HOUSTON LAWYERS' ASS'N, ET AL., PETITIONERS V. THE ATTORNEY GENERAL OF TEXAS, ET AL. LEAGUE OF UNITED LATEN AMERICAN CITIZENS, ET AL., PETITIONERS V. THE THE ATTORNEY GENERAL OF TEXAS, ET AL. No. 90-813, No. 90-974 In The Supreme Court Of The United States October Term, 1990 On Writ Of Certiorari To The United States Court Of Appeals For The Fifth Circuit Brief For The United States As Amicus Curiae Supporting Reversal TABLE OF CONTENTS Questions Presented Interest of the United States Statement A. The structure of the Texas judicial system 1. District courts 2. Specialized subject matter jurisdiction courts 3. Other courts B. The proceedings in this case 1. The district court's decision 2. The court of appeals' decision Summary of argument Argument: The results test of Section 2 of the Voting Rights Act of 1965 applies to the election of Texas district court judges A. The amended version of Section 2 applies to the election of state judges B. The election of Texas district judges is not exempt from the Section 2 results test on the ground that district judges occupy "single person offices" C. A state's strong, nondiscriminatory reasons for at-large judicial elections are among the "totality of circumstances" that courts must consider in determining whether there has been a violation of Section 2 1. A "strong state policy divorced from racial discrimination" supporting at-large judicial elections is a "circumstance" for the courts to consider in determining whether Section 2 has been violated 2. The district court did not consider whether the state has a strong, nondiscriminatory policy supporting the use of at-large judicial elections a. A representative yet impartial judiciary b. Electoral accountability c. Specialized courts d. Administrative benefits e. Other aspects of at-large judicial elections QUESTIONS PRESENTED 1. Whether the results test of Section 2 of the Voting Rights Act of 1965, 42 U.S.C. 1973, applies to the election of state court judges. 2. Whether the results test of Section 2 of the Voting Rights Act of 1965 applies to the election of offices that can be held by only one person. INTEREST OF THE UNITED STATES This case, like Chisom v. Roemer, No. 90-757 (consolidated with United States v. Roemer, No. 90-1032) involves the application of the vote dilution analysis of Section 2 of the Voting Rights Act of 1965, 42 U.S.C. 1973, to state judicial elections. The United States has the primary responsibility for enforcing Section 2, and the decision of the Court in this case will have an important effect on federal enforcement efforts. In addition, the Attorney General is responsible under Section 5 of the Voting Rights Act for preclearing voting changes, and under existing regulations must withhold preclearance if he concludes that such action "is necessary to prevent a clear violation of amended section 2." 28 C.F.R. 51.55(b)(2). This decision therefore will affect the manner in which the government reviews proposed voting changes in judicial election procedures under Section 5. The United States filed an amicus curiae brief when the case was heard by the panel and the en banc court, and presented oral argument to the en banc court. STATEMENT A. The Structure Of The Texas Judicial System 1. District courts. Texas district courts are the State's trial courts of general jurisdiction. Tex. Const. Art. 5, Section 8; see generally Tex. Judicial Council, Texas Judicial System: 61st Annual Report 10-17 (1989) (hereinafter Judicial Report). The Texas legislature creates district court seats as the nened arises, assigns the seats individual numbers, and defines the geographic area that they are to encompass. Tex. Gov't Code Ann. Section 24.101 et seq. (Vernon 1988). With one exception, all of the district court seats at issue in this case encompass single counties. /1/ A 1985 amendment to the Texas Constitution allowed redistricting, on the approval of the voters, of district court seats into districts consisting of portions of counties. Tex. Const. Art. 5, Section 7a(i). /2/ District court judges are elected to office for four-year terms through at-large, countywide elections. Candidates run for numbered seats in partisan contests. Tex. Const. Art. 5, Section 7a(i). While party primaries require a majority of votes to be nominated, the general election requires only a plurality to win. Tex. Election Code Section 172.003 (Vernon 1986). Interim district court vacancies are filled by gubernatorial appointment. Tex. Const. Art. 5, Section 28. See generally 90-813 Pet. App. 190a-191a. The jurisdiction of Texas district courts is statewide, while venue (i.e., the designation of which court with competent jurisdiction should hear a case) is determined by county. See Nipper v. U-Haul Co., 516 S.W.2d 467, 470 (Tex. Civ. App. 1974). /3/ The venue rules regarding the county in which a case should be filed are defined by general laws, see Tex. Civ. Prac. & Rem. Code Ann. Section 15.001 (Vernon 1986); Tex. Code Crim. Proc. Ann. Art. 13.10 et seq. (Vernon 1977), or by the mandatory or permissive provisions of particular statutes, see, e.g., Tex. Civ. Prac. & Rem. Code Ann. Sections 15.011-15.040 (Vernon 1986); see generally 72 Tex. Jur. 3d Venue Sections 1-99 (1990). District courts conduct proceedings at the county seat of the county in which the case is pending. Tex. Const. Art. 5, Section 7. Jury selection, case assignment, and record retention are handled on a countywide basis. See, e.g., 3 Tr. 267; 4 Tr. 255-256 (lawsuits in Harris County are filed at a central intake division and are randomly assigned to a district court, as are county residents reporting for jury duty), 4 Tr. 144 (Dallas County). Administrative coordination extends across counties as well. District court judges are sometimes assigned to hear cases in district courts in other counties, 5 Tr. 120, and this practice has been upheld against the challenge that judges have jurisdiction only in their own counties. Reed v. State, 500 S.W.2d 137, 138 (Tex. Crim. App. 1973). Under Texas law, the district courts of the State are divided into nine multicounty administrative judicial regions. See Tex. Gov't Code Ann. Section 74.042 (Vernon 1988). The presiding judge of each region is granted authority to promulgate regional rules of administration, advise judges on case flow management, recommend organizational changes to the Supreme Court, and oversee the assignment of judges within and between regions. Id. Sections 74.046-74.060. 2. Specialized subject matter jurisdiction courts. Although the state legislature cannot restrict the constitutional jurisdiction of district courts, /4/ laws containing new district court seats often express the intent that the judges elected to these posts give preference to certain types of matters. In this manner, the legislature has instructed certain district courts to give preference to family law matters, criminal matters, civil cases, or some combination of these areas, and some others appear to possess an informally designated subject-matter specialty. Judicial Report 17-19. Specialty courts are generally employed in populous, metropolitan counties. Id. at 10-11. Since these countis have many district courts, subject-matter preferences can be allocated among district court judges so that relatively large numbers of seats are assigned to each specialty. /5/ 3. Other courts. There are 14 intermediate courts of appeals providing appellate review of the judgments of the district courts. Each such appellate court has between 3 and 13 justices, all of whom are elected in partisan elections from geographic districts. Judicial Report 10-11. The jurisdiction of these appellate courts is "co-extensive with the limits of their respective districts." Tex. Const. Art. 5, Section 6. The two courts of last resort in the State, the Supreme Court and the Court of Criminal Appeals, have nine justices each, all of whom are elected in statewide at-large partisan elections. Judicial Report 10. The Texas system also includes county courts, municipal courts, and justice of the peace courts. Justices of the peace are elected from "commissioner's precincts," which divide each county into between four and eight subdistricts. Tex. Const. Art. 5, Section 18. In larger counties, more than one justice of the peace may be elected from each commissioner's precinct. Ibid. The Texas courts have held that justices of the peace have countywide territorial jurisdiction despite their election from individual precincts. See, e.g., Zulauf v. State, 591 S.W.2d 869, 872 & n.5 (Tex. Crim. App. 1979); Bradley v. Swearingen, 525 S.W.2d 280, 282 (Tex. Civ. App. 1975). B. The Proceedings In This Case The plaintiffs, black and Hispanic voters residing in nine Texas counties, /6/ filed this suit on July 11, 1988, alleging that the use of an at-large election scheme to elect state district court judges in these counties diluted the voting strength of black and Hispanic voters, in violation of Section 2 of the Voting Rights Act of 1965, 42 U.S.C. 1973. After a bench trial, the district court found that the plaintiffs had proved violations of Section 2 in each county. See Pet. App. 183a-304a. 1. The district court's decision. At the outset, the district court held that Section 2 applies to the election of judges. The court rejected the State's argument that Section 2 does not apply to the election of trial judges since trial judges act individually, not collegially. Relying on the Fifth Circuit's decision in Chisom v. Edwards, 839 F.2d 1056, cert. denied, 488 U.S. 955 (1988) (Chisom I), the court held that Section 2 applies to all judicial elections. Pet. App. 289a & n.32. The court then applied the vote dilution test set forth in Thornburg v. Gingles, 478 U.S. 30 (1986), to analyze plaintiffs' claims. The court found that plaintiffs had shown, in each county, (1) the existence of a minority group (black, Hispanic, or both) of sufficient size and geographical compactness to constitute a majority in a single-member district; (2) political cohesiveness within that minority group; and (3) racial bloc voting by the white majority that had consistently defeated the preferred candidates of the minority group. Pet. App. 289a-301a; see id. at 210a-273a. The court also found relevant three other factors drawn from the Senate Judiciary Committee Report accompanying the 1982 amendments to the Voting Rights Act, see S. Rep. No. 417, 97th Cong., 2d Sess. 28-29 (1982) (hereinafter Senate Report): the effects of past discrimination on minorities in Texas and the historical domination of the Texas judicial system by whites; use of numbered posts and a majority vote requirement in primaries; and the large size of five couties (Harris, Dallas, Tarrant, Bexar, and Travis). Pet. App. 273a-277a. The court also noted that minorities had not enjoyed electoral success since 1980. Id. at 279a-281a. The State maintained that at-large elections were necessary for several reasons: Judges elected from areas smaller than the existing county-wide districts would be susceptible to influence by organized crime and special interest or political groups; the at-large system had administrative advantages that would be lost if courts had less than county-wide jurisdiction; and modifying the present at-large system would be unduly costly. Pet. App. 281a. The State also maintained that judges should be elected from the same area in which they exercised their primary jurisdiction, and that members of the electorate would be disenfranchised in counties that used specialized courts if persons could not vote for judges who exercise each type of subject matter jurisdiction. Id. at 281a-282a. The district court found none of these arguments persuasive. It concluded that jury selection and other administrative functions could still be handled centrally on a county-wide basis, Pet. App. 284a, and that no compelling state policy required specialty court designations to be retained, id. at 283a-285a. /7/ In sum, the court found, based on the totality of the circumstances, that the plaintiffs had proved a violation of Section 2. Id. at 300a-301a. /8/ 2. The court of appeals' decision. The defendants appealed, and the en banc court of appeals reversed by a divided vote. League of United Latin American Citizens Council No. 4434 v. Clements, 914 F.2d 620 (5th Cir. 1990) (LULAC) (Pet. App. 1a-182a). /9/ The court held that although Section 2 generally applies to judicial elections, the vote dilution test of Section 2(b) does not, because judges are not "representatives" under Section 2(b) or as a general matter. Pet. App. 14a-35a. The court noted that the concept of minority vote dilution was modeled on the vote dilution standards developed in "one-person, one-vote" cases and that by 1981 numerous federal court decisions, including one by this Court -- Wells v. Edwards, 409 U.S. 1095 (1973), summarily aff'g 347 F. Supp. 453 (M.D. La. 1972) (three-judge court) -- had ruled that "the judicial office is not a representative one, most often in the context of deciding whether the one-man, one-vote rubric applied to judicial elections." Pet. App. 16a; id. at 16a n.9 (collecting cases). Applying the canon of construction that Congress is presumed to be aware of and endorse "the uniform construction" placed on a term, id. at 24a, the majority determined that Congress used the term "representatives" in order to apply the new results test of Section 2 to elections for representative, political offices but not to vote dilution claims in judicial contests. Id. at 16a-27a. The majority found unpersuasive the fact that the definitional provision of the Act, 42 U.S.C. 1973l(c)(1), defined "voting" by reference to "candidates for public or party office," because the term "representatives" in Section 2 was more specific. Pet. App. 27a-29a. Because Section 5 of the Voting Rights Act does not use the word "representatives," the majority also found irrelevant the fact that Section 5 applies to judicial elections. Pet. App. 29a. Six members of the en banc court, in three separate opinions, concluded that the dilution test of Section 2 applies to judicial elections. Judge Higginbotham, joined by three other judges, concluded that the term "representatives" encompasses elected judges. Pet. App. 51a-90a. He nevertheless concluded that the at-large election of trial judges in Texas does not violate Section 2 since each trial judge, like each governor, occupies a so-called "single member office" whose electorate cannot be further subdivided. In such instances, he said, electing all trial judges on an at-large basis does not dilute minority voting strength. Id. at 90a-114a. Concurring specially, Chief Judge Clark said that he agreed with Judge Higginbotham, adding that vote dilution analysis might be appropriate when a State elects its judges from single-member districts. Id. at 36a-46a. Judge Johnson dissented. In his view, the Section 2(b) vote dilution test applies to judicial elections, and the "single member office" exception did not apply to district court judges in Texas, because there were multiple officeholders at that level. Pet. App. 115a-182a. SUMMARY OF ARGUMENT A. The results test of Section 2 of the Voting Rights Act of 1965 applies to the election of state court judges. The original version of Section 2 of the Voting Rights Act covered the election of judges, and the amended version of Section 2 did nothing to change that. Congress did not amend the law to shorten its reach; instead, Congress revised Section 2 to enact the "results" test that this Court had rejected in City of Mobile v. Bolden, 446 U.S. 55 (1980). Thus, amended Section 2, like its predecessor, covers every election for public office. The Fifth Circuit erred in ruling that the term "representatives" in Section 2(b) exempts the election of judges from the results test of the statute. The legislative history of the 1982 amendments does not indicate that Congress used that term in order to limit the scope of Section 2. Congress used that term interchangeably with "candidates" and "elected officials" to refer to those elected to public or party office. The "one person, one vote" line of decisions holding that judges are not representatives is inapplicable here, since racial vote dilution differs from geographic vote dilution. B. The "single member office" doctrine is inapplicable to this case. That doctrine recognizes that a minority cannot prove a case of vote dilution when there is only one person who can hold a unique office. District court posts in Texas, however, are not unique. Each office is identical to every other such office. It is immaterial that each judge holds an office separately designated as a specific district court, because there is no functional difference between the differently enumerated district courts. C. The courts must consider a strong, nondiscriminatory state interest among the "totality of circumstances" under Section 2. This Court in Whitcomb v. Chavis and the Fifth Circuit in Zimmer v. McKeithen endorsed consideration of this factor. Although it does not "trump" proof of vote dilution, such an interest can spell the difference between a finding of unlawful vote dilution and a lawful state electoral practice. In this case, the district court did not apply the correct legal standard to the facts, so the case should be remanded for further proceedings. ARGUMENT THE RESULTS TEST OF SECTION 2 OF THE VOTING RIGHTS ACT OF 1965 APPLIES TO THE ELECTION OF TEXAS DISTRICT COURT JUDGES A. The Amended Version Of Section 2 Applies To The Election of State Judges The district court found that the Texas at-large system of electing trial judges violates Section 2 of the Voting Rights Act under the standard that this Court applied in Thornburg v. Gingles, 478 U.S. 30 (1986). The Fifth Circuit concluded that the Section 2 results test does not apply to judicial elections and therefore had no occasion to address the correctness of the district court's application of Gingles to the facts of this case. The questions, then, are whether the results test of Section 2 applies to judicial elections and, if so, how that test applies in such a case. In Chisom v. Roemer, No. 90-757 (consolidated with United States v. Roemer, No. 90-1032), we have taken the position that the results test of Section 2 applies to the election of state judges. /10/ As we explain in our brief in that case, the original version of Section 2 of the Voting Rights Act of 1965 covered the election of judges, and the amended version of Section 2 also applies to judicial elections. Congress did not amend that statute to shorten its reach; instead, Congress revised Section 2 in order to enact the "results" test this Court rejected in City of Mobile v. Bolden, 446 U.S. 55 (1980). In addition, the Fifth Circuit erred in ruling that the term "representatives" necessarily excludes elected judges. That term is best read to include all elected officeholders, whether in the legislative, executive, or judicial branch. For these reasons, the Section 2 results test applies to the election of state judges. B. The Election Of Texas District Judges Is Not Exempt From The Section 2 Results Test On The Ground That District Judges Occupy "Single Person Offices" In the Fifth Circuit, respondents argued that, even if Section 2 generally applies to judicial elections, district judges in Texas should be exempt from the Section 2 results test. Each district judge acts independently of every other judge, respondents argued, unlike an appellate judge, who is but one of several members of a collegial decisionmaking body, like a state legislature or a city council. Texas district judges, according to respondents, occupy what has been termed a "single person" or "single member office," to which a vote dilution claim is inapplicable. Several members of the court of appeals found that argument persuasive and would have reversed the district court's judgment on that ground. See Pet. App. 90a-114a (Higginbotham, J., concurring in the judgment). That conclusion, we believe, is mistaken. The "single member office" exception is less an exception to the Section 2 results test than it is a limitation on the ability of vote dilution theory to provide a useful means of analyzing and challenging the at-large election of offices that are held by only one person. Vote dilution theory has been used as a means of proving that an electoral scheme can dilute the collective voting strength of minority groups even though individual minority voters are free to cast ballots that are given the same weight as the ballots cast by the members of every other group in the community. As the Court explained in Thornburg v. Gingles, a successful challenge to an at-large system depends on proof that a politically cohesive minority group "is sufficiently large and geographically compact to constitute a majority in a single-member district." 478 U.S. at 50. Otherwise, a minority group's electoral failure does not necessarily signify that minority votes have been effectively cancelled out. Gingles also made clear that minority voters cannot claim to have been injured by a particular electoral structure, procedure, or practice (such as the use of an at-large election) unless minority voters have the potential to elect representatives in the absence of the challenged voting device. For example, if a minority group has the ability to elect legislators from single-member districts, but has been unable to do so in an at-large system, the group can challenge the State's use of an at-large system on the ground that it dilutes the votes of the minority group. A violation can be remedied by redrawing the geographic boundaries of the at-large system into single-member units. But if a minority group is dispersed throughout the districts, that group lacks the ability to elect representatives in a single-member system, and the at-large nature of the electoral system cannot be said to dilute minority votes. See id. at 50-51 & n.17. The vote dilution theory recognized in Thornburg v. Gingles cannot be applied to the at-large election of an office that is occupied by only one person to represent the entire district. In that event, a minority group, by definition, is not "sufficiently large and geographically compact to constitute a majority in a single-member district." 478 U.S. at 50. Under those conditions, it is not the at-large nature of a system that causes the dilution of minority votes, but the decision to elect only one person to the office involved and the impossibility of subdividing the one office into components. The vote dilution theory endorsed in Thornburg v. Gingles does not extend so far as to invalidate the use of such offices by requireing, for example, multiple governors or mayors. The decision in Butts v. City of New York, 779 F.2d 141 (2d Cir. 1985), cert. denied, 478 U.S. 1021 (1986), illustrates the "single person office" concept. Butts involved a Fourteenth Amendment and Section 2 challenge to a New York law requiring run-off primary elections if no candidate received 40% of the vote for the positions of mayor, city council president, and comptroller, each of which were held by only one person. The court held that the challenged law did not improperly dilute the votes of minorities because dilution principles governing election to multimember bodies, such as a city council, do not apply to single-member offices, like mayor, because "(t)here can be no equal opportunity for representation within an office filled by one person." 779 F.2d at 148. When elections are held for a multimember body, a minority group has the opportunity to secure "a share of representation" equal to other groups of citizens by electing members from districts in which the minority is dominant. But because "there is no such thing as a 'share' of a single-member office," the Second Circuit stated, vote dilution theory cannot be applied in the case of offices that govern the entire electorate and that are held by only one person. Ibid. See also United States v. Dallas County Comm., 850 F.2d 1430, 1432 n.1 (11th Cir. 1988) (relying on Butts v. New York to hold that the at-large election of a single probate judge is permissible). Cf. City of Port Arthur v. United States, 459 U.S. 159 (1982) (striking down run-off requirement for seats on multimember city council without mentioning the run-off requirement for mayor). Compare Dillard v. Crenshaw County, 831 F.2d 246, 251-252 (11th Cir. 1987) (chair of county commission was not treated as a single-member office); Southern Christian Leadership Conference v. Siegelman, 714 F. Supp. 511, 518-520 (M.D. Ala. 1989) (single-member office principle is inapplicable to elected Alabama circuit and district judges). /11/ Of course, it is theoretically possible to eliminate a single member office by creating a new office that can be occupied by independent officeholders either simultaneously or on a rotating basis. See Note, Applying Section 2 of the Voting Rights Act to Single-Member Offices, 88 Mich. L. Rev. 2199 (1990) (endorsing that proposal). The Voting Rights Act, however, does not reach that far. As Judge Higginbotham wrote, the Act does not require the States to restructure their elected positions, Pet. App. 90a-91a; instead, it guarantees minorities the same "opportunity" as other members of the electorate "to participate in the political process and to elect representatives of their choice." 42 U.S.C. 1973(b). Section 2 expressly disavows any requirement of proportional representation for minority groups, and Congress made clear that the Act does not prohibit all use of at-large systems. See Senate Report 2, 16, 31-33, 68. See also Whitcomb v. Chavis, 403 U.S. 124 (1971) (upholding an at-large system over a racial note dilution claim); id. at 156-160 (holding that multimember systems are not per se unconstitutional); White v. Regester, 412 U.S. 755, 765 (1973), and Zimmer v. McKeithen, 485 F.2d 1297, 1304 (5th Cir. 1973), aff'd on other grounds sub nom. East Carroll Parish School Bd. v. Marshall, 424 U.S. 636 (1976) (stating that at-large systems are not per se unconstitutional), cited at Senate Report 33. Accordingly, the vote dilution theory of Thornburg v. Gingles does not invalidate single member offices. The single member office principle, however, only applies to a narrow category of elected offices: ones for which the electroal district cannot be subdivided because only one official is elected from a geographic region. That principle, accordingly, does not apply in this case, given the structure of the Texas district court system. Each county elects between 3 and 59 district court judges; none elects only one. Nor does each district judge have unique responsibilities; all have the same authority and exercise the same responsibility. Some judges specialize in criminal, civil, or family law cases, but each district court within each county is interchangeable with every other office, and courts with the same specialization are entirely fungible. Under these circumstances, the Texas electoral system in each county can be subdivided into separate units without having to perform surgery on the functions performed by the office of district court judge. The single member office barrier to the application of the Section 2 vote dilution analysis does not exist in this case. Judge Higginbotham concluded that a function of the office of district judge is to represent the entire community, and that creating subdivisions within counties would thus alter the nature and function of that office. Pet. App. 103a-112a. But if that analysis were correct, no at-large election to a multimember body could ever be successfully challenged under Section 2, because every State could maintain that it is the function of each officeholder to represent the entire community. That proposition may be true, but it is also largely immaterial to the question of coverage. Rather, it is an interest to be considered under the "totality of circumstances." What is pertinent is whether the functions that an office is empowered to carry out under state law are unique, not whether an officeholder can be said to represent a particular geographic community when exercising the power of his office. The district courts in this case are not unique, and the single member office doctrine thus does not apply. Respondents also maintained below that district judges occupy single member offices because each judge is elected from a separate numbered district. Although it is true that district judges are technically elected to different judicial districts, each district in a county is identical to every other district. The designation of a judge as sitting for a particular numbered judicial district thus does not affect the nature of the position that judge holds, and therefore is of no consequence to the Section 2 analysis. For example, by designating 59 judicial districts in Harris County, all of which cover the same geographic area, Texas has just identified different numbered posts for election purposes, thereby allowing candidates to run in head-to-head contests. That designation does not give rise to a "single person office" for purposes of Section 2 because it does not affect the nature of the office itself. C. A State's Strong, Nondiscriminatory Reasons For At-Large Judicial Elections Are Among The "Totality Of Circumstances" That Courts Must Consider In Determining Whether There Has Been A Violation Of Section 2 Texas maintains that the at-large election of district judges helps ensure that they are responsive to the community while at the same time lessening the risk that they will be susceptible to undue influence by particular components of it. We disagree with respondents that this justification is sufficient automatically to exempt judges from Section 2, but we believe that such a justificaton is a legitimate one and must be considered by the courts among the "totality of circumstances" in applying Section 2. 42 U.S.C. 1973(b). In some instances, that interest can spell the difference between a lawful and unlawful electoral scheme. Since the district court did not apply the correct legal standard to the facts, the judgment should be vacated and the case remanded for further proceedings. 1. A "strong state policy divorced from racial discrimination" supporting at-large judicial elections is a "circumstance" for the courts to consider in determining whether Section 2 has been violated Historically, electoral schemes were initially challenged on the ground that the malapportionment of representatives effectively diluted the votes of individual electors. Reynolds v. Sims, 377 U.S. 533 (1964). An at-large election can remedy that problem, but also can lead to a different one, group vote dilution. Group vote dilution occurs when the practical operation of an electoral system effectively erases or minimizes the voting strength of a particular group, such as a racial minority. Fortson v. Dorsey, 379 U.S. 433, 439 (1965). To prove that an at-large electoral system dilutes minority votes, a plaintiff must "produce evidence to support findings that the political processes leading to nomination and election were not equally open to participation by the group in question -- that its members had less opportunity than did other residents in the district to participate in the political processes and to elect legislators of their choice." White v. Regester, 412 U.S. at 766. See also Whitcomb v. Chavis, 403 U.S. at 149-153. In Zimmer v. McKeithen, the Fifth Circuit distilled from this Court's decisions in Whitcomb and White a list of factors for the courts to use when analyzing a racial vote dilution claim. 485 F.2d at 1305. The later significance of the Fifth Circuit's discussion merits lengthy quotation, ibid. (footnotes omitted): The Supreme Court has identified a panoply of factors, any number of which may contribute to the existence of dilution. Clearly, it is not enough to prove a mere disparity between the number of minority residents and the number of minority representatives. Where it is apparent that a minority is afforded the opportunity to participate in the slating of candidates to represent its area, that the representatives slated and elected provide representation responsive to minority's needs, and that the use of a multi-member districting scheme is rooted in a strong state policy divorced from the maintenance of racial discrimination, Whitcomb v. Chavis, supra, would require a holding of no dilution. Whitcomb would not be controlling, however, where the state policy favoring multi-member or at-large districting schemes is rooted in racial discimination. Conversely, where a minority can demonstrate a lack of access to the process of slating candidates, the unresponsiveness of legislators to their particularized interests, a tenuous state policy underlying the preference for multi-member or at-large districting, or that the existence of past discrimination in general precludes the effective participation in the election system, a strong case is made. Such proof is enhanced by a showing of the existence of large districts, majority vote requirements, anti-single shot voting provisions and the lack of provision for at-large candidates running from particular geographical subdistricts. The Fifth Circuit and other courts relied on what became known as the "Zimmer factors" in the vast majority of subsequent cases. E.g., David v. Garrison, 553 F.2d 923, 927 (5th Cir. 1977). The Senate Judiciary Committee Report accompanying the Voting Rights Act Amendments of 1982 noted that fact, Senate Report 23, and endorsed that approach to the resolution of Section 2 claims, Senate Report 27-30. The Report also set out a list of "(t)ypical factors" that a plaintiff could invoke to prove racial vote dilution, id. at 28-29, that was "derived from the analytical framework used by th(is) Court in White, as articulated in Zimmer." Id. at 28 n.113. See Thornburg v. Gingles, 478 U.S. at 44-45. One factor expressly mentioned in Zimmer was whether "the use of a multi-member districting scheme is rooted in a strong state policy divorced from the maintenance of racial discrimination." 485 F.2d at 1305. The Fifth Circuit in Zimmer believed -- correctly, in our view -- that this Court had endorsed that factor in Whitcomb v. Chavis. 485 F.2d at 1305 (citing Whitcomb). The Fifth Circuit often reiterated that factor in later cases, although it did not rely on it to uphold an at-large electoral scheme. /12/ Nevertheless, for several reasons, we believe that strong state policies underlying the at-large election of judges are factors that the courts must consider in determining whether such a system violates Section 2. First, the text of Section 2 requires the courts to consider the "totality of circumstances" when analyzing a Section 2 claim. This Court has recognized that the State's justification for its electoral system is a proper factor for the courts to assess in a racial vote dilution inquiry, see Whitcomb v. Chavis, 403 U.S. at 149, while the Fifth Circuit in Zimmer expressly approved the use of this particular factor in the balance of considerations. 485 F.2d at 1305. The Senate Report on the 1982 Amendments favorably cited Whitcomb and Zimmer when discussing how amended Section 2 would operate. See Senate Report 24-33. Second, Section 2 applies to all state electoral mechanisms, not simply the use of at-large elections. For instance, a state law requiring a person to be a member of the state bar for a set number of years before becoming a judge could conceivably be challenged on the ground that it dilutes the pool of candidates for the bench. For that matter, a rule that the candidate with the most votes wins could conceivably be challenged on the ground that it prevents minority groups from electing a representative to a single-member office, like governor. If so, it would be odd in the extreme for Congress to have prevented the courts from considering the legitimacy and weight of the State's interest in such rules, focusing instead solely on their dilutive effect. What is more, focusing exclusively on the dilutive effects of a state electoral practice while not giving any weight to the State's interests could well lead to requiring proportional representation, which the Act expressly states is not required. Congress hardly intended that the operation of Section 2 would result in an outcome that its text expressly disavows. At the same time, we do not mean to suggest that proof of a strong state interest automatically trumps proof of racial vote dilution. Section 2 is a broadly based prohibition on electoral practices that, intentionally or not, deny minority voters equal access to the electoral process. As a footnote in the Senate Report explains, "even a consistently applied practice premised on a racially neutral policy would not negate a plaintiff's showing through other factors (derived from Zimmer v. McKeithen, supra) that the challenged practice denies minorities fair access to the process." Senate Report 29 n.117. See Bolden v. City of Mobile, 571 F.2d 238, 244 (5th Cir. 1978) ("(c)ity-wide representation is a legitimate interest, and at-large districting is ordinarily an acceptable means of preserving that interest (but) the district court was warranted in finding that the city's interests in its at-large plan did not outweigh the strong showings by the appellees under the other Zimmer criteria"), rev'd, 446 U.S. 55 (1980). The often dilutive effect of at-large elections on minority voting strength was of considerable importance to Congress in 1982. Congress knew that many localities had enacted at-large systems for legitimate governmental reasons, and no one disputed that there was legitimate governmental support for at-large elections. "The reason usually given in support of at-large elections for municipal offices is that at-large representatives will be free from possible ward parochialism and will keep the interests of the entire city in mind as they discharge their duties." Wallace v. House, 515 F.2d 619, 633 (5th Cir. 1975), vacated on other grounds, 425 U.S. 947 (1976). See also McMillan v. Escambia County, 638 F.2d 1239, 1244 (5th Cir. 1981) (describing "good government" basis for at-large elections). Congress explicitly relied on a list of 23 lower court decisions that applied dilution principles. Senate Report 32. In several of those cases, where defendants had alleged race-neutral grounds for at-large elections, the courts held that the elections were unlawfully dilutive and required the adoption of a single-member districting system for future elections to ensure that the State's electoral methods did not dilute minority voting. /13/ Analysis should also be informed, however, by the fact that the role of judges differs from those of legislative and executive officials. In balancing the strength of race-neutral state policies against evidence of vote dilution under the "totality of circumstances," it is obviously pertinent to consider the nature of the office at issue. The most obvious difference is that while legislators, and, perhaps, to a lesser extent, executive officials are expected to advance and protect the interests of their constituents, and are elected to do just that, judges are expected to be fair and impartial. Thus, "responsiveness" to minority voters in not a relevant concern in evaluating judicial elections. The important state interest in ensuring a fair and impartial judiciary must also be carefully considered in evaluating a State's decision to elect judges at-large. The State may believe that judges should be discouraged from thinking of themselves as representing only a portion of a particular jurisdiction. /14/ In addition, a State may determine that small electoral districts must be avoided in order to prevent a relatively discrete segment of the jurisdiction from controlling the election of trial judges. A State may determine that fairness, impartiality, and public confidence are significantly aided where all the people who may generally appear before a particular judge have a voice in the election of that judge. /15/ Finally, to the extent that there are legitimate and strong state interests in the at-large election of trial or appellate judges, that is powerful evidence that minority electoral failure is not the product of a "built-in bias" against minorities but stems, instead, from other, neutral factors. Whitcomb v. Chavis, 403 U.S. at 153. Considering the State's interests -- which may be different in both nature and magnitude for the at-large election of judges than for the at-large election of legislators -- is therefore consistent with the principle in the case law and legislative history that "'the question whether the political processes are 'equally open' depends upon a searching practical evaluation of the 'past and present reality' and on a 'functional' view of the political process." Thornburg v. Gingles, 478 U.S. at 45 (quoting Senate Report 30 & n.120). See White v. Regester, 412 U.S. at 766-767; Whitcomb v. Chavis, 403 U.S. at 149-155. Section 2 is broad in its reach, but there is no reason to believe that, in passing the amendment in 1982, Congress sought to alter the fundamental nature of judicial offices established by the State or require a method of election that irreconcilably conflicts with those offices. After all, Section 2 is addressed to voting practices, not to the definition of state offices. Thus, if Texas consistently elects trial or appellate judges at-large in order to ensure the appearance of fairness and impartiality in the judicial process, its interest may not be just "strong" but "compelling." Such an interest should be considered along with other factors in the totality of circumstances. 2. The district court did not consider whether the State has a strong, nondiscriminatory policy supporting the use of at-large judicial elections Judge Higginbotham concluded that Texas has tied together a district court's electoral and jurisdictional base as a means of protecting each judge's independence without compromising the fact and appearance of impartiality. The independent decisionmaking authority of each trial judge is "a structure we must accept," Pet. App. 104a, Judge Higginbotham concluded, and a federal court cannot create or direct the State to create single-member districts for judicial elections because doing so would interfere with the State's achievement of those critical governmental interests. Id. at 108a-112a. We believe that Judge Higginbotham's conclusions are at least premature, as the facts necessary to a proper assessment of the way in which Texas uses at-large elections to advance governmental interests, and the strength of those interests, have been neither fully developed nor analyzed. It is not clear that Texas has met the standard in this case of demonstrating that at-large elections of trial judges serve a strong state policy. Nor has the state policy been weighed against the evidence of vote dilution in this case. The case therefore should be remanded so the district court can carry out the appropriate analysis. a. A representative yet impartial judiciary. The State's principal justification for holding at-large judicial elections is that using subdistricts could lead to the fact and perception of judicial bias and undue influence by special interest groups. This rationale has been used to defend at-large elections for legislative and executive offices, see City of Mobile v. Bolden, 446 U.S. at 70 n.15 (plurality opinion); Zimmer v. McKeithen, 485 F.2d at 1301; cf. The Federalist No. 10, at 82-84 (J. Madison) (C. Rossiter ed. 1961), /16/ but assumes heightened importance in judicial elections, for all the reasons the judiciary is different from the other branches. Several witnesses testified that at-large election of judges was necessary for this reason, /17/ and this factor may be one of the "totality of circumstances" showing that minority electoral failure cannot be attributed to a state system that improperly dilutes minority votes. There is, however, evidence tending to disprove the claim that at-large county-wide elections are essential to ensure a representative and impartial bench. The Texas Constitution does not require county-wide election of the district judges; it permits the voters to select them from sub-county districts. Tex. Const. Art. 5, Section 7a(i). Justices-of-the-peace are elected from precincts smaller than a county, and some election districts of a similar size could be created here. /18/ There also was testimony that justices-of-the-peace, who are elected from such precincts, had no difficulties with bias or allegations of bias. 4 Tr. 90. Electoral districts that are smaller than a county can still be quite large. Harris County, for example, has a population of nearly 2.5 million and 59 district judges. Even were Harris County to be divided into 59 subdistricts (a remedy that we do not contend is required), each district would contain approximately 41,000 people. /19/ Moreover, were a remedy necessary, the use of several multimember districts in large counties might be far preferable to the use of single-member districts. Finally, the potential for the fact or appearance of bias is present whenever judges are elected, while the interest in making judges accountable to minority voters can perhaps be served only by eliminating an at-large system. b. Electoral accountability. A related justification for holding countywide at-large elections is that it helps to ensure that judges are accountable to the persons over whom they exercise jurisdiction. /20/ That rationale is a legitimate one, but its force is weakened by the fact that Texas law does not treat all judges alike. Justices-of-the-peace, which are trial courts below the level of district courts, are elected from sub-county precincts even though these officers exercise jurisdiction over an entire county. /22/ District courts also can exercise jurisdiction over cases arising beyond the county boundaries since the parties can agree to permit a court to adjudicate a case that does not arise within the county. Page 3 & note 3, supra. Furthermore, district court judges often hear cases in other counties to help manage the docket, 5 Tr. 120, and the residents of a county in which a district judge temporarily sits have no recourse against him at the polls. This disuniformity may signify that the accountability rationale is less weighty in practice than might otherwise be the case. c. Specialized courts. District Judge Davidson testified that applying the results test of Section 2 could disrupt the use of specialized courts. Because the counties that use specialized courts generally have many of each type of court, pages 4-5 note 5, supra, the State could retain the use of specialty courts in large counties and still remedy dilution, were any such remedy necessary, by, for example, dividing the specialty courts among districts and having each district elect each type of judge. d. Administrative benefits. Several witnesses testified that county-wide elections have administrative advantages that would be lost if smaller electoral districts were used, such as county-wide records retention, random assignment of cases to judges within the county (which aids docket control), and county-wide jury selection. 3 Tr. 257, 264; 4 Tr. 257, 261. Such benefits may well be valuable, but by themselves are insufficient to justify a racially dilutive electoral process. See Westwego Citizens for Better Gov't v. Westwego, 872 F.2d 1201, 1211 (5th Cir. 1989). e. Other aspects of at-large judicial elections. It may be possible to modify other aspects of the electoral apparatus in a fashion that ameliorates the dilutive effect of an at-large system. For instance, judges are elected in head-to-head contests. That factor, which the legislative history of Section 2 identified as a common method of diluting minority voting strength, Senate Report 29, could be modified without also eliminating at-large elections. In sum, the careful assessment of the strength of the State's interests in the present method of election has not been properly performed, nor have those interests been balanced against the showing of vote dilution. The case should be remanded for that purpose. CONCLUSION The judgment of the court of appeals should be reversed and the case remanded for further proceedings. Respectfully submitted. KENNETH W. STARR Solicitor General JOHN R. DUNNE Assistant Attorney General JOHN G. ROBERTS, JR. Deputy Solicitor General ROGER CLEGG Deputy Assistant Attorney General PAUL J. LARKIN, JR. Assistant to the Solicitor General JESSICA DUNSAY SILVER MARK L. GROSS Attorneys MARCH 1991 /1/ The 72nd Judicial District encompasses Lubbock and Crosby counties. See Tex. Gov't Code Ann. Section 24.174 (Vernon 1988) /2/ Although Art. 5, Section 7a(i) permits voters to create subcounty districts, it is not clear whether a court so created could still exercise county-wide authority. /3/ Nipper held that the Travis County district court had jurisdiction over a claim arising in Bexar County, and that the parties, both of whom lived in Bexar County, had waived their venue right to have the case transferred to Bexar County. See also Jernigan v. Jernigan, 467 S.W.2d 621 (Tex. Civ. App. 1971) (drawing same distinction between jurisdiction and venue). /4/ See, e.g., Lord v. Clayton, 163 Tex. 62, 67, 353 S.W.2d 718, 721-722 (1961); Ex parte Richards, 137 Tex. 520, 155 S.W.2d 597, 599 (1941). /5/ According to the most recent report of the Texas Judicial Council, the 59 district court seats in Harris County are assigned subject matters as follows: 31 courts have general district court jurisdiction or a civil preference; 16 courts have preference for criminal cases; and 12 courts specialize in family law. Judicial Report 54. The 37 seats in Dallas County are divided as follows: 13 courts have general district court jurisdiction or a civil preference; 15 courts are assigned a criminal preference; and 9 courts are designated as family courts or assigned a family law preference. Id. at 45. Tarrant County has 11 courts with general jurisdiction or a civil preference; 11 courts with a criminal preference or designation; and 7 courts with a family law preference. Id. at 76. Jefferson County has 4 courts of general district court jurisdiction; 2 courts handling criminal matters; and 2 courts handling family law. Id. at 58. One of Travis County's 13 district court judges is assigned a criminal preference, id. at 78, and one of Midland County's three courts is designated as a family court, id. at 66. Ector and Lubbock Counties do not use specailized courts. Id. at 48, 63. /6/ These nine counties are: Harris County, which has a population of more than 2,400,000 and chooses 59 district court judges in at-large elections; Dallas County, which has a population of more than 1,500,000 and elects 37 judges at-large; Tarrant County, which has more than 860,000 people and elects 23 judges at-large; Bexar County, wose population of nearly 1,000,000 chooses 19 judges in at-large elections; Travis County, which elects 13 judges and has a population of nearly 420,000; Jefferson County, which has 8 judges and more than 250,000 people; Lubbock County, with 6 judges and more than 211,000 people; Ector County, with 4 judges and more than 115,000 people; and Midland County, whose 82,000 plus residents elect 3 judges at-large. 90-813 Pet. App. 200a-209a. Hereinafter we will use "Pet. App." to refer to the petition appendix in 90-813. /7/ The district court also rejected the State's defense that partisan politics and not race explained the results of elections in counties where political party primaries were not the impediment to minority electoral success. Because the court of appeals held that judges are exempt from the results test of Section 2, it did not address this issue. /8/ In January 1990, the district court entered an orde postponing upcoming elections until November 1990 in order to give the Texas legilature additional time to enact its own remedy. The court of appeals thereafter stayed the district court's order, permiting the elections to go forward. /9/ A panel also initially reversed the district court's judgment. League of United Latin American Citizens Council No. 4434 v. Clements, 902 F.2d 293 (5th Cir. 1990). The majority held that Section 2 of the Voting Rights Act applies to the election of judges, 902 F.2d at 295-303, but that the results test of Section 2 does not apply to Texas trial judges, since they occupy "single person" or "single member offices," id. at 303-308. Judge Johnson dissented. He agreed with the majority that Section 2 applies to judicial elections, but he did not believe that trial judges in Texas hold "single member offices." Id. at 309-321. The court then decided to rehear the case en banc. 902 F.2d 322 (1990). /10/ We have provided copy of our brief in Nos. 90-757 & 90-1032 to counsel in this case. /11/ In Butts, the court of appeals held that a plurality vote runoff requirement for a single-person office was not subject to Section 2. 799 F.2d at 151. In our view, the method of electing a single-member office is not immune from attack under Section 2; minority voters should be able to challenge hurdles to success other than the at-large structure of an election. /12/ See Turner v. McKeithen, 490 F.2d 191, 194 (5th Cir. 1973) ("the strength of the state interest in multi-member or at-large voting"); Robinson v. Anderson County Comm'rs Court, 505 F.2d 674, 680 (5th Cir. 1974) (considering and rejecting justification for apportionment; "the mere fact that an approtionment plan may satisfy some legitimate governmental goals does not automatically immunize it from constitutional attack on the ground that it has offended more fundamental criteria"); Bradas v. Rapides Parish Police Jury, 508 F.2d 1109, 1112 (5th Cir. 1975) (quoting Turner); David v. Garrison, 553 F.2d 923, 927, 930 (5th Cir. 1977); Parnell v. Rapides Parish School Bd., 563 F.2d 180, 184 (5th Cir. 1977); Bolden v. City of Mobile, 571 F.2d 238, 244 (5th Cir. 1978), rev'd 466 U.S. 55 (1980). Cf. Hendrix v. Joseph, 559 F.2d 1265, 1269 (5th Cir. 1977); McGill v. Gadsden County Comm'n 535 F.2d 277, 280-281 (5th Cir. 1976) (plaintiff did not show that state policy was tenuous or discriminatory). Compare Moore v. Leflore County Bd. of Election Comm'rs, 502 F.2d 621, 624-625 (5th Cir. 1974) (in reviewing the remedy imposed by the district court, "(e)qualization of land area and road mileage is extremely important here"). /13/ Zimmer v. McKeithen, 485 F.2d at 1307; Hendrix v. Joseph, 559 F.2d 1265 (5th Cir. 1977); Kendrick v. Walder, 527 F.2d 44 (7th Cir. 1975). See also Wallace v. House, 515 F.2d at 632-633; Bolden v. City of Mobile, 571 F.2d at 244. /14/ Of course, where the plaintiff can prove that the adoption or maintenance of an at-large system, whenever it occurred, was motivated by racial discrimination, the State's interests are not entitled to deference. /15/ A State may conclude that its court of last resort is the autoritative source of state law, including the state consitution, and that all of the people in the State should have a vote for each member of the court entrusted with that responsibility. Once a State, as here, decides to elect judges from areas smaller than the entire State, closer scrutiny is appropriate, as the State has already determined to make the judge accountable to only part of the State's electorate. /16/ See also Dallas County v. Reese, 421 U.S. 477, 479-480 (1975); Chapman v. Meier, 420 U.S. 1, 20 n.14 (1975); Dusch v. Davis, 387 U.S. 112 (1967); Fotson v. Dorsey, 379 U.S. at 438 (all concluding that multimember districts can be justified on the ground that an officeholder represents the entire geographic region, not simply one district). /17/ For example, District Judge Mark Davidson testified that holding elections for judges from districts that are smaller than the existing county-wide districts woul mean that "the political pressures, at least on judges from Harris County, would increase substantially," 3 Tr. 264, and "forum shopping * * * would increase drematically," as parties sought to "find an attorney * * * with some political pressure, for example, over the Judge of the Court whose case you fell in," 3 Tr. 265. Professor Anthony Champagne opposed creating subdistricts because "(t)heoretically the larger the population you serve the more insulated a Judge would be from special interest group pressure." 4 Tr. 146. District Judge Carolyn Wright expressed the same opinion. 4 Tr. 191. District Judge Harold Entz testified that using smaller elctoral districts could give rise to a public perception that some judges would be elected, and some litigants would be successful, simply because of their race. 4 Tr. 82, 88-90. He also expressed concern that smaller judicial districts could help enable large-scale drug traffickers to influence the outcome of elections. 4 Tr. 83-84. /18/ County population ranges from 82,000 (Midland County) to 2.4 million (Harris County). Under the Texas Constitution, counties with a population of 30,000 have 4-8 justices-of-the-peace subdistricts, whole counties with a population of 18,000 can have 2-5 such precincts. Tex. Const. Art. 5, Section 18(a). /19/ Dallas County, with a population of 1.5 million people, has 37 judges. Were it divided into 37 subdistricts, each one would contain about 42,000 people. Ther are currently judges elected county-wide from counties with populations of similar size. By our count, 114 of Texas's 362 district courts are elected from areas of less than 100,000 people, and 61 of those are elected from areas of 50,000 or less. (County populations were determined by reference to United States Department of Commerce, 1980 Census of Population -- Texas (1982).) /20/ For instance, Texas Supreme Court Chief Justice Thomas Phillips, testifying for the State, said that at-large elections assured that judges "ought to be accoutable to those people who can be hailed into their Court," 5 Tr. 120, and trial judges therefore ought to be elected from the same area over which the court has judisdiction. /21/ Witnesses also testified that votes for judges were cast based on factors such as party affiliation, name recognition, or other considerations unrelated to judicial performance. 3 Tr. 270; 4 Tr. 120; 5 Tr. 129. Although that fact would tend to undercut the State's asserted interest in accountability, because the same criticism can be lodged agains any electoral scheme, it is not clear that this criticism has any weight here. /22/ Justices-of-the-peace can try cases that arise in other precincts. Bradley v. Swearingen, 525 S.W.2d 280, 282 (Tex. Civ. App. 1975). See also Zulauf v. State, 591 S.W.2d 869, 872 & n.5 (Tex. Crim. App. 1979). APPENDIX