Opinion by Judge WIGGINS; Dissent by Judge PREGERSON.

Johnny Roy Footracer, a Native American, claims that he was denied his Sixth Amendment right to a jury venire reflecting a fair cross-section of the community because his criminal trial was transferred from the District of Arizona’s Prescott Division, the division in which the alleged crimes were committed, to the Phoenix *1060Division, which contains a much smaller percentage of Native Americans. We affirm Footracer’s conviction.

I.

In September 1996, a federal grand jury indicted Johnny Roy Footracer on seven counts of various sex offenses allegedly committed against minor girls on the Navajo Indian Reservation in Arizona. The Reservation lies within the Prescott Division of the District of Arizona, and Footracer’s case originally was set for trial in the Prescott Division. The district court, however, sua sponte transferred the trial to the U.S. Courthouse in Phoenix, Arizona.1 Footracer moved to have his trial returned to Prescott, where Native Americans constitute a much larger percentage of the population than they do in Phoenix, but the district court denied the motion and held the trial in Phoenix. The jury convicted Footracer of four of the seven counts in the indictment, and the district court sentenced Footracer to 46 months incarceration. Footracer now appeals his conviction, arguing that the transfer of his trial from' Prescott to Phoenix violated his Sixth Amendment right to a jury venire composed of a fair cross-section of the community.2 We have jurisdiction under 28 U.S.C. § 1291. “We review independently and non-deferentially a challenge to the composition of grand and petit juries.” Thomas v. Borg, 159 F.3d 1147, 1149 (9th Cir. 1998) (quoting United States v. Sanchez-Lopez, 879 F.2d 541, 546 (9th Cir. 1989)).

II.

Footracer relies on this court’s dicta in United States v. Etsitty, 130 F.3d 420 (9th Cir. 1997), to support the claim that transferring his case from the Prescott Division to the Phoenix Division denied him the right to a jury venire reflecting a fair cross-section of the community. In Etsitty, a Prescott district court transferred a criminal case to Phoenix for trial. The district court then denied the defendant’s motion to have the case transferred back to Prescott. A Phoenix jury convicted the defendant, and the defendant appealed to this court. He argued, inter alia, that the transfer of his case to Phoenix violated his Sixth Amendment right to a jury venire reflecting a fair cross-section of the community as articulated in Duren v. Missouri, 439 U.S. 357, 99 S.Ct. 664, 58 L.Ed.2d 579 (1979). The defendant based his claim on undisputed evidence showing that the Prescott Division had a much larger percentage of Native Americans than did the Phoenix Division. Faced with this evidence, the Etsitty court opined that the systematic removal of cases from the Prescott Division to the Phoenix Division would pose “a strong case for finding an exclusion of Indians in violation of Duren.” Id. at 425. Nonetheless, the Etsitty court refused to decide the issue, leaving “for another day” the issue of whether the systematic transfer of criminal trials from Prescott to Phoenix violates the fair cross-section requirement.3 Etsitty, 130 F.3d at 426. We believe it is time to resolve this issue.

In Duren v. Missouri, 439 U.S. 357, 99 S.Ct. 664, 58 L.Ed.2d 579 (1979), the Supreme Court announced a three-part test for determining the constitutionality of jury selection under the Sixth Amendment:

*1061[T]o establish a prima facie violation of the fair cross-section requirement, the defendant must show (1) that the group alleged to be excluded is a “distinctive” group in the community; (2) that the representation of this group in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community; and (3) that this underrepresentation is due to systematic exclusion of the group in the jury selection process.

Id. at 364, 99 S.Ct. 664.

Footracer claims that Native Americans are the relevant “distinctive” group in this case. If we accept this characterization, Footracer satisfies the first Duren prong because Native Americans undoubtedly qualify as a “distinctive” group. See United States v. Cannady, 54 F.3d 544, 547 (9th Cir. 1995) (reaffirming that ethnic minority groups qualify as distinctive groups). Footracer also argues that he satisfies Duren’s second prong because Native Americans were underrepresented on his jury venire. We need not reach this issue, however, because even if Native Americans were underrepresented, Footracer cannot satisfy Duren’s third prong because he cannot show that Native Americans were systematically excluded from jury venires.

To prevail on his claim, Footracer must show that a distinctive group was underrepresented because of the “systematic exclusion of the group.” Duren, 439 U.S. at 364, 99 S.Ct. 664 (emphasis added). Footracer must show that Native Americans, in particular, were excluded. To “exclude” means to “bar from participation, consideration, or inclusion.” Webster’s Ninth New Collegiate Dictionary 433 (1984). Therefore, Footracer must show that the jury selection process barred Native Americans from participating on juries. This use of the term “exclude” implies that the Sixth Amendment is violated only where the particular distinctive group was treated differently. The case law supports this interpretation. In Duren, the Supreme Court found constitutionally infirm a jury selection plan that failed to treat all groups equally. Under the jury selection process in that case, women were treated differently than men because all women and older men, but not men under age 65, were given the option to return a questionnaire to the jury commissioner for the purpose of avoiding jury duty. Duren, 439 U.S. at 361-62, 99 S.Ct. 664. In addition, under the jury selection process in Duren, if women failed to report for jury duty, they were presumed to have claimed an exemption from service. Id. at 362, 99 S.Ct. 664. This same presumption was not extended to men. The different treatment accorded women caused an underrepresen-tation of women on jury venires, violating the Sixth Amendment.

Footracer cannot satisfy this third Duren requirement. He makes no allegation that Native Americans were treated differently than other racial or ethnic groups. According to Footracer, all criminal trials from Prescott are held in Phoenix. He claims that the jury venires for Phoenix trials are composed solely of people residing in the Phoenix Division. Obviously, such an approach would exclude the residents of the Prescott Division, which has a much larger percentage of Native Americans than does the Phoenix Division. Footracer argues, therefore, that the jury selection process, by excluding the large Native American population in the Prescott Division, systematically excludes Native Americans. But Footracer misinterprets Duren’s systematic exclusion requirement. If the Prescott district courts transfer all of their trials to Phoenix, then every potential juror in the Prescott division — Native American and nonNative American alike — is excluded from jury service. Native Americans are not treated differently; they are excluded to the same extent as all other racial and ethnic groups in the Prescott Division.4 Thus, instead of arguing that disparate *1062treatment infected the jury selection plan in Arizona, Footraeer argues that the neutral jury selection plan violated the Sixth Amendment because it had a disparate impact on a distinctive group. As such, his argument runs counter to the overwhelming weight of authority, which establishes that a jury selection system does not systematically exclude a distinctive group where the system treats all groups equally but has a disparate impact on one or more.5

This principle is seen most clearly in the numerous cases where defendants have challenged the use of voter registration lists in compiling lists of eligible jurors. Many defendants have claimed that, because racial minorities register to vote in smaller percentages than whites, drawing jurors from voter registration lists guarantees that racial minorities will be underrepresented on jury venires. Courts have rejected this argument consistently. Some courts have assumed the fact that the use of voter registration lists leads to the un-derrepresentation of racial minorities on jury venires. These courts still have rejected defendants’ Sixth Amendment challenges to these jury selection plans on the basis of Duren’s systematic exclusion prong. For example, in Truesdale v. Moore, 142 F.3d 749 (4th Cir. 1998), the Fourth Circuit rejected such a claim because Truesdale “[did] not advance[ ] any direct evidence of ‘systematic exclusion’ of African Americans from the venire.” Id. at 755. Truesdale argued that use of voter registration lists decreased the percentage of African Americans on jury venires, but he offered no evidence that African Americans were treated differently than any other group. “Instead he [relied] on the bare assertion of substantial underrepresentation to prove that there was a structural or systemic impediment to voter registration by African Americans.” Id. The court rejected the claim, holding that it has “consistently required more to make out a violation of the ‘fair cross-section’ guarantee.” Id. To find a Sixth Amendment violation where a criminal defendant shows only that a distinctive group was underrepresented on jury venires, without evidence that the group was treated differently than other groups, would “allow [the defendant] to substitute evidence of substantial under-*1063representation for evidence of systematic exclusion” and would “exalt racial proportionality over neutrality].” Id.; see also United States v. Cecil, 836 F.2d 1431, 1446 (4th Cir. 1988) (holding “that a showing of mere statistical underrepresentation, without evidence of actual discrimination or exclusionary practices was insufficient to establish” a Sixth Amendment violation.) (emphasis added) (internal citations and quotation marks omitted).

The Fourth Circuit is not alone. In United States v. Ireland, 62 F.3d 227 (8th Cir. 1995), the Eighth Circuit also rejected a claim that the use of voter registration lists decreased minority participation on juries in violation of the Sixth Amendment. The Ireland court held that even if the defendant could provide evidence of under-representation, such “evidence, without more, fails to establish that Native Americans were systematically excluded.” Id. at 231. “In addition to a numerical disparity, a defendant must demonstrate that the voter-registration qualifications are suspect, or that the jury-selection procedure is administered in a discriminatory manner.” 6 Id. Other circuits have made similar rulings. See Schanbarger v. Macy, 77 F.3d 1424 (2d Cir. 1996) (“[Ajbsent positive evidence that some groups have been hindered in attempting to register to vote, a jury venire drawn from voter registration lists [does not] violate[ ] the Sixth Amendment’s fair cross-section requirement.”); United States v. Ashley, 54 F.3d 311, 314 (7th Cir. 1995); Ford v. Seabold, 841 F.2d 677, 685 (6th Cir. 1988).7

All of these cases establish that a defendant cannot prevail on a fair cross-section claim without evidence that a distinctive group is treated differently than other groups; disparate impact is not enough.8 Even if a group is underrepre*1064sented on jury venires, there is no constitutional violation where the jury selection process is neutral in its treatment of all groups. Here, Native Americans were treated the same as all other groups. Thus, we reject Footracer’s argument.

III.

Footraeer cannot show that Native Americans were systematically excluded from Arizona’s jury venires. But assuming as true Footracer’s alleged facts, he has shown the systematic exclusion of all adults eligible for jury service who reside in the Prescott area. If every Prescott criminal case is tried in Phoenix, and if Phoenix jury venires are drawn solely from residents in the Phoenix Division, no resident of the Prescott Division would ever sit on a criminal jury. Geography, not race, would determine the composition of Arizona’s jury venires.

Accepting Footracer’s facts, Arizona’s selection process treats differently the entire, jury-eligible population of the Prescott Division. The question, then, is whether this “group” satisfies Duren’s requirement of distinctiveness. It does not. This circuit has adopted the Eleventh Circuit’s three-part test for determining whether a group is distinctive. United States v. Fletcher, 965 F.2d 781, 782 (9th Cir. 1992) (adopting the test used in Willis v. Zant, 720 F.2d 1212 (11th Cir. 1983)).9 Under this test, a group is distinctive if: (1) the group is defined and limited by some factor, (2) a common thread or basic similarity in attitude, ideas, or experience runs through the group, and (3) there is a community of interests among members of the group such that the group’s interest cannot be adequately represented if the group is excluded from the jury selection process. Id. Using this test, we have held that “college students” are not a “distinctive” group under Duren. Id. We have reached the same conclusion regarding young adults, non-working people, and non-high school graduates. United States v. Kleifgen, 557 F.2d 1293, 1296 (9th Cir. 1977). All of these alleged groups lacked the requisite similarity in “attitudes, ideas, and experiences” to qualify as a distinctive group. Fletcher, 965 F.2d at 782. It seems clear that the population residing within the Prescott Division’s large geographical area lack sufficient shared “attitudes, ideas, and experiences” to qualify as a “distinctive” group. Thus, Footraeer does not satisfy Duren’s first prong.

This understanding of Duren is in harmony with those circuits that have analyzed whether the populations of large, geographic groups like the Prescott Division can constitute “distinctive” groups. In Zicarelli v. Dietz, 633 F.2d 312 (3rd Cir. 1980), the Third Circuit rejected a claim similar to Footracer’s. There, a defendant who was accused of committing a crime in Hudson County had his trial transferred to Burlington County. Jurors were drawn only from Burlington County. The defendant claimed that this transfer violated the Sixth Amendment because of *1065the different racial compositions of the two counties. The Third Circuit rejected this argument, holding that residents of a geographic group cannot be considered a distinctive group for cross-section analysis unless the group is “profoundly culturally distinct.” Id. at 320. Numerous courts have made similar rulings.10 See United States v. Butera, 420 F.2d 564, 572 (1st Cir. 1970) (“[W]e are not aware that residents of counties can be said to hold views and attitudes which are in any way ‘distinct’ from those of their neighbors”); Cobbs v. Robinson, 528 F.2d 1331, 1336 (2d Cir. 1975) (“Residents of the City of Bridgeport are also not necessarily a cognizable class.”); United States v. Foxworth, 599 F.2d 1, 4 (1st Cir. 1979) (“[I]t can hardly be asserted that the registered voters in a given city or town are sufficiently ‘distinct’ to constitute a cognizable group.”).11

IV.

Our interpretation of Duren is not only consistent with established case law, it also avoids undermining Federal Rule of Criminal Procedure 18 and imposing onerous burdens on district courts. A decision adopting Footracer’s argument would do both.12

Under Rule 18, Fed.R.Crim.P., a criminal trial must be held in the district in which the crime was committed, but it need not be held in the same division in which the crime was committed. See United States v. Herbert, 698 F.2d 981, 984 (9th Cir. 1983). In fact, this rule was meant to “vest[ ] discretion in the court to fix the place of trial at any place within the district.” Id. (citing Fed.R.Crim.P. 18). Adopting Etsitty’s dicta would destroy Rule 18 and thwart its policy goals. Without question, there are numerous districts with internal divisions that vary widely in their ethnic makeups. Thus, even if all jury selection procedures were perfect— and were completely free of discrimination against any group — the jury venires in many divisions still would have a higher percentage of a particular racial minority than would the venires in neighboring divisions. See United States v. Gottfried, 165 F.2d 360, 364 (2d Cir. 1948) (“There are probably no districts in the Union, which can be divided without disclosing in the sections different racial, religious, political, social or economic percentages.”). Thus, *1066anytime a district court exercises its Rule 18 right to hold trial in a division other than the division in which the alleged crime was committed, the defendant is likely to have an argument that the jury venire does not reflect the racial composition of the division in which the alleged crime was committed.

Footracer argues that these differences in racial composition could serve as the basis for a Sixth Amendment violation. To rule in Footracer’s favor, we first would have to accept that Native Americans, and other racial minorities, qualify as distinctive groups. This we already have done. Second, we would have to accept the premise that a neutral jury selection system that has a disparate impact on minority groups systematically excludes these minority groups. This premise we reject. But assuming, arguendo, that we accepted both premises, the key to resolving claims like Footracer’s would be whether the distinctive group is underrepresented as articulated in Duren’s second prong. This analysis hinges on the absolute disparity between the group’s percentage of the population as a whole and its percentage of the relevant jury venire. See Thomas, 159 F.3d at 1150. There is no bright line that separates constitutionally acceptable disparities from constitutionally infirm disparities, but the Supreme Court has found disparities as small as 14.7% sufficient to establish underrepresentation. See Jones v. Georgia, 389 U.S. 24, 88 S.Ct. 4, 19 L.Ed.2d 25 (1967).

Adopting Footracer’s argument, then, would sound the death knell of Rule 18, because, as a practical matter, it would deny district courts the discretion to hold trial anywhere within a district. Under Footracer’s proposed rule of law, district courts would have two unsavory options. They could try all criminal cases in the same divisions in which the crimes were committed. But this, of course, abolishes Rule 18. Alternatively, to keep Rule 18 alive, district courts would have to constantly update their demographic studies of the divisions in their districts, in order to avoid transferring a case to a neighboring division where the percentage of some racial, ethnic, or otherwise “distinctive” group is 14% less than in the division in which the crime was committed. Importantly, the district court would have to analyze not only the particular group to which a particular defendant belongs, but also every other distinctive group.13 The text of the Sixth Amendment does not demand such an extravagant waste of judicial resources, and common sense strongly counsels against it.

V.

Footracer has not shown that he was deprived of a jury venire reflecting a fair cross-section of the community. Native Americans, though a distinctive group, were not systematically excluded by the District of Arizona’s jury-selection process. The residents of the Prescott Division, though excluded by the jury-selection process, are not a distinctive group.

Footracer’s conviction is AFFIRMED.

dissenting:

Given the disparity between the racial and ethnic composition of the Phoenix Division jury pool and that of the communities comprising the Prescott Division,1 I think justice would have been better *1067served if Footracer’s trial had been returned to Prescott. In light of the serious constitutional concerns that Footracer’s motion to return raised, I would find that the district court acted outside its discretion by refusing to return the trial to Prescott.

This court has previously warned that the systematic transfer of cases from the Prescott Division to the Phoenix Division would raise serious Sixth Amendment concerns. See United States v. Etsitty, 130 F.3d 420, 425 (9th Cir. 1997), amended by 140 F.3d 1274 (1998). The majority disagrees that the Sixth Amendment would be violated by this systematic transfer, but reaches its decision by importing the equal protection concept of discriminatory intent into what is a straightforward Sixth Amendment fair cross-section challenge. Because I believe that this court’s warning in Etsitty was well-founded, and because the majority’s approach is directly contrary to Duren and its progeny, I respectfully dissent.

This court has warned, “[i]f ... trials were systematically removed from the Prescott Division to the Phoenix Division, we would be confronted with a strong case for finding an exclusion of Indians in violation of Duren [v. Missouri, 439 U.S. 357, 99 S.Ct. 664, 58 L.Ed.2d 579 (1979)]”.2 Etsitty, 130 F.3d at 425. I believe that this warning was well-founded.3

As the majority sets out above, see supra at 1060, a prima facie Duren violation has three components: (1) the excluded group must be a distinctive group in the community; (2) the excluded group must be underrepresented in jury venires; and (3) this underrepresentation must be due to systematic exclusion of the group in the jury-selection process. See Duren, 439 U.S. at 364, 99 S.Ct. 664. Once the defendant establishes a prima facie case, the government bears the burden of showing that a “significant [governmental] interest is manifestly and primarily advanced by those aspects of the jury-selection process ... that result in the disproportionate exclusion of a distinctive group.” Id. at 367-68, 99 S.Ct. 664.

The majority concedes that Native Americans are a “distinctive” group in the community. See supra at 1060; see also United States v. Brady, 579 F.2d 1121 1131 (9th Cir. 1978). It is also hardly controversial that the representation of Native Americans in the Phoenix venire is not “fair and reasonable” in relation to the number of Native Americans in the Prescott Division:4 the absolute disparity be*1068tween the percentage of jury-eligible Native Americans in the Prescott Division and the percentage of Native Americans in the Phoenix Division jury wheel is 20.48% (20.78% — 0.3%).5

But the majority rejects as dicta this court’s warning in Etsitty, 130 F.3d at 425, that the systematic transfer of eases from Prescott to Phoenix would constitute systematic exclusion of Native Americans in the jury selection process. See supra at 1060. The majority argues that the use of the term “exclusion” in Duren’s third prong, see 439 U.S. at 364, 99 S.Ct. 664 (“that underrepresentation is due to systematic exclusion of the group in the jury selection process”), implies that the Sixth Amendment is violated only where the particular distinctive group is treated differently than other groups. See supra at 1060-61. This argument not only goes beyond that advocated by the government,6 but also directly contradicts Duren and its progeny.

The Supreme Court made it clear in Duren, 439 U.S. at 368 n. 26, 99 S.Ct. 664, that it is not disparate treatment but disproportion itself that is the hallmark of a Sixth Amendment violation: “[I]n Sixth Amendment fair-cross section cases, systematic disproportion itself demonstrates an infringement of the defendant’s interest in a jury chosen from a fair community cross section.” Given this statement, there can be no question that it is disparate impact and not disparate treatment that the Sixth Amendment forbids.

Perhaps the majority got off on the wrong foot by focusing on the term “exclusion,” rather than the term “systematic.” See supra at 1060-61. But both the Du-ren opinion and the overwhelming weight of authority establish that the critical inquiry of Duren’s third prong is whether the underrepresentation of the distinctive group is systematic. See Duren, 439 U.S. at 366, 99 S.Ct. 664 (“his undisputed demonstration that a large discrepancy occurred not just occasionally but in every weekly venire for a period of nearly a year manifestly indicates that the cause of the underrepresentation was systematic ”) (emphasis added); id. at 367, 99 S.Ct. 664 (“[wjomen were therefore systematically underrepresented within the meaning of Taylor[ v. Louisiana, 419 U.S. 522, 95 S.Ct. 692, 42 L.Ed.2d 690 (1975) ]”) (emphasis added). As the First Circuit explained:

The issue is not whether the discrepancy was purposeful; Duren ... defined “systematic exclusion” as simply one that was “inherent in the particular jury-selection process;” viz., the system’s result, regardless of intent. We are not concerned with perfection, but, assuming a distinctive group, a process which, however neutral on its face, consistently underran by some 70%, providing one juror when there should have been four, is surely excessive.

LaRoche v. Perrin, 718 F.2d 500, 503 (1st Cir. 1983), overruled on other grounds by Barber v. Ponte, 772 F.2d 982 (1985). See also United States v. Jackman, 46 F.3d 1240, 1244, 1248 (2nd Cir. 1995) (holding that the “underrepresentation [of Hartford and New Britain residents, and thus blacks and Hispanics, in the jury pool] ‘was quite obviously due to the system by *1069which juries were selected’ ”) (quoting Duren, 439 U.S. at 367, 99 S.Ct. 664) (emphasis added).

By focusing on “exclusion,” and by defining “exclusion” to require a defendant to present evidence of disparate treatment, the majority mistakenly imports an equal protection concept into a fair cross-section challenge.7 Simply put, requiring a defendant to show disparate treatment requires him to show discriminatory intent,8 which is not an element of a Sixth Amendment fair cross-section challenge. See Thomas v. Borg, 159 F.3d 1147, 1150 (9th Cir. 1998).

The Sixth Amendment establishes a defendant’s right to be tried by a jury selected from a fair cross-section of the community. See Taylor v. Louisiana, 419 U.S. 522, 526-531, 95 S.Ct. 692, 42 L.Ed.2d 690 (1975). The Fifth Amendment protects against intentional discrimination in the selection of venires, see Castaneda v. Par-tida, 430 U.S. 482, 492-94, 97 S.Ct. 1272, 51 L.Ed.2d 498 (1977) (Fourteenth Amendment case), but the Sixth Amendment protects against unintentional deviations from the constitutional standard. See United States v. Perez-Hernandez, 672 F.2d 1380, 1384 n. 5 (11th Cir. 1982) (“[i]n a fair cross section analysis, purposeful discrimination is irrelevant since the emphasis is purely on the structure of the jury venire”); LaRoche, 718 F.2d at 503 (“Duren ... defined ‘systematic exclusion’ as simply one that was ‘inherent in the particular jury-selection process;’ viz., the system’s result, regardless of intent.”). When a defendant shows that a facially neutral jury selection process has the unintended effect of causing a substantial underrepresentation of a distinctive group, he makes out a prima facie fair cross-section violation. See United States v. Benmuhar, 658 F.2d 14, 19 (1st Cir. 1981) (holding that defendant made out a prima facie Sixth Amendment case that English proficiency requirement systematically excluded cognizable groups).

Nor do the cases the majority discusses support its contrary conclusion. The line of cases upholding the use of voter registration lists to compile eligible voter lists9 *1070is inapposite. As these cases demonstrate, courts’ acceptance of the use of voter registration lists has a two-pronged rationale: (1) courts attribute the underrepresentation of minorities on voter registration lists not to any governmental act, but to “personal predilection not to register” and “sloth,” see United States v. Cecil, 836 F.2d 1431, 1447-48 (4th Cir. 1988) (internal quotation marks and citation omitted); and (2) the alternative to using voter registration lists is daunting and impractical. See id. (“voter registration lists are imperfect ... but this does not render use of those lists unconstitutional, especially considering the alternative”). See also Truesdale v. Moore, 142 F.3d 749, 755 (4th Cir. 1998) (focusing on the lack of structural or systemic impediments to voter registration); Schanbarger v. Macy, 77 F.3d 1424, 1424 (2nd Cir. 1996) (same); United States v. Ireland, 62 F.3d 227, 231 (8th Cir. 1995) (same).

This two-pronged rationale does not support the systematic transfer of cases from the Prescott Division to the Phoenix Division. Given the history of the United States’ reservation policy, the location of the Native American reservations and the concentration of Native Americans on those reservations can hardly be attributed to “personal predilection” or “sloth.” And the alternative to transferring trials from Prescott to Phoenix is obvious: holding trials in Prescott. Therefore, the circumstances that justify the use of voter registration lists to compile eligible juror lists do not justify the systematic transfer of trials from Prescott to Phoenix. The majority’s reliance on the voter registration list cases to support its conclusion that Footracer has failed to satisfy the third Duren requirement is thus misplaced.

Once a defendant has established a pri-ma facie Sixth Amendment violation, the government bears the burden of showing that a significant governmental interest is manifestly and primarily advanced by the exclusionary practice. See Duren, 439 U.S. at 367-68, 99 S.Ct. 664. The government has not presented any interest, let alone a significant interest, that is served by the systematic transfer of cases from Prescott to Phoenix.

The majority’s discussion of Federal Rule of Criminal Procedure 18 suggests that the district court’s “Rule 18 right,” see supra at 1065, might provide such an interest. Rule 18 provides that prosecution shall be had in the district in which the offense was committed and vests discretion in the court to fix the place of trial at any place within the district. See United States v. Herbert, 698 F.2d 981, 984 (9th Cir. 1983). I doubt that even a proper exercise of discretion under Rule 18 could trump the Sixth Amendment. But because the district court did not apply Rule 18 with due regard to the relevant considerations, I would not reach the question whether a proper exercise of Rule 18 discretion would provide a governmental interest sufficient to justify the systematic exclusion of Native Americans.

As the majority notes, see supra at 1060, n. 1, the district court’s original order transferring Footracer’s trial to Phoenix did not offer any reason for the transfer. Moreover, even the reasons the district court later gave in its denial of Footracer’s motion to return-the convenience of Phoenix and the prompt administration of justice-do not justify its decision.

Rule 18 specifies that the relevant measure for “convenience” is the “convenience of the defendant and the witnesses,” not the convenience of the judge. Fed. R.Crim.P. 18; see Fed.R.Crim.P. 18, advisory committee’s note to 1979 amend, (noting that Rule 18 does not authorize consideration of the judge’s convenience when fixing of the place of trial). Footracer made a strong argument that Prescott was as, if not more, convenient than Phoenix: all of the witnesses resided either on the Navajo reservation or other areas of Northern Arizona; Prescott is several hours closer than Phoenix to the reservation; there are two federal courtrooms in Prescott; both the U.S. Attorney and the Federal Public Defender have offices in Prescott; and lodging is readily available.

*1071Moreover, Rule 18’s requirement that courts give due regard to the “prompt administration of justice” was intended to ensure compliance with the Speedy Trial Act of 1974. See Fed.R.Crim.P. 18, advisory committee’s note to 1979 amend. As such, the measure for prompt administration of justice is the speediness with which the defendant’s trial can be conducted. The district court, however, found that the prompt administration of justice favored trying the case in Phoenix “because of the court’s heavy civil docket which occasionally requires interruptions during trial.” (Emphasis added). In light of the serious constitutional concerns that Footracer raised in his motion to return, neither of the reasons the district court gave justified its decision to hold the trial in Phoenix.

Therefore, I believe that the district court’s refusal to return the trial to Prescott was at least outside her discretion and may have risen to a Sixth Amendment violation.10 For these reasons, I respectfully dissent.