United States Parole Comm’n, 908 F.2d 635, 638 (10th Cir.1990). This circuit has recently held that “once the government has shown that [defendant intended to commit the robbery and that a killing occurred in the course of that robbery, no additional proof of state of mind is necessary [to support a conviction under

section 1111(a) ].” United States v. Nguyen, 155 F.3d 1219, 1225 (10th Cir.1998). Here, the evidence presented to the jury, viewed in the light most favorable to the government, indicates that Mr. Pearson committed a robbery. Moreover, the jury was properly instructed that “[a] killing is done with malice aforethought when it is done deliberately and with the intent to kill another person, or if it results from the commission of a robbery.” Rec. vol. II, doc. 166, Instruction No. 29. Accordingly, we conclude that the evidence was sufficient for the jury to find that Mr. Pearson committed felony murder under § 1111(

a) and that the jury was properly instructed on the elements of this offense. We therefore reject these related challenges to Mr. Pearson’s § 924(j)(

l) conviction. G. Mr. Pearson’s request for lesser included offense instructions Mr. Pearson asked the district court to instruct the jury on second degree murder and manslaughter as lesser included offenses to felony murder. The district court refused. We engage in de novo review of the legal question of whether an offense for which an instruction is sought actually qualifies as a lesser included offense of the offense charged. United States v. Duran, 127 F.3d 911, 914 (10th Cir.1997), cert. denied, 523 U.S. 1061, 118 S.Ct. 1389, 140 L.Ed.2d 648 (1998).

However, we review the district court’s decision as to whether there is enough evidence to justify a lesser included offense instruction for an abuse of discretion. Id. Mr. Pearson was entitled to an instruction on these lesser included offenses if: (1) he made a proper request; (2) the lesser included offenses included some but not all of the elements of the offense charged; (3) the elements differentiating the lesser included offenses from the charged crime are in dispute; and (4) a jury could rationally convict the defendant of one of the lesser offenses and acquit him of the greater offense. United States v.

Moore, 108 F.3d 270, 272 (10th Cir.1997) (citing Fitzgerald v. United States, 719 F.2d 1069, 1071 (10th Cir.1983)). Applying this standard, we conclude that the district court properly denied Mr. Pearson’s requested lesser included offense instructions because both fail the first half of the fourth requirement: a jury could not rationally convict him of either second degree murder or manslaughter. The first paragraph of 18 U.S.C. § 1111(

a) defines first degree murder. The second paragraph describes second degree murder as “[a]ny other murder.” Replacing the word “murder” with its definition, found in the first sentence of 1111(a), results in the reconstituted statutory expression of second degree murder as “any other unlawful killing of a human being with malice aforethought.” However, we,must again look to the common law to find the definition of “malice afo

[…]