League of United Latin American Citizens, Council No. 4434,plaintiffs-appellees,andjessie Oliver, et al., Intervening Plaintiffs-appellees, v. William P. Clements, Etc., et al., Defendants.jim Mattox, et al., Defendants-appellees, Appellants, v. Judge F. Harold Entz, Etc., Judge Sharolyn Wood, Etc., Andgeorge S. Bayoud, Jr., Etc., Defendants-appellants,andtom Rickhoff, Susan D. Reed, John J. Specia, Jr., Sid L.harle, Sharon Macrae and Michael P. Pedan, Bexarcounty, Texas State District Judges, Appellants, 986 F.2d 728 (5th Cir. 1993)

This case is before us on remand from the Supreme Court's decision in Houston Lawyers' Association v. Attorney General of Texas, --- U.S. ----, 111 S. Ct. 2376, 115 L. Ed. 2d 379 (1991), in which the Court held that Section 2 of the Voting Rights Act, 42 U.S.C. § 1973, applies to all Texas judicial elections. We must now address a question that is undoubtedly easier to frame than to answer. In particular, we must decide whether the district court erred in concluding that the method by which Texas elects district court judges--as that method operates in nine counties--violates Section 2 of the Voting Rights Act. After a careful review of the record, we hold (1) that the district court correctly concluded, in eight of the counties at issue, that Texas' method of electing district court judges violates Section 2, but (2) that the district court erred in finding a Section 2 violation in Travis County. We therefore affirm the district court's decision in part, reverse the decision in part, and remand the case to the district court for consideration and imposition of an appropriate remedy.

In view of the length of this opinion,1 a summary of the pertinent facts and major legal conclusions may be helpful to the reader:

As with all cases under the Voting Rights Act, this one is driven by the facts. In this case, certain key facts are best summarized by the following table which sets forth, with respect to each of the nine Texas counties at issue, the population of the relevant minority group, the total number of district court judges elected in the county, the number of judges who are members of the relevant minority group and the percentage of the total number of district judges who are members of the relevant minority group.

NOTE--Some parts of this form are wider than one screen. To view

material that exceeds the width of this screen, use the right arrow

key. To return to the original screen, use the left arrow key.

County Relevant Total no. Number of Percentage minority of judges who of judges population district are members who area judges of the members ofb relevant relevant minority minority groupc group ----------------------------------------------------------------------------------------- Bexar 46.6% Hispanic 19 5 26.3% Dallas 18.5% Black 36 2d 5.5% Ector 26.0% Black & 4 0 0.0% Hispanic Harris 19.7% Black 59 3 5.1% Jefferson 28.2% Black 8 0 0.0% Lubbock 27.0% Black & 5 0 0.0% Hispanic Midland 23.5% Black & 3 0 0.0% Hispanic Tarrant 11.8% Black 23 2e 8.7% Travis 17.2% Hispanic 13 0 0.0%

PATRICK E. HIGGINBOTHAM, Circuit Judge, dissenting:

* History will with the perspective of time mark the Voting Rights Act as the single most important act of government in the black civil rights movement since the Emancipation Proclamation and the Civil Rights Amendments. Its constitutionality was anchored in the explicit grants of power to the Congress of the Thirteenth, Fourteenth, and Fifteenth Amendments. At the same time the Voting Rights Act as the act of the representative branch stands on ground beyond the reach of the most dedicated process-oriented legal critics. Its mission was to deliver the constitutionally secured promise of power. Its result was more swift and immediately far-reaching than Brown v. Board of Education, 347 U.S. 483, 74 S. Ct. 686, 98 L. Ed. 873 (1954). Its allocation of power found expression at every level. The story can be elaborated with reams of data and a library of events but its soaking penetration into the daily world of the black, and the southern black in particular is captured by two riveting scenes.

George Wallace campaigning in Alabama in 1963 told voters, "I draw the line in the dust and toss the gauntlet before the feet of tyranny, and I say: segregation now, segregation tomorrow, segregation forever." Within ten years Governor Wallace moved from his pugnacious stance in the door at the University of Alabama to the center of the playing field of Bryant-Denny Stadium crowning a black homecoming queen. It was simultaneously a trip of a quarter of a mile and light years. Wallace's rhetoric of segregation now and forever gave way to "We're all God's children. All God's children are equal."

The Constitution allocates power and at every turn attempts to check its use. As wise as its diffusion has proved to be, our constitutional structure ultimately rests on the ballot box. The Voting Rights Act of 1965 built on this proven political theory by cutting at barriers to the opportunity to vote. The enfranchisement of protected minorities moved swiftly and the opportunity to vote quickly became a reality. Nonetheless many local political structures facilitated the perpetuation of old power, frustrating the leveraging power of voting minorities. Many white majorities were unable to jump traces that in combination with various election arrangements, such as at-large voting, facilitated the freezing of factions and left the white majority unchallenged and unresponsive. The fight to remove those structures that diluted the hard won votes of minorities encountered in 1980 the requirement that the structures be shown to have been adopted or perpetuated with the purpose of denying voting strength to minorities. The fight returned to the Congress resulting in the 1982 amendments to the Act, the interpretation of which today confounds this court. Before turning to these difficulties, one distraction must be laid aside.

I challenge the assertion that this court is somehow less devoted than in times past to the implementation of this congressional directive and the constitutional commands it implements. That assertion rests on a failure to appreciate the new claims now being pushed under the old clothing of the civil rights movement. The assertion exposes unawareness that this great surge in the effort to implement the political ideal that we are all the children of this political order has, perhaps inevitably, gathered the baggage of contentions that federal courts must not only strip away the effects of racial and ethnic discrimination but must now shelter the work of political parties and political coalitions that may today but not tomorrow offer aid to blacks and Hispanics. It is here that I part company with my colleagues, a parting that leaves us far apart. The 1982 amendments reached to effects of racial discrimination; it did not and could not have dispensed with proof of race as a causative factor. As I will explain, the majority has read the Voting Rights Act as the protector of factions, not of protected racial minorities. It is wrong. By ignoring race the majority loses the moral force of the Civil Rights Amendments. More to the point it sorely taxes the constitutionality of the Act as the majority would apply it.

This insistence upon a tie to race received expression and effect in Whitcomb v. Chavis, 403 U.S. 124, 91 S. Ct. 1858, 29 L. Ed. 2d 363 (1971), and White v. Regester, 412 U.S. 755, 93 S. Ct. 2332, 37 L. Ed. 2d 314 (1973), as well as the decisions of this court. See, e.g., Zimmer v. McKeithen, 485 F.2d 1297 (1973), aff'd sub nom East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S. Ct. 1083, 47 L. Ed. 2d 296 (1976). The 1982 amendment's codification of the "results test" set out in Whitcomb and White ensured that the focus of § 2 would remain on those cases "where a combination of public activity and private discrimination have joined to make it virtually impossible for minorities to play a meaningful role in the electoral process." Hearings on the Voting Rights Act Before the Subcom. on the Constitution of the Senate Comm. of the Judiciary, 97th Cong., 2d Sess. 1367-68 (statement of Professor Drew Days) (emphasis added). The Supreme Court confirmed this understanding of the Act in Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752, 92 L. Ed. 2d 25 (1986). There, Justice Brennan's attempt to sever the crucial link between alleged unequal opportunity and race was rebuffed by a majority of the Justices as a standard crafted to shield political minorities from the vicissitudes of "interest-group politics rather than a rule hedging against racial discrimination." Gingles, 478 U.S. at 83, 106 S. Ct. at 2782 (White, J., concurring); id. at 101, 101 S. Ct. at 2792 (O'Connor, J., joined by Burger, C.J., Powell and Rehnquist, JJ., concurring).

Similarly, the effects test that lies at the heart of statutory anti-discrimination law has always contemplated that defendants may attempt to rebut a prima facie case of disparate impact by showing that the challenged practice promotes a substantial interest. This "balancing test," despite its indeterminacy, has been an integral part of Title VII cases and was expressly recognized in the context of § 2 in Houston Lawyers' Ass'n v. Attorney General, --- U.S. ----, 111 S. Ct. 2376, 115 L. Ed. 2d 379 (1991), where the Supreme Court held that the state's interest in maintaining its electoral scheme should be weighed against proven vote dilution. Id. --- U.S. at ----, 111 S. Ct. at 2381. It is a direct draw upon equal protection constructs.

The district court did not apply these legal standards. To the contrary, the court held that the causes of divergent voting patterns, including the extraordinary explanatory power of partisan voting, were legally irrelevant. Not even the majority seriously defends the ruling of the district court that partisan voting is simply irrelevant. In Dallas County, for example, the Republican Party dominates judicial elections--black Republicans win and white Democrats lose. There are virtually no exceptions save the most recent elections in which a black Democratic appointee defeated a white Republican. But this bear in the plaintiff tent is to be ignored, I am told. This dependence upon a winning Republican Party for proof of bloc voting is an embarrassment to the findings of voting dilution in this case, unless it's the case that racial prejudice cements white majorities only in these nine counties and not in the Democratic strongholds of Texas. This alone highlights the failure of the plaintiffs' model to distinguish between bloc and party voting. Plaintiffs' and the majority's model produces a second perverse result. By the model as you proceed back in time there is less dilution of minority votes because with each step backward into Texas history the Democratic Party became stronger; the end of the closed Texas primaries would mark the absence of bloc voting.

All this comes from a majority willing to find a racial appeal in a Dallas County judicial race sufficient to warrant changing the structure of all its courts from a report that state district court Judge Baraka was described as a Black Muslim. The majority fails to point out that Judge Baraka was elected with 61% of the white vote--and now serves on that court. In sum, the majority does not fairly deal with the facts and its view of partisan elections is not supported by the law. It equally defies common sense.

The district court also refused to consider the strong state interests advanced by Texas' method of electing its district judges. Unable to explain away the error in failing to apply governing law, the majority nonetheless, with one unsupportable exception, affirms the district court's holding in its entirety. It might fairly be asked how reversal or at least a remand might be avoided once an appellate court concludes that a district court's holding is rooted in a fundamental misapprehension of not one, but two foundational elements of § 2 jurisprudence. The majority accomplishes this surprising feat by nominally accepting the relevant Supreme Court decisions, but then misinterpreting and misapplying them so as to give this controlling authority no practical effect.

The majority's interpretation of cases directing courts to inquire into the causes of voting patterns renders these decisions, particularly Whitcomb, beyond recognition. Freed from the constraints of statute and precedent, Congress' powerful and distinct command forbidding any practice "which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color" is transformed in the imaginative hands of the majority into a "test" that establishes a violation of § 2 whenever "some mix of factors" discloses an undefined measure of unequal opportunity in a way that is somehow "linked to race or color." Majority op. at 744. The majority's avoidance of Houston Lawyers' is less inventive but no less effective. The court concedes that the state's interest must be weighed against vote dilution, but then dismisses Texas' interest in maintaining its electoral plan, one virtually as old as statehood, and one that several members of this court, upon our earlier review, found so weighty as to preclude § 2 liability as a matter of law, as no more than "rational." We are bound to give Supreme Court decisions and the will of Congress their full effect. I dissent from today's refusal to do so.

II

The majority properly notes that a § 2 attack on an at-large system is governed by the framework established in Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752, 92 L. Ed. 2d 25 (1986). Under Gingles, a challenger of an at-large election scheme must demonstrate that (1) it is sufficiently large and geographically compact to constitute a majority in a single-member district; (2) it is politically cohesive; and (3) the white majority votes sufficiently as a bloc to enable it usually to defeat the minority's preferred candidate. 478 U.S. at 48-51, 106 S. Ct. at 2765-67. Satisfaction of these three threshold factors is necessary but not sufficient to establish liability; plaintiffs must also show that, under the "totality of circumstances," they do not possess the same opportunities to participate in the political process enjoyed by other voters. The heart of my disagreement with the majority lies in its explication and application of the Gingles inquiry into white bloc voting and the closely related Zimmer factor directing courts to examine "the extent to which voting ... is racially polarized." S.Rep. 417 at 29, reprinted in 1982 U.S.Code Cong. & Admin.News at 206. According to the majority, causation has no role in examining divergent voting patterns; bloc voting exists whenever "whites and blacks vote differently." This approach is overly simplistic and in stark tension with § 2's express command that minority groups shall not be denied equal opportunity "on account of race or color." Contrary to the majority's contentions, Whitcomb, the text and legislative history of the 1982 amendments, Gingles, and the decisions of this circuit preclude a finding of racial bloc voting where divergent voting patterns are best explained by partisan affiliation.

In Whitcomb, black citizens residing in one part of Marion County, referred to as the "ghetto" by the Court, claimed that the county's at-large method of electing members to the state legislature impermissibly diluted their votes. "Ghetto" residents "voted heavily Democratic," but since the county's white majority consistently voted Republican, their preferred candidates were defeated in four of the five elections between 1960 and 1968. 403 U.S. at 150, 91 S. Ct. at 1873. The Whitcomb Court recognized that the at-large electoral scheme caused the "voting power of ghetto residents [to be] 'cancelled out,' " id. at 153, 91 S. Ct. at 1874, but stated that this result by itself did not provide grounds for relief. Noting that blacks enjoyed full access to the political process, id. at 149-50, 91 S. Ct. at 1872-73, the Court reasoned that "had the Democrats won all of the elections or even most of them, the ghetto would have no justifiable complaints about representation." Id. at 152, 91 S. Ct. at 1873. For this reason, the Court concluded that the "failure of the ghetto to have legislative seats in proportion to its population emerges more as a function of losing elections than of built-in bias against poor Negroes." Id. at 153, 91 S. Ct. at 1874.

The Whitcomb Court was reluctant to view the plaintiffs' claims of vote dilution as anything more than "a euphemism for political defeat at the polls," id., for, absent evidence of a lack of access to the political system, there was no principle by which the Court could distinguish the "ghetto's" claims and those of other unsuccessful political groups:

[A]re poor Negroes of the ghetto any more underrepresented than poor ghetto whites who also voted Democratic and lost, or any more discriminated against than other interest groups or voters in Marion County with allegiance to the Democratic Party, or, conversely, any less represented than Republican areas or voters in years of Republican defeat? We think not. The mere fact that one interest group or another concerned with the outcome of Marion County elections has found itself outvoted and without legislative seats of its own provides no basis for invoking constitutional remedies where, as here, there is no indication that this segment of the population is being denied access to the political system.

Id. at 154-55, 91 S. Ct. at 1874-75. To grant relief to black residents in this case "would make it difficult to reject claims of Democrats, Republicans, or members of any political organization in Marion County who live in what would be safe districts in a single-member district system but who in one year or another, or year after year, are submerged in a multi-member district vote." Id. at 156, 91 S. Ct. at 1876.

As the Seventh Circuit recently observed, § 2 is "a balm for racial minorities, not political ones--even though the two often coincide." Baird v. Consolidated City of Indianapolis, 976 F.2d 357, 361 (7th Cir.1992) (citing Whitcomb ). Black Democrats are entitled to relief when their defeats at the polls are attributable to their race, but not when such losses are due to their party. The majority rejects the rule established in Whitcomb and its opinion is replete with assertions that any suggestion that evidence of partisan voting might be introduced to explain divergent voting patterns is presumptively inconsistent with § 2's focus on "results." See, e.g., Majority op. at 748; see also Richard L. Engstrom, The Reincarnation of the Intent Standard: Federal Judges and At-Large Election Cases, 28 How.L.J. 495, 498 (1985). The relevant authorities, however, indicate that Whitcomb 's causation analysis is entirely consistent with an approach that looks not to "intent," but "effects." In Mobile v. Bolden, 446 U.S. 55, 100 S. Ct. 1490, 64 L. Ed. 2d 47 (1980), for example, the plurality's assertions that Whitcomb had required a showing of intent, see id. at 66, 70, 100 S. Ct. at 1499, 1501, prompted Justice Marshall in dissent to explain the decision:

In Whitcomb v. Chavis, we again repeated and applied the Fortson [v. Dorsey, 379 U.S. 433, 85 S. Ct. 498, 13 L. Ed. 2d 401 (1965) ] [effects] standard, but determined that the Negro community's lack of success at the polls was the result of partisan politics, not racial vote dilution. The Court stressed that both the Democratic and Republican Parties had nominated Negroes and several had been elected. Negro candidates lost only when their entire party slate went down to defeat. In addition, the Court was impressed that there was no finding that officials had been unresponsive to Negro concerns.

Id. at 109, 100 S. Ct. at 1522 (Marshall, J., dissenting) (emphasis added) (citations omitted). The Senate Report agreed with Justice Marshall that Whitcomb had employed an effects test and echoed his emphasis in presenting what it understood to be the kernel of the decision:

The failure of the ghetto to have legislative seats in proportion to its population emerges more as a function of losing elections than of built-in bias against poor Negroes. The voting power of ghetto residents may have been "cancelled out," as the district court held, but this seems a mere euphemism for political defeat at the polls.

S.Rep. 417 at 21 (quoting Whitcomb, 403 U.S. at 153, 91 S.Ct. at 1874) (emphasis added), reprinted in 1982 U.S.Code Cong. & Admin.News at 198. These references to the "lack of success at the polls" as a "result" or "function" of "partisan politics," not "racial vote dilution" or "built-in bias against poor Negroes," only confirm Whitcomb 's central meaning: where black Democrats lose not because they are black, but because they are Democrats, or, put another way, where black Republicans win and white Democrats lose, there can be no finding of illegal vote dilution.

The majority offers a very different reading of Whitcomb. According to the majority, this case stands for a very narrow proposition: "Where there is evidence of partisan voting or interest group politics and no evidence that members of the minority group have an unequal opportunity to participate in the political process on account of race or color, the minority group's vote dilution claim will fail." Majority op. at 809 (emphasis in original). Stated in this manner, Whitcomb presents no obstacle to liability in this case, for here there is "substantial evidence that minority voters in those counties have an unequal opportunity to participate in the political process on account of race." Id. at 811.

This reading renders Whitcomb nonsensical. The majority would have the decision stand for the following proposition: "Where there is evidence of partisan voting and no evidence of vote dilution, the minority's vote dilution claim will fail." This odd result is accomplished through a bit of anachronistic wordplay, as the majority infuses one of the Court's subsidiary findings with the import it would have under the amended § 2. The majority quotes the Whitcomb Court's statement that black voters could not prevail " 'absent evidence and findings that [black] residents had less opportunity than did other Marion County residents to participate in the political process and elect legislators of their choice.' " Majority op. at 808 (quoting Whitcomb, 403 U.S. at 149, 91 S.Ct. at 1872). The language used by the Court here was of course codified by the 1982 amendments, and today constitutes the ultimate showing of vote dilution necessary to make out a violation of § 2. The Whitcomb Court had something very different in mind, however, as the text immediately following this statement discloses:

We have discovered nothing in the record or in the court's findings indicating that poor Negroes were not allowed to register or vote, to choose the political party they desired to support, to participate in its affairs or to be equally represented on those occasions when legislative candidates were chosen. Nor did the evidence purport to show or the court find that inhabitants of the ghetto were regularly excluded from the slates of both major parties, thus denying them the chance of occupying legislative seats.

Id. at 149-50, 91 S. Ct. at 1872-73. This evidence of an equal "opportunity ... to participate in the political process and elect legislators of their choice" pertains solely to formal barriers to political participation, and is examined under "Zimmer" factors today. As such, it is, contrary to the majority's contentions, quite remote from anything resembling an ultimate finding of nondilution.

The flaws in the majority's reading of Whitcomb become even more pronounced when it attempts to distinguish this decision from the instant case. Recall that Whitcomb "held" that a minority group's vote dilution claim will fail "[w]here there is evidence of partisan voting or interest group politics and no evidence that members of the minority group have an unequal opportunity to participate in the political process on account of race or color." Majority op. at 809 (emphasis in original). The majority concludes from this that "Whitcomb does not preclude a finding of vote dilution" in this case because there is "substantial evidence that minority voters in those counties have an unequal opportunity to participate in the political process on account of their race or color." Majority op. at 811. In these counties

there is evidence of a history of discrimination against minorities that violated their rights to participate in the political process, evidence of strong to severe racially polarized voting, evidence that minorities continue to bear the effects of past discrimination, and evidence that minority candidates, especially those who are sponsored by the minority community, are unsuccessful in their bids to hold offices like the district court bench.

Id. at 811. The difficulty is that all of these circumstances constituting "substantial evidence" of unequal opportunity were present to an equal or greater extent in Whitcomb, a case in which, according to the majority, there was "no evidence." A comparison of the two cases both confirms the contrived nature of the majority's interpretation of Whitcomb and discloses the utter meaninglessness of the court's "findings."

First, with respect to the effects of past discrimination, the Whitcomb Court did not dispute the district court's findings that "[s]trong differences" between "ghetto" residents and the adjacent communities "were found in terms of housing conditions, income and educational levels, rates of unemployment, juvenile crime, and welfare assistance." 403 U.S. at 132, 91 S. Ct. at 1863. Moreover, in contrast to the district court's "inquiry" into the effects of past discrimination in this case, which even the majority concedes amounted to no more than the taking of "judicial notice," majority op. at 772, the lower court findings before the Whitcomb Court were the product of careful and detailed analysis. See Chavis v. Whitcomb, 305 F. Supp. 1364, 1376-81 (S.D.Ind.1969). Second, the presence of "polarized voting" does not distinguish this case, for the evidence in Whitcomb indicated that the "ghetto" voted overwhelmingly Democratic while white voters consistently supported Republicans. See Whitcomb, 403 U.S. at 149-53, 91 S. Ct. at 1872-74.1 Finally, the record in Whitcomb discloses that the "ghetto" was less successful than the minority groups in several of the counties here in electing minority candidates. "Ghetto" residents totaled nearly 18% of the Marion County population and were elected to 4.75% of the Senate seats and 5.97% of the Assembly seats during the relevant time period. See Whitcomb, 403 U.S. at 164, 91 S. Ct. at 1880 (appendix). This level of success is comparable to minority candidates' election rate in several of the counties in which the majority finds a violation of § 2, such as Tarrant County (11.8% black population, 8.7-13.1% of judges), Bexar County (46.6% Hispanic population, 15.8-31.6% of judges), Dallas County (18.5% black population, 8.3% of judges), and Harris County (19.7% black population, 5.1% of judges). Here again, the proof of this factor is if anything more compelling in Whitcomb, for unlike judicial elections in which only lawyers are eligible, there are no demanding prerequisites for citizens wishing to run for state legislature.2

In sum, were facts identical to those present in Whitcomb before the court today, the majority's analysis would compel it to find, in direct contradiction with the Supreme Court's holding, illegal vote dilution. The obvious impermissibility of this result makes a mockery of its announcement, apparently offered in all seriousness, that "we decline to redefine the concept of racially polarized voting to include a causation element." Majority op. at 748. An inquiry into whether divergent voting patterns are attributable to partisan voting does not involve a "redefin[ition]" of a "concept," but the application of binding Supreme Court precedent. Further, that this court would find "substantial evidence" of unequal opportunity on the same record in which the Supreme Court, according to the majority, found "no evidence" demonstrates just how far we have strayed in examining the other elements of a voting rights claim. Whitcomb dictates that no dilution should be found where polarized voting is the product of partisan politics, even where a minority group suffers from the lingering effects of past discrimination and the percentage of minority officeholders does not mirror the group's population.

The majority strenuously resists both the general applicability of Whitcomb and the particular conclusion that an application of the approach it employs today to the facts of that case would require us to find illegal vote dilution. The court contends that "Whitcomb must be read in the context of the community in which the vote dilution claim was being brought." Majority op. at 810. The "context" of Whitcomb differs from the record before us, according to the majority, in two respects. The plaintiffs in Whitcomb, unlike the plaintiffs in this case, did not present evidence (1) of "a history of official discrimination touching the rights of Blacks to vote and participate in the political process" and (2) "that the socioeconomic disparities from which they suffered were the result of, or on account of, race or color." Majority op. at 810.

The majority's attempt to distinguish Whitcomb on these grounds is wholly unpersuasive. This argument does provide, however, an ample demonstration of the literally dispositive weight the court attaches to the two Zimmer factors pertaining to the effects of discrimination. In the first instance, the majority offers no reasons in support of its contention that Whitcomb must be read "in context." The Whitcomb Court itself certainly gave no indications that its analysis of vote dilution, particularly its emphasis on the role of partisan voting, was somehow tied to the immediate facts of the case or otherwise inapplicable to cases in which there is evidence of discrimination with respect to voting. In addition, there are no suggestions in either the Senate Report or Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752, 92 L. Ed. 2d 25 (1986), that Whitcomb should be limited in the manner proposed by the majority. Gingles is especially relevant, given that the Court had before it a § 2 claim arising in North Carolina, a state which of course shares much of Texas' tragic history of discrimination. To the extent that the court's limited reading of Whitcomb rests on the perceived importance of the effects of discrimination, the rejection of this interpretation by the relevant authorities discloses that the court has grossly overestimated the significance of these factors.

Moreover, even if the Zimmer factors concerning the effects of past discrimination could bear the dispositive weight the court imposes, the lack of evidence relating to these effects in this case precludes the court from distinguishing Whitcomb. Recall that Whitcomb differs from this case because of an absence of evidence (1) of "a history of official discrimination touching the rights of Blacks to vote and participate in the political process" and (2) "that the socioeconomic disparities from which they suffered were the result of, or on account of, race or color." Majority op. at 810. The second part of this statement is not accurate, for residents of the "ghetto," like many of the minority voters in this case, were victims of discrimination in education and housing, as a 1971 decision finding de jure segregation in the Indianapolis public schools indicates. See United States v. Board of School Commissioners, 332 F. Supp. 655 (S.D.Ind.1971). Thus, the two cases are distinguishable only by Texas' history of voting discrimination. The difficulty with the majority's reliance on this factor, however, is that there is no evidence in the record that this history has any present-day effects in seven of the eight counties in which the court finds § 2 liability. The majority concedes that plaintiffs have failed to show that black citizens register to vote at a lower rate than whites, the typical way in which this Zimmer factor is established, see, e.g., Gingles, 478 U.S. at 39, 106 S. Ct. at 2760, and instead purports to rely on the paucity of minority lawyers as proof. But this lack of minority lawyers is attributable, as the court's citation of Sweatt v. Painter, 339 U.S. 629, 70 S. Ct. 848, 94 L. Ed. 1114 (1950), suggests, to discrimination in education, not voting. Since Marion County and Texas have this unfortunate history of segregated education in common, the majority's attempt to distinguish Whitcomb ultimately fails.

As I explain in more detail in the attached opinion, Whitcomb precludes a finding of vote dilution in most, but not all, of the counties at issue. The only possible justification for the majority's failure to give the decision full force in this case is that Whitcomb enjoys lesser significance under the amended § 2. The majority in fact makes a brief attempt to slight the decision's importance, suggesting only that Whitcomb "continues to have relevance under the amended Section 2." Majority op. at 808. But the very passages in the Senate Report the court cites explicitly state that the 1982 amendment is intended to codify the results test as employed in Whitcomb and White. See S.Rep. 417 at 20-23, 32-33, reprinted in 1982 U.S.Code Cong. & Admin.News at 197-201, 210-11; see also Jones, 727 F.2d at 379 (the amended § 2 "codifies pre-Bolden voting dilution law"). A review of the text and legislative history of the 1982 amendment and Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752, 92 L. Ed. 2d 25 (1986), confirms this legislative intent.

Whitcomb 's teaching that illegal vote dilution does not lie where the inability of minority groups to elect candidates of their choice is the product of partisan voting--where black Republicans win and white Democrats lose--is reflected in the text of amended § 2 itself. Plaintiffs may establish a violation of § 2 by proving that members of a minority group "have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." 42 U.S.C. § 1973(b). The statute then singles out "[t]he extent to which members of a protected class have been elected to office in the state or political subdivision [as] one circumstance which may considered." Id. This express reference to the success of minority candidates led the Gingles Court to cite this factor along with racial bloc voting as most probative of illegal vote dilution. 478 U.S. at 48 n. 15, 106 S. Ct. at 2765 n. 15.

The influence of Whitcomb is evident in Congress' provision of "[t]he extent to which members of a protected class have been elected," rather than the extent to which members of a minority group have "elect[ed] representatives of their choice," as the yardstick for minority electoral success. The minority-preferred candidate will not always be a minority herself, and successful minority candidates will not always have the support of minority voters. The distinction between the two terms contained in § 2 loses its significance only in cases where race is the primary determinant of voter behavior, as the white majority's refusal to vote for minority candidates ensures they will be elected only with the support of minority voters, if at all. By contrast, where, as here, party affiliation best explains voting patterns, Congress' decision to confine the inquiry into electoral success to whether "members of a protected class have been elected" becomes quite relevant. In order to establish this crucial element in their vote dilution claim, plaintiffs must show not that the white majority does not support minority-preferred candidates, but that whites do not vote for minority candidates. Simply put, it is quite possible that minority plaintiffs will be unable to demonstrate a relevant lack of electoral success as part of the "totality of circumstances" inquiry even where few, if any, minority-preferred candidates have been elected, as in the case, for example, where black Republicans win and white Democrats lose. Section 2 thus preserves Whitcomb 's sharp distinction between cases of illegal vote dilution and those instances in which the "cancell[ing] out" of a minority group's "voting power" constitutes no more than a "political defeat at the polls." Whitcomb, 403 U.S. at 153, 91 S. Ct. at 1874.

This interpretation of § 2 of course presumes that Congress intended to convey legal meaning through the use of two distinct phrases, that it did not use "members of the protected class" and "representatives of [the minority group's] choice" synonymously on the assumption that whites will only support white candidates and minorities will only support minority candidates. In Gingles, however, Justice Brennan proposed this contrary reading of § 2 in contending, as the majority does here, see, e.g., majority op. at 743, that the focus of a court's inquiry into electoral success should remain on whether members of a minority group are able to "elect representatives of their choice." Gingles, 478 U.S. at 67-68, 106 S. Ct. at 2774-75 (emphasis in original). Justice Brennan maintained that "[u]nder § 2, it is the status of the candidate as the chosen representative of a particular racial group, not the race of the candidate, that is important." Id. (emphasis in original). The obvious difficulty with this position is that § 2's explicit reference to "members of a protected class" would seem to make the race of the candidate, not the source of her support, the significant factor in gauging electoral success. Justice Brennan attempted to ease this tension by intimating that "members of a protected class" should be read "representatives of [the minority group's] choice." After noting that "it will frequently be the case that a black candidate is the choice of blacks, while a white candidates is the choice of whites," he explained that this "tendency" made it "convenien[t]" for him to follow the district court in "refer[ring] to the preferred representative of black voters as the 'black candidate' and to the preferred representative of white voters as the 'white representative.' " Id. Given that Congress no doubt had the opportunity to observe this same "tendency" among voters, is it unreasonable to believe that it employed the same background assumption in drafting § 2? Regardless, Justice Brennan concluded, contrary to the text of the statute, that "the fact that race of voter and race of candidate is often correlated is not directly pertinent to a § 2 inquiry." Id.

The Senate Report, however, expressly disclaimed any reliance on a purported correlation between the race of the voter and the candidate. The Senate Subcommittee on the Constitution's Report charged the authors of the amended statute with proceeding under the same assumption attributed to them by Justice Brennan, "that race is the predominant determinant of political preference ... [and] that blacks will only vote for black candidates and whites only for white candidates." S.Rep. 417 at 107, 148 (Report of the Subcommittee on the Constitution), reprinted in 1982 U.S.Code Cong. & Admin.News at 278, 321. Rather than disclose its adoption of this assumption or dismiss the asserted tendency among white voters to vote only for white candidates as "not directly pertinent to a § 2 inquiry," as Justice Brennan's analysis might have suggested, the Senate Report cited such incidents of "racial politics" as the harm § 2 was intended to remedy: "Unfortunately ... there still are some communities in our Nation where racial politics do dominate the electoral process. In the context of such racial bloc voting, and other factors, a particular election method can deny minority voters equal opportunity to participate meaningfully in elections." Id. at 33, reprinted in U.S.Code Cong. & Admin.News at 211. The Senate Report continued:

The results test makes no assumptions one way or the other about the role of racial political considerations in a particular community. If plaintiffs assert that they are denied fair access to the political process, in part, because of the racial bloc voting context within which the challenged election system works, they would have to prove it.

Id. at 34 (emphasis in original), reprinted in U.S.Code Cong. & Admin.News at 212.

Far from assuming a tendency among white voters to vote for white candidates in drafting the statute, Congress identified such "racial political considerations" as the linchpin of a successful vote dilution claim. We have no license to amend § 2, as the majority would do here, by analyzing the electoral success of "members of a protected class" through a "representatives of [the minority group's] choice" filter, with the success of minority candidates not supported by the minority group being discounted accordingly. Evidence of the electoral success of minority candidates should not, of course, be accepted uncritically, but the Senate Report, consistent with the text of § 2, expressly limits the scope of our inquiry in cautioning that the

election of a few minority candidates does not "necessarily foreclose the possibility of dilution of the black vote," in violation of this section. Zimmer, 485 F.2d at 1307. If it did, the possibility exists that the majority citizens might evade the section e.g., by manipulating the election of a "safe" minority candidate. "Were we to hold that a minority candidate's success at the polls is conclusive proof of a minority group's access to the political process, we would merely be inviting attempts to circumvent the Constitution ... Instead we shall continue to require an independent consideration of the record." Ibid.

Id. at 29 n. 115, reprinted in 1982 U.S.Code Cong. & Admin.News at 207 n. 115. This passage does not suggest that courts should depreciate the success of minority candidates elected without widespread support from minority groups, as the majority might have it. The concern that the election of a " 'safe' minority candidate" might facilitate an evasion of § 2 is a function of the number of candidates, not the source of their support. The election of several, rather than few, minorities precludes an inference that the process has somehow been manipulated. In any event, the record in this case discloses that Republicans have in no way attempted to limit their ticket to minority candidates adjudged to be "safe." Plaintiff-intervenors Oliver, Tinsley, and White, for example, all of whom ran unsuccessfully as Democrats, testified that they were heavily lobbied by the Dallas Republican Party to run as Republicans. White in fact acknowledged what the majority has refused to recognize in this case, that "if I ran as a Republican ... the likelihood is that I would win."

Section 2 provides that plaintiffs' proof of a lack of electoral success turns solely on the election rate of "members of a protected class" without regard to whether these individuals are "representatives of [the minority group's] choice." Since the support of the majority of white voters will ensure victory in most communities, the statute, in so limiting our inquiry, requires plaintiffs seeking to demonstrate a lack of electoral success to establish that white voters do not vote for minority candidates. This principle seems sensible, for the willingness of whites to support minority candidates tends to negate any inference that their failure to give their votes to minority-preferred candidates, regardless of color, is attributable to race. As such, the losses of minority-preferred candidates in these instances are best explained by political, rather than racial, differences.3

It would seem anomalous for our inquiry into racial bloc voting, the other crucial element of a vote dilution claim identified by the Gingles Court, to be governed by different principles. Plaintiffs may establish this factor by demonstrating "that the white majority votes sufficiently as a bloc to enable it ... usually to defeat the minority's preferred candidate." Gingles, 478 U.S. at 51, 106 S. Ct. at 2766-67. As discussed above, § 2's provision of the race of the candidates, not their status as the choice of the minority group, as the relevant factor in measuring electoral success will often preclude plaintiffs from establishing legally relevant electoral losses where these defeats are attributable to partisan differences. A consideration of the race of the candidate in examining racial bloc voting leads to a similar and, I would submit, proper result.

The role of the candidate's race in assessing racial bloc voting formed the basis of the fundamental disagreement among the Justices in Gingles. Justice Brennan advanced the textual argument discussed above as part of his claim that "[u]nder § 2, it is the status of the candidate as the chosen representative of a particular racial group, not the race of the candidate, that is important." Id. at 68, 106 S. Ct. at 2775 (emphasis in original). Justice Brennan's assertion that the race of the candidate was irrelevant in examining racial bloc voting, however, was squarely rejected by five Justices in Gingles. 478 U.S. at 83, 106 S. Ct. at 2783 (White, J., concurring); id. at 100-01, 106 S. Ct. at 2792-93 (O'Connor, J., joined by Burger, C.J., Powell and Rehnquist, JJ., concurring). Justice White argued that

Justice Brennan states in Part III-C that the crucial factor in identifying polarized voting is the race of the voter and that the race of the candidate is irrelevant. Under this test, there is polarized voting if the majority of white voters vote for different candidates than the majority of the blacks, regardless of the race of the candidates. I do not agree. Suppose an eight-member multimember district that is 60% white and 40% black, the blacks being geographically located so that two safe black single-member districts could be drawn. Suppose further that there are six white and two black Democrats running against six white and two black Republicans. Under Justice Brennan's test, there would be polarized voting and a likely § 2 violation if all the Republicans, including the two blacks, are elected, and 80% of the blacks in the predominately black areas vote Democratic.... This is interest-group politics rather than a rule hedging against racial discrimination. I doubt that this is what Congress had in mind in amending § 2 as it did, and it seems quite at odds with the discussion in Whitcomb v. Chavis, 403 U.S. 124, 149-160, 91 S. Ct. 1858, 1872-78, 29 L. Ed. 2d 363 (1971).

Id. 478 U.S. at 83, 106 S. Ct. at 2783 (White, J., concurring) (emphasis added). Justice O'Connor joined Justice White in maintaining that evidence that white and minority voters generally supported different candidates did not constitute legally significant racial bloc voting where these patterns were attributable to partisan affiliation rather than the race of the candidate. She therefore rejected Justice Brennan's position thatevidence that the divergent racial voting patterns may be explained in part by causes other than race, such as an underlying divergence in the interests of minority and white voters.... can never affect the overall vote dilution inquiry. Evidence that a candidate preferred by the minority group in a particular election was rejected by white voters for reasons other than those which made that candidate the preferred choice of the minority group would seem clearly relevant in answering the question whether bloc voting by white voters will consistently defeat minority candidates. Such evidence would suggest that another candidate, equally preferred by the minority group, might be able to attract greater white support in future elections.

I believe Congress also intended that explanations of the reasons why white voters rejected minority candidates would be probative of the likelihood that candidates elected without decisive minority support would be willing to take the minority's interests into account. In a community that is polarized along racial lines, racial hostility may bar these and other indirect avenues of political influence to a much greater extent than in a community where racial animosity is absent although the interests of racial groups diverge. Indeed, the Senate Report clearly stated that one factor that could have probative value in § 2 cases was "whether there is a significant lack of responsiveness on the part of elected officials to the particularized needs of the members of the minority group." S.Rep., at 29. The overall vote dilution inquiry neither requires nor permits an arbitrary rule against consideration of all evidence concerning voting preferences other than statistical evidence of racial voting patterns. Such a rule would give no effect whatever to the Senate Report's repeated emphasis on "intensive racial politics," on "racial political considerations," and on whether "racial politics ... dominate the electoral process" as one aspect of the "racial bloc voting" that Congress deemed relevant to showing a § 2 violation. Id., at 33-34. Similarly, I agree with Justice White that Justice Brennan's conclusion that the race of the candidate is always irrelevant in identifying racially polarized voting conflicts with Whitcomb and is not necessary to the disposition of this case. Ante, [478 U.S.] at 83 [106 S. Ct. at 2783] (concurring).

Id. at 100-01, 106 S. Ct. at 2792 (O'Connor, J., concurring) (emphasis added).

The majority here nonetheless adheres to Justice Brennan's position that causation is irrelevant in determining whether racial bloc voting exists. Evidence of racial considerations is acknowledged only as an "other" factor to be taken into account under the totality of circumstances: "After all, as Justice O'Connor recognized in her Gingles concurrence, the Senate Report accompanying the amendment to § 2 repeatedly emphasizes the relevance of 'racial politics' and 'racial political considerations' to the vote dilution inquiry." Majority op. at 753 (citing Gingles, 478 U.S. at 101, 106 S.Ct. at 2792-93). The court does not quote Justice O'Connor directly, and in the very passage cited she asserts that a rule precluding consideration of "all evidence of voting patterns ... would give no effect whatever to the Senate Report's repeated emphasis on 'intensive racial politics,' on 'racial political considerations,' and on whether 'racial politics ... dominate the electoral process' as one aspect of the 'racial bloc voting' that Congress deemed relevant to showing a § 2 violation." Gingles, at 101, 106 S. Ct. at 2792 (quoting S.Rep. at 33-34) (emphasis added); see also id. at 83, 106 S. Ct. at 2783 (White, J., concurring) ("Under [Justice Brennan's] test, there is polarized voting if the majority of white voters vote for different candidates than the majority of the blacks, regardless of the race of the candidates. I do not agree.") (emphasis added). Far from being an "other" factor of which the Senate made no mention, the extent to which voting patterns are attributable to causes other than race is an integral part of the inquiry into racial bloc voting, as five Justices in Gingles expressly held. The majority's failure to treat it as such is unjustified.4

As the Seventh Circuit recently noted in Baird v. Consolidated City of Indianapolis, 976 F.2d 357 (7th Cir.1992), the agreement between Justice O'Connor and Justice White that a "system leading to the election of Black Republicans could not be dismissed as discriminatory" confirms the teaching of Whitcomb: "The Voting Rights Act does not guarantee that nominees of the Democratic Party will be elected, even if black voters are likely to favor that party's candidates." Id. at 361. In most areas where voting patterns are best explained by partisan affiliation, evidence that a majority of whites support one party will necessarily lead to the election of several minority candidates. (The contrary result might give rise to an inference that minorities did not enjoy full access to the party's slating process). Thus, Justice White and Justice O'Connor specifically charged that it was Justice Brennan's refusal to consider the race of the candidates that was inconsistent with Whitcomb. See Gingles, 478 U.S. at 101, 106 S. Ct. at 2792-93 (O'Connor, J., concurring) ("I agree with Justice White that Justice Brennan's conclusion that the race of the candidate is always irrelevant in identifying racially polarized voting conflicts with Whitcomb ").

This close tie between a consideration of the race of the candidates and an inquiry into the causes of a minority-preferred candidate's defeat poses additional problems for the majority's rejection of causation, for even the majority must acknowledge that this Circuit has sided with Justice White and Justice O'Connor against Justice Brennan in regarding those elections pitting minority candidates against white candidates as the most probative of bloc voting. See, e.g., Citizens for a Better Gretna v. City of Gretna, 834 F.2d 496, 503-04 (5th Cir.1987). This leaves the majority in the untenable position of simultaneously maintaining that the race of the candidate is relevant but evidence of partisan voting is not. As Whitcomb and the various opinions in Gingles demonstrate, a consideration of the race of the candidate is necessarily related to examination of the reasons for white voters' reluctance to support the minority-preferred candidate. Thus, in holding that the races in which a member of a minority group runs against a white candidate are most probative of racial bloc voting, the Gretna court cited the same "concerns about the dangers in advancing interest group politics or enforcing proportional representation," 834 F.2d at 503 (footnote omitted), identified by Justice O'Connor and Justice White as reasons for examining causation. The majority's sole response to this analysis is the assertion that "we have never held or suggested ... that for white bloc voting to be significant, it must be motivated by racial animus." Majority op. at 747 (citing Westwego Citizens for Better Gov't v. City of Westwego, 946 F.2d 1109, 1118-1120 (5th Cir.1991)). This single citation offers no support for this proposition, as we expressly noted in Westwego that "no non-racial explanation for the correlation between race and the selection of candidates has been offered." Id. at 1119 n. 14.

In sum, the majority's failure to consider partisan affiliation as an explanation of divergent voting patterns is in flagrant contradiction with § 2 and the relevant Supreme Court and Fifth Circuit precedents. These authorities have established § 2 as "a balm for racial minorities, not political ones--even though the two often coincide." Baird v. Consolidated City of Indianapolis, 976 F.2d 357, 361 (7th Cir.1992) (citing Whitcomb ). The majority appears to challenge this distinction between race and party, suggesting in various places that the Republican and Democratic Parties are mere proxies for racial and ethnic groups in Texas. In this view, the difference between supporting candidates on the basis of partisan affiliation rather than race is insignificant, for a Republican vote is a vote for whites and against minorities. I agree with the majority that courts should remain sensitive to this question, and should not be quick to dismiss plaintiffs' vote dilution claims in cases where divergent voting patterns correlate with partisan affiliation as "political defeats" not cognizable under § 2.

Here, the majority in effect treats party as a proxy for race by refusing to acknowledge the distinction between "partisan politics" and "racial vote dilution" that lies at the heart of § 2. My refusal to join the majority in equating race and party on the record before us is rooted in two considerations. First, white voters constitute the majority of not only the Republican Party, but also the Democratic Party, in several of the counties at issue. Even in those areas where the Republican Party is strongest, in Dallas County, for example, 30-40% of white voters consistently support Democrats, making white Democrats more numerous than all of the minority Democratic voters combined. These white Democrats have in recent years experienced the same electoral defeats as minority voters. If we are to hold, as the majority in effect does today, that these losses at the polls, without more, give rise to a racial vote dilution claim warranting special relief for minority voters, a principle by which we might justify withholding similar relief from white Democrats is not readily apparent.

I am also reluctant to treat party as a proxy for race in this case because a white partisan voter found herself not infrequently supporting a minority candidate rather than his white opponent. The undisputed evidence discloses that white voters in most counties, both Republican and Democratic, without fail supported the minority candidates on their party's ticket at levels equal to or greater than those enjoyed by white candidates, even where the minority candidate was opposed by a white candidate. In Dallas County, for example, Judge Wright received the greatest recorded percentage of the white vote (77%) in her race against a white Democrat. To conclude on this record that political parties serve as proxies for race is simply unwarranted. Because the evidence in most instances unmistakably shows that divergent voting patterns among white and minority voters are best explained by partisan affiliation, I would hold that plaintiffs have failed to establish racial bloc voting in most, but not all, of the counties.

On the other hand, where the reluctance of white voters to support minority-preferred candidates cannot readily be attributed to partisan affiliation, I would hold that plaintiffs have established an inference, one to be tested by a consideration of the Zimmer factors, that race is at work. The majority argues that a rule providing that racial bloc voting exists only where divergent voting patterns are not caused by partisan affiliation imposes a requirement on plaintiffs that they prove that white voters' rejection of minority-preferred candidates is due to "racial animus." In one sense this is true, in that evidence of partisan voting is probative under § 2 as well as Whitcomb and Gingles precisely because it tends to show that voters are not animated by racial considerations in selecting candidates.

In another sense, however, the majority's interpretation is incorrect. There are many other possible non-racial causes of voter behavior beyond partisan affiliation. A rule conditioning relief under § 2 upon proof of the existence of racial animus in the electorate would require plaintiffs to establish the absence of not only partisan voting, but also any other potential innocent explanation for white voters' rejection of minority candidates. I think that a rule requiring proof that the failure of white voters to support a particular minority candidate was not attributable, for example, to her inexperience or her reputation as a "bad" judge, see the discussion of the court's treatment of Travis County, infra, would impose far too great a burden on plaintiffs. Such a requirement would make racial bloc voting both difficult and, considering the additional analysis that would be needed, expensive to establish. Moreover, it would facilitate the use of thinly-veiled proxies by permitting, for example, testimony that a Hispanic candidate was a "bad" judge to defeat a claim that white voters' refusal to support him was based on race or ethnicity. Finally, an inquiry into causation beyond partisan affiliation seems inconsistent with the fundamental division between "partisan politics" and "racial vote dilution" under § 2. For these reasons, I would limit our inquiry into racial bloc voting to determining whether divergent voting patterns are caused by partisan differences, with the appropriate inference being drawn from this finding.

The importance of the distinction in § 2 jurisprudence between illegal vote dilution and political defeat, between protecting racial minorities and fostering the work of political coalitions, has led some to question whether different racial or ethnic minority groups, usually blacks and Hispanics, should be permitted to form a single minority group within the meaning of the Voting Rights Act. League of United Latin American Citizens v. Midland Indep. School District, 812 F.2d 1494, 1505-07 (5th Cir.1987) (Higginbotham, J., dissenting); Katherine I. Butler & Richard Murray, Minority Vote Dilution Suits and the Problem of Two Minority Groups: Can a 'Rainbow Coalition' Claim the Protection of the Voting Rights Act?, 21 Pacific L.J. 619, 641-57 (1990). We have determined, however, that such suits are permissible where the evidence suggests that the two groups are politically cohesive, see, e.g., Midland I.S.D., 812 F.2d at 1500-02, and I accept the authority of these decisions. At the same time, I do not believe that plaintiffs should be able to avoid its full effects. The defendants claim that the district court erred in refusing to consider elections involving Hispanic and white candidates in Harris and Tarrant Counties, counties in which plaintiffs proceed on behalf of black voters only, but where the indisputable evidence showed that blacks and Hispanics were politically cohesive. I agree.

Blacks and Hispanics have joined forces for purposes of this suit in Midland, Lubbock, and Ector Counties. In these counties, I would agree that white-Hispanic elections are relevant to show legally significant white bloc voting, because the Hispanic candidate provides the combined Hispanic-black minority with a viable minority choice. But the plaintiffs contend that where they represent only black voters, white-Hispanic elections in which the Hispanic candidate received the support of black voters are irrelevant. A difference in litigation strategy cannot support this distinction. Cohesion is a fact, not a strategic card to be played at the caprice of a plaintiff. As we stated in Campos v. City of Baytown, 840 F.2d 1240 (5th Cir.1988), "if the statistical evidence is that Blacks and Hispanics together vote for the Black or Hispanic candidate, then cohesion is shown." Id. at 1245 (footnote omitted). If blacks and Hispanics vote cohesively, in effect making them a single minority group, then elections with a candidate from this single minority group are elections with a viable minority candidate. The plaintiffs cannot have it both ways.

Plaintiffs contend that there is evidence is the record suggesting that blacks and Hispanics were not politically cohesive in Harris and Tarrant Counties. They do not tell us to which evidence they refer, and understandably so. The record shows that blacks and Hispanics were more cohesive in Harris and Tarrant Counties than in Midland and Ector Counties, where the plaintiffs represent both blacks and Hispanics and the district court found cohesion.

In Harris County, Dr. Taebel studied 46 elections in which he determined the percentage of black and Hispanic votes cast for the minority/winning candidate. In 35 elections the black and Hispanic vote percentages varied by less than 10%. Similarly, the levels of black and Hispanic support for the same candidate were within ten percentage points in 13 of the 17 elections studied in Tarrant County. In Midland County, by contrast, the black and Hispanic voting percentages differed by less than 10% in only 4 of the 8 elections analyzed; in Ector County, this close correlation between the preferences of Hispanic and black voters was shown in just 2 of 10 elections. I do not dispute the district court's findings of cohesion in Midland, Ector, and Lubbock Counties, for in those counties a significant number of blacks and Hispanics usually voted for the same candidates. Gingles, 478 U.S. at 56, 106 S. Ct. at 2769. But these findings compel the conclusion that there is also black-Hispanic cohesion in Harris and Tarrant Counties. The district court thus clearly erred in ignoring elections involving Hispanic and white candidates in these counties.

The majority attempts to salvage the plaintiffs' argument by recasting it in terms of appellate procedure. According to the court, the district court's refusal to consider the results of elections pitting Hispanic candidates against whites "was not properly raised before the district court and therefore is not properly before us on appeal." Majority op. at 789. The court correctly notes that the defendants did not ask the district court to find that blacks and Hispanics were cohesive in Tarrant and Harris Counties, but that is not the nature of the claim they press here. Defendants do not challenge the lower court's failure to find that Hispanics and blacks were cohesive; rather, they argue that the district court committed legal error by excluding Hispanic-white elections in determining whether white bloc voting existed in Harris and Tarrant Counties. It is true that this claim is, as the majority suggests, "fact-based," in that the validity of the district court's exclusion of these elections turns on whether blacks and Hispanics are cohesive, but the defendants contend that the evidence of cohesiveness is indisputable, a claim that is amply supported by the record and not contested by the majority. In sum, the court's rejection of defendants' claim is wholly unpersuasive and its refusal to consider white-Hispanic elections in Harris and Tarrant Counties is unjustified.

Finally, the majority's holding of no § 2 violation in Travis County, while upholding the district court's finding of dilution in all others, is worthy of special note. This illustrates the difficulties present in the case, for the majority must modify its test in order to avoid liability. The majority concludes that plaintiffs have failed to establish legally significant white bloc voting, Gingles 's third factor. I cannot say with such confidence that plaintiffs did not satisfy the threshold test. Instead, I would hold that the plaintiffs failed to show dilution under the totality of the circumstances.

The majority departs from its previous analysis on three significant points. First, Travis County was one of only two counties where voter registration statistics could have been introduced, tending to prove that past discrimination hindered the ability of Hispanics to participate in the political process. The evidence showed that only 38.5% of eligible Hispanic voters were registered to vote, a much lower rate than in Bexar County, the only other instance in which evidence tending to prove this Zimmer factor was offered. The fact that the majority could find that Hispanics enjoy an equal opportunity to participate in the political process in spite of this evidence betrays the court's analysis of other counties, where the Zimmer factors pertaining to past discrimination often seem to have nearly dispositive weight in the absence of any evidence of depressed minority participation.

Second, the majority has upheld the district court's determination and plaintiffs' argument that black plaintiffs are entitled to limit their proof to those elections involving black candidates and may exclude those elections pitting a Hispanic candidate against a white candidate, even where these Hispanic candidates are preferred by an overwhelming majority of black voters. In Travis County, however, where plaintiffs represent only Hispanics, the significance of a loss by a Hispanic candidate is discounted because a black candidate, who was not the Hispanic-preferred candidate, received a bare majority of the white vote. This analysis goes well beyond the approach I have urged, which would include elections involving candidates from a different minority group only upon a showing that blacks and Hispanics vote cohesively. The election of a black candidate who is not the choice of Hispanics in the end seems relevant because it tends to demonstrate that the voting preferences of whites are not attributable to racial considerations, a showing that I believe quite probative but one that the majority, at least in the other eight counties at issue, finds insignificant.5

The majority's reliance on causation, intimated above, becomes unmistakable in its treatment of another election loss by a Hispanic judge. The court asserts that this defeat is explained by "testimony indicating that the reason Gallardo lost is because he was a 'bad judge.' " Majority op. at 802 (emphasis in original). As in its reliance on the election of a black candidate to explain the defeat of a Hispanic candidate preferred by Hispanics, the court's analysis goes far beyond the position it has previously discarded. As I have indicated, our analysis of causation should be limited largely to an inquiry into whether divergent voting patterns among minorities and whites are attributable to partisan voting. A demonstration that minority and white candidates receive roughly equal levels of support from their party creates the inference that a minority group's inability to elect representatives of its choice is not attributable to race. On the other hand, a showing that white voters are less inclined to give their votes to the minority candidates on their party's ticket creates an inference, one to be tested by an analysis of the Zimmer factors, that race is at work. I would be quite reluctant to rely, as the majority does here, on anecdotal testimony concerning the characteristics of a particular candidate to establish that white voters were not motivated by racial or ethnic considerations. There are several reasons for this reluctance, some of which have been addressed above, but the most important one is that the Supreme Court has authorized a consideration of partisan voting in Whitcomb and Gingles and we should not be quick to stray from this mark.

As I stated, I agree with the majority that plaintiffs have not proven that they do not have an equal opportunity to participate in the political process and elect representatives of their choice in Travis County. That the court must employ an approach incorporating much of the analysis it has previously rejected to reach what it understands to be the right result would seem to provide ample reason by itself to question whether the approach used to find illegal vote dilution in the other eight counties at issue is the proper test.

III

In addition to severing racial discrimination and the Voting Rights Act, the majority errs by failing to weigh correctly the state's interest in defining the judicial office of district judge. As this panel previously recognized, see LULAC I, 902 F.2d 293, 307-09 (5th Cir.1990), vacated, 902 F.2d 322 (5th Cir.1990) (en banc), important reasons exist to link the electoral bases of district courts to their primary jurisdiction. The Supreme Court directed that on remand this interest would be weighed in determining whether a violation of the Act has occurred.

The majority flees this command with an over-the-shoulder characterization of the Supreme Court's opinion as containing "isolated passages" "suggesting" that the state's interest will be weighed. Majority op. at 760. This flight fails. The Court focused on this linkage, and its language belies the asserted ambiguity:

[T]he State's interest in electing judges on a district-wide basis ... is a factor to be considered by the court in evaluating whether the evidence in a particular case supports a finding of a vote dilution violation in an election for a single-member office.

* * * * * *

[W]e believe that the State's interest in maintaining an electoral system--in this case, Texas' interest in maintaining the link between a district judge's jurisdiction and the area of residency of his or her voters--is a legitimate factor to be considered by courts among the "totality of circumstances" in determining whether a § 2 violation has occurred. A State's justification for its electoral system is a proper factor for the courts to assess in a racial vote dilution inquiry, and the Fifth Circuit has expressly approved the use of this particular factor in the balance of considerations. See Zimmer v. McKeithen, 485 F.2d 1297, 1305 (CA5 1973), aff'd sub nom. East Carroll Parish School Bd. v. Marshall, 424 U.S. 636, 96 S. Ct. 1083, 47 L. Ed. 2d 296 (1976). Because the State's interest in maintaining an at-large, district-wide electoral scheme for single-member offices is merely one factor to be considered in evaluating the "totality of circumstances," that interest does not automatically, and in every case, outweigh proof of racial vote dilution.

Houston Lawyers', --- U.S. at ----, 111 S. Ct. at 2380-81 (emphasis in original).

No other court has had difficulty in reading this language. All correctly understood the Court's decision and weighed the state's interest in an electoral scheme in the totality of the circumstances. See e.g., Nipper v. Chiles, 795 F. Supp. 1525 (M.D.Fla.1992); Magnolia Bar Ass'n, Inc. v. Lee, 793 F. Supp. 1386 (S.D.Miss.1992); Southern Christian Leadership Conf. of Ala. v. Evans, 785 F. Supp. 1469 (M.D.Ala.1992). Commentators have also recognized that Houston Lawyers' directs courts to consider the strength of the state's interest at the liability stage. "[T]he Court recognized that in balancing the many factors in the totality of the circumstances test, the state interest in districtwide judicial elections may, in some cases, outweigh proof of racial voter dilution." Mary T. Wickham, Note, Mapping the Morass: Application of Section 2 of the Voting Rights Act to Judicial Elections, 33 Wm. & Mary L.Rev. 1251, 1285 (1992).

The majority sees the weighing of the linkage interest in the liability determination as a radical reworking of the traditional balancing test--which could not have been intended. I read the Supreme Court to say what it means, and find nothing radical in that meaning. The amended Act, like court decisions before and after it, considers the totality of the circumstances. Significantly, the Court in Houston Lawyers' cited Zimmer at page 1305, where we noted the significance of a strong, as opposed to merely non-tenuous state policy:

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, [cite], would require a holding of no dilution.

485 F.2d at 1305 (citing Whitcomb v. Chavis, 403 U.S. 124, 91 S. Ct. 1858, 29 L. Ed. 2d 363 (1971)) (emphasis added). The Senate Report on the 1982 amendment to § 2 endorsed both Zimmer and Whitcomb as parts of the pre-Bolden precedential standards which Congress intended to reinstate. S.Rep. 417 at 20-21, 23, 27-29; reprinted in 1982 U.S.Code Cong. & Admin.News at 197-98, 200-01, 204-07.

The majority refuses to hold that we must weigh the state's interest, see majority op. at 743, and then proceeds to weigh it, finding it insubstantial. That effort is flawed, misapprehending both the role and importance of linking jurisdiction to electoral base. While a state's interest will not automatically preclude a § 2 violation, Texas' substantial interest in preserving linkage can, at the least, outweigh marginal evidence of dilution.

Linking primary jurisdiction to the electoral base advances Texas' vital interest in judicial fairness and independence.6 Each district judge remains accountable to all voters within his or her jurisdiction, preserving "the fact and appearance of independence and fairness [which] are so central to the judicial task." LULAC II, 914 F.2d at 646 (Higginbotham, J., concurring); see also John L. Hill, Jr., Taking Texas Judges Out of Politics: An Argument for Merit Election, 40 Baylor L.Rev. 339, 364 (1988) ("the very perception of impropriety and unfairness undermines the moral authority of the courts"). Each judge acts alone, rather than in a collegial body, distinguishing this office from most elective positions. Thus, Texas' interest in district-wide elections goes beyond its interest in choosing officials a certain way. This election scheme is an integral component of the structure and effectiveness of the state judiciary. See LULAC II, 914 F.2d at 649-51 (Higginbotham, J., concurring). One member of the majority has twice previously agreed with this conclusion. See id. at 634; LULAC I, 902 F.2d at 294.

The substantiality of similar linkage interests in Florida and Alabama has recently been acknowledged. See Nipper v. Chiles, 795 F. Supp. 1525, 1548 (M.D.Fla.1992) (holding Florida's legitimate interest in linkage weighed against violation); Southern Christian Leadership Conf. of Ala. v. Evans, 785 F. Supp. 1469, 1478 (M.D.Ala.1992) (holding Alabama's strong interest in linkage weighed against violation). At least twenty-six other states currently elect judges of their principal trial courts.7 Twenty-two of them, like Texas, Alabama, and Florida, employ district-wide, at-large elections. Of the remaining four, North Carolina allows every elector within a judge's jurisdiction to vote on that judge by holding statewide elections for district courts. Louisiana and Mississippi only recently abandoned at-large judicial elections in order to end prolonged litigation. Only in Cook County, Illinois, have judges with county-wide jurisdiction been historically elected from electoral districts smaller than the county--and those courts have hardly been exemplars of judicial propriety.8 On the whole, the overwhelming preservation of linkage in states which elect their judges strongly supports Texas' contention that linkage is integral to the judicial role and is not simply Texas' preferred method of election.

The Supreme Court on the same day it decided Houston Lawyers' observed that, in our federal system, "the States' power to define the qualifications of their officeholders [referring to judges] has force even as against the proscriptions of the Fourteenth Amendment." Gregory v. Ashcroft, --- U.S. ----, ----, 111 S. Ct. 2395, 2405, 115 L. Ed. 2d 410 (1991).9 Gregory described the state's interest in maintaining a capable judiciary as "legitimate, indeed compelling." Id. --- U.S. at ----, 111 S. Ct. at 2407. Likewise, Texas has at least a substantial interest in defining the structure and qualifications of its elective judiciary. It has done so by refusing throughout its entire history to allow the jurisdictional bases of its trial courts to fall short of its election boundaries. Texas as a distinct political entity is entitled to strive for all the judicial independence that an elected judiciary can deliver. If setting the age requirements for judges lies at "the heart of representative government," surely a state's explicit definition of the authoritative reach of its judicial offices is also a powerful interest. Id. --- U.S. at ----, 111 S. Ct. at 2402.

The suggestion that Texas abandoned this interest by permitting county residents to decide to create districts smaller than counties is false. The opposite is true. Texas gave voters the power to reconsider the interest. Not a single county in Texas has voted to break the link between jurisdiction and electoral base--254 distinct reaffirmations of policy. Even if a county were to subdivide, the state preserves the interest for all other counties. Cf. Mahan v. Howell, 410 U.S. 315, 93 S. Ct. 979, 35 L. Ed. 2d 320, modified, 411 U.S. 922, 93 S. Ct. 1475, 36 L. Ed. 2d 316 (1973). In Mahan, the Court recognized that although Virginia divided one county when reapportioning its state legislature, it retained its interest in preserving boundaries of all other political subdivisions. Id. 410 U.S. at 327, 93 S. Ct. at 986.

The majority mistakenly suggests that Texas' interest deserves little weight because it might be accommodated by some remedies. Dismissing linkage on the basis that it might be preserved by certain remedies simply fails to weigh that interest in determining whether a violation has occurred. The majority shunts consideration of the interest from the liability stage to the remedy stage, contrary to Houston Lawyers'. Plaintiffs have asserted that at-large election of district judges violates § 2. We must decide whether the state's substantial interest in that electoral scheme outweighs plaintiffs' proof of dilution.

In several contexts courts balance states' interests against undesirable results. In doing so, we test the validity of the asserted interest or goal by considering how it may be achieved. Here, Texas' interest is in linking judges' electoral bases to their jurisdiction. That goal may only be achieved by a district-wide election system. The majority's suggestions to the contrary are red herrings. Cumulative or limited voting are election mechanisms which preserve district-wide elections. Thus, they are not even remedies for the particular structural problem which the plaintiffs have chosen to attack.

Plaintiffs went to trial attempting to prove the Gingles elements. Those elements test whether, but for district-wide, at-large voting, minority preferred candidates would likely prevail. That is all the Gingles elements prove. They do not establish whether other features of an electoral scheme, such as anti-single shot rules or majority runoffs, cause dilution. Since plaintiffs tried this case on the basis of Gingles, they did not show that the absence of cumulative or limited voting within the at-large scheme diluted minority votes. The question is whether Texas' district-wide election of trial judges is illegal. To answer that question, we must determine the weight of its linkage interest without regard to purported remedies which preserve the challenged at-large scheme. Plaintiffs cannot attack at-large voting as a violation of § 2 and ignore the state's interest in preserving linkage--unique to trial judges--by insisting that they will embrace a remedy which preserves that scheme. This cannot be done if Houston Lawyers' holds that the state's interest must be considered in the liability phase--which is precisely what it held.

The majority revives the argument that Texas has little interest in linking jurisdiction and electoral base because non-resident litigants may appear before judges they did not elect. Nothing has changed about this claim since it was previously rejected. See LULAC II, 914 F.2d at 651 (Higginbotham, J., concurring). The majority's efforts to distinguish jurisdiction from venue prove little, if anything. Texas venue law, as amended in 1983, has been influenced by both Spanish and English principles. Besides protecting defendants from inconvenient forums, the rules strive to ensure that local matters are tried in local courts.10 In addition to conducting trials, district judges have responsibilities such as selecting a county's grand jurors. Tex.Code Crim.Proc.Ann. art. 19.01 (Vernon Supp.1993). Thus, despite the state-wide jurisdiction which the majority emphasizes, the district court remains essentially a territorial court focused upon the county or counties comprising its district.

Perhaps the majority understates the local nature of the district court by overlooking the area of Texas practice where it is most evident: criminal law. There domicile and hence convenience to the defendant have never been a consideration. Although the state constitution permits a district court to try any Texas felony, Ex Parte Watson, 601 S.W.2d 350, 352 (Tex.Crim.App.1980) (en banc), the Code of Criminal Procedure limits the reach of that jurisdiction. In criminal practice a charging instrument must establish that the district court's county is a proper venue. Tex.Code Crim.Proc.Ann. arts. 21.02(5), 21.21(5) (Vernon 1989). Since the propriety of venue goes to the authority of the court, it is "quasi-jurisdictional in nature." George E. Dix, Texas Charging Instrument Law: The 1985 Revisions and the Continuing Need for Reform, 38 Baylor L.Rev. 1, 71 (1986).11 But whether the area of practice is criminal, civil, or family law, the conclusion reached in the concurring opinion in LULAC II remains valid. "[T]he state recognized that elimination of [the] risk and appearance of bias was essential to the office it was creating by an elaborate set of rules controlling venue." 914 F.2d at 651 (Higginbotham, J., concurring). If a Code of Judicial Conduct were alone sufficient to dispel the public's concern with "home cooking" and bias, there would be less need for venue rules. Instead, the district-wide election of judges joins with those rules to assure fairness and independence in the district courts.

Conclusion

The aggressive play in this case with Texas's judicial offices was as predictable as it is unwise. Federal courts were asked to do a political and legislative task and this panel responded with self assurance enjoyed only by officials steeped in the mythology that Courts are the sole arbiters of constitutional right and custom and consensus about those rights carry no sway. The image of the noble federal judge carrying out his task of protecting people from themselves is heady stuff indeed. That a state has throughout its history thought it essential to preserve linkage is of no moment ... of no weight and it is cast aside in a flood of legalisms. In doing so the majority proceeds, as we often do, with the certitude that judges and lawyers are the anointed defenders of constitutional values and liberties. True, as far as it goes but we know and admit in quieter moments that constitutional values rest on the broader foundations of consensus, and the genetic-like grip of generation after generation of understanding. The point is made by a look at the constitutional histories of the former Soviet Union and many South American countries with their beautiful documents unsupported by a culture of commitment to liberty. As T.S. Eliot put it:

The general ethos of the people they have to govern ... determines the behavior of politicians.

Congress may read a new consensus and strike a new balance, as the representative branch locates a new sense. But when it fails to do so it cannot simply hand its task uncharged to the courts with little more than "here, you figure it out." Courts must accept such assignments with caution, asking for clearer statements as faithful agents probing for their charge. Much can be said, but I need not pause here because we need not reach so far. I ask only that we be a bit more modest and not so bent on announcing new truths while draped with the hypocrisy that it is duty's call that we answer. We ought not cast aside the interests of Texas as little more than an ephemeral election rule in collision with voting rights. Its weight is substantial and the found "violations" are here more remote from the Voting Rights Act or other normative voting rule than any we have encountered.

Congress was on sound ground in explicating the Civil Rights Amendments, but it failed to grip the difficult choices in resolving the political choices confronting it in 1982. Instead of taking hold Congress lifted itself from the maelstrom in which it found itself by lifting the debate to a level of generality about election rights that sapped statutory clarity. The 1982 amendments to the Voting Rights Act gave the courts little help in defining the dilution it forbids. One thing is sure, Congress did not, because it could not, sever race from the voting rights inquiry. The majority in every practical way today does precisely that.

We are enjoined to make pragmatic assessments of the political realities in these voting cases. The majority confounds common sense with its pretense that partisan voting sheds no light while accepting its totals as proof of bloc voting. This despite the inescapable fact that partisan voting is the single most powerful explanatory variable in the data. Its glib dismissal makes this a fairyland. As uncertain as our congressional instructions are, they surely did not contemplate the collapse of the Voting Rights Act into a partisan fray gleefully attended by federal judges, none of whom are of course political in any partisan way. When the Republican National Committee and the NAACP join hands in attacking at-large districts surely we can be honest enough to observe that dilution of the voting rights of minorities has now become cover for political factions, coalitions, and parties to gain advantage. This is politics as usual except it has now walked through the courthouse door and so far we haven't said they are in the wrong hall. It is time we do that.

The next step must be to vacate the panel opinion and take this case en banc. Having lost the compass of race the majority is unable to develop a meaning for even dilution. The result is that the majority has looked inward and what we see is their own display and sense of the result they prefer in this case. In their march to result, the powerful confluence of tradition and state sovereignty has no voice. I would have resolved this case with the proposed opinion which is attached hereto as Appendix A.

APPENDIX A

This panel again turns to the Voting Rights Act and the election of state judges. In our first effort we held that the Act covers judicial elections but concluded that the election of state district judges in county-wide elections in Texas did not violate § 2. League of United Latin American Citizens v. Clements, 902 F.2d 293 (5th Cir.1990) ("LULAC I "). We considered the history of judicial elections in Texas and the office of district judge--the court of general jurisdiction. Our study persuaded a majority of the panel that Texas had a special interest in linking the jurisdictional and elective base of the trial courts, an interest accented by unwavering support throughout Texas history. Finding no truly informing analogues for resolving such an attack on at-large voting supported by a unique state interest, we looked to the weighing constructs familiar to the act. We concluded that, as a matter of law, the state interest linking jurisdiction and elective base outweighed its potentially dilutive effect. LULAC I, 902 F.2d at 308. The parties questioned only the at-large election of judges and have not questioned other elements of process for electing Texas district court judges.

A majority of this court sua sponte ordered reconsideration of the panel decision en banc. League of United Latin American Citizens v. Clements, 914 F.2d 620 (5th Cir.1990) ("LULAC II "). The en banc court held that § 2 of the Act did not apply to judicial elections, rejecting the contrary view of the panel.

The Supreme Court granted certiorari and reversed, holding that § 2 does apply to state judicial elections. Houston Lawyers' Ass'n v. Attorney General, [--- U.S. ----] 111 S. Ct. 2376 [115 L. Ed. 2d 379] (1991). The Supreme Court also held Texas has an especial interest in linking the elective and jurisdictional bases of district judges. Houston Lawyers', [--- U.S. at ----], 111 S. Ct. at 2381. The Court did not agree, however, that this state interest outweighed its dilutive effect in all cases, as a matter of law. Rather, the Court held that outcomes of the balancing will be case-specific. It also held that the balance will be struck in an inquiry into the totality of the circumstances. Justice Stevens explained that the state interest in linkage was to be weighed in deciding "whether a § 2 violation occurred." Id. Justice Stevens made plain that assessing the linkage interest is part of the determination of liability and not remedy alone. It was at this juncture that the Supreme Court disagreed with the Department of Justice position, choosing to rest between the "matter of law" view of the concurring opinion and the "goes only to remedy" view of the Department of Justice.

On remand the en banc court in turn remanded to the panel. We have had the benefit of oral argument and additional briefing and now must decide whether the district court erred in holding that Texas' system of electing judges at-large in nine Texas counties violates § 2 of the Voting Rights Act.

This is nominally a single case with common legal issues. However, after setting the common legal standard, we face nine distinct factual patterns. We will examine the findings of the district court in each county because each has unique political histories and cultures. In doing so, we heed the Supreme Court's admonishments to look to the political realities.

On July 11, 1988, individual voters and organizations, including the League of United Latin American Citizens sued in federal district court alleging that Texas' system of electing state trial judges violated plaintiffs' federal rights in several Texas counties1 under § 2 of the Voting Rights Act and the U.S. Constitution. Texas voters elect their trial judges in county wide elections. A voter in Dallas County, for example, may vote for all of the trial courts of general jurisdiction in the county, although each candidate runs for a particular court within the county's judicial system.

Plaintiffs contend that this method of electing trial judges violates § 2 of the Voting Rights by impermissibly diluting the voting power of Hispanics and blacks. Plaintiffs proceed on behalf of different ethnic groups or combinations of ethnic groups in different counties. Depending on the county, they argue that Hispanic voters, black voters, or the combination of both Hispanic and black voters "have less opportunity than other members of the electorate to participate in the political process and elect representatives of their choice."

The case was tried in September 1989 in the U.S. District Court for the Western District of Texas. On November 8, 1989, the district court issued its opinion, sustaining plaintiffs' claim in every county. The district court found that, in all nine counties, minority voters were geographically compact and voted cohesively for the same candidates and that these candidates were consistently defeated by a majority of white voters who voted as a bloc. These findings, according to the court, constituted a prima facie case of illegal vote dilution under Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752 [92 L. Ed. 2d 25] (1986).

The district court also found that, under the totality of the circumstances, this dilution violated § 2 of the Voting Rights Act. In all counties, the district court based its finding of dilution on five facts. These were (1) a general history of discrimination; (2) underrepresentation of minorities on the district court bench; (3) the majority run-off requirement in primary races; and (4) the prohibition against single-shot voting. In the five larger counties, the district court also found that the counties' large size could hinder minority electoral success. Finally, in Dallas County only, the district court found evidence of racial appeals in elections.

Opting to follow Part III(c) of Justice Brennan's opinion in Gingles, a part of the opinion that a majority of justices did not join, the district court ruled after trial that the cause of county voting patterns was legally irrelevant. The district court decided to ignore fully developed evidence regarding voters' partisan affiliation--evidence pointing unerringly to the conclusion that partisan affiliation of candidates was the single most powerful explanation of election outcomes.

In evaluating the totality of the circumstances, the district court gave little weight to the state's historical insistence on linkage. The district court held that any state interest in linkage could be protected under a system of electing judges from single-member districts. As we will explain, this both under-valued and misapprehended the state interest.

The district court judgment declared that Texas' at-large system of electing judges in nine counties violated § 2 of the Voting Rights Act and enjoined future elections under an at-large system. The district court at the same time rejected the constitutional attack finding no intent to dilute minority votes by county-wide election of district judges.

The parties presented statistical studies of county voting patterns in both judicial and non-judicial elections as well as opinions and anecdotal evidence of local politics. Supporting data differed, although studies included many of the same elections. Defendants' expert witness, Dr. Taebel, sometimes analyzed elections with no minority candidate. Plaintiffs' experts, Drs. Brischetto and Engstrom, limited their studies to those elections with a minority candidate.

Plaintiffs, however, did not examine every election in which there were minority candidates. In counties where plaintiffs were proceeding on behalf of black voters only, plaintiffs studied only those elections in which black candidates had run and ignored those elections with Hispanic candidates. Likewise, in counties where plaintiffs were proceeding exclusively on behalf of Hispanic voters, plaintiffs examined only those elections in which Hispanic candidates had participated, ignoring those races with black candidates. Simultaneously, plaintiffs urged that blacks and Hispanics were members of a single cohesive voting group in Ector, Midland, and Lubbock Counties. Plaintiffs considered all elections in which a minority had been a candidate only in these three counties.

Both parties offered evidence that minority voters in each county voted cohesively and that the minority-preferred candidate was the Democratic candidate. There was also evidence that minority residence was sufficiently compact that a minority district could be drawn in every county.

A successful claim that an at-large election dilutes the votes of protected minorities under § 2 of the Voting Rights Act must first meet the three threshold criteria explained in Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752 [92 L. Ed. 2d 25] (1986): (1) the minority group in question is sufficiently large and geographically compact that it could constitute a majority in a single-member district; (2) the minority group is politically cohesive; and (3) the white majority votes sufficiently as a bloc to cause it usually to defeat the minority's preferred candidate. Id. at 51, 106 S.Ct. at 2766-67; Westwego Citizens for Better Gov't v. City of Westwego, 872 F.2d 1201, 1205-06 (5th Cir.1989).

Proving the three Gingles factors, however, does not prove liability under § 2. Plaintiffs must further show that, under the "totality of circumstances," the members of the protected class "have less opportunity than other members of the electorate to participate in the political processes and to elect representatives of their choice." 42 U.S.C. § 1973(b); see also Monroe v. City of Woodville, 881 F.2d 1327, 1330 (5th Cir.1989).

This inquiry into the "totality of circumstances" is guided, in part, by the nine criteria discussed in the Senate Judiciary Committee Report accompanying the 1982 amendment to the Voting Rights Act. These are "Zimmer factors." Drawn in large part from this court's opinion in Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir.1973) (en banc) aff'd on other grounds sub nom., East Carroll Parish School Bd. v. Marshall, 424 U.S. 636 [96 S. Ct. 1083, 47 L. Ed. 2d 296] (1976), they consist of nine signals of diminished opportunity for political participation of the minority group and election of the representatives of their choice. As the Senate Judiciary Report makes clear, these Zimmer factors are not the exclusive measures of dilution under the totality of the circumstances. S.Rep. 417, at 29, reprinted in 1982 U.S.Code, Cong. & Admin.News at 207.

The Zimmer factors guide the inquiry into whether the complaining minority group lacks "an equal opportunity to participate in the political process and to elect candidates of their choice." Gingles, 478 U.S. at 44, 106 S. Ct. at 2763. Inquiry into the totality of the circumstances is more than mechanical "factor-counting" with its illusion of neutral decisionmaking. It is "a searching practical evaluation of the past and present reality" and "a functional view of the political process." Id. at 45 [106 S. Ct. at 2764]. Its focus is always whether the minority group has an opportunity for political participation equal to that of other voters.

But describing the trip is far easier than making it. It becomes especially complicated when we consider elections for trial court judges. We have earlier explained and will not here again rehearse that the judicial office differs from other elective offices in that the "appearance of independence and fairness [is] so central to the judicial task." LULAC II, 914 F.2d at 646 (Higginbotham, J., concurring); see also Gregory v. Ashcroft, [--- U.S. ----, ----] 111 S. Ct. 2395, 2407 [115 L. Ed. 2d 410] (1991) (discussing how the deliberately fostered independence of judicial offices makes them different from other elective positions). The Supreme Court recognized that Texas' special interest in linking elective and jurisdictional base of its trial court judges was a "legitimate factor" to be weighed among the totality of circumstances to determine liability. Houston Lawyers', [--- U.S. at ----], 111 S. Ct. at 2381.

Our task is to determine coherent means for weighing Texas' linkage interest considered as part of the totality of the circumstances. This task is made more difficult because the traditional totality of circumstances analysis asks only whether there was dilution. Including the linkage interest in the § 2 assessment changes the usual weighing process, since the state's interest is not a signal of dilution like the Zimmer factors. Instead, it is a factor to be balanced against such signals. We ask not only whether, under the totality of the circumstances, there was dilution but also whether such found dilution is sufficiently substantial to support a finding of liability, given the state's interest in linkage. In other words, the state's interest is weighed against proven dilution. Houston Lawyers', [--- U.S. at ----] 111 S. Ct. at 2381 (noting that linkage interest does not always "outweigh proof of racial vote dilution").

We must also review the district court's legal conclusion that evidence of the defeat of minority-preferred candidates because of partisan affiliation rather than race was not relevant. Finally, we must determine the universe of elections by which we test success of the minority-preferred candidate. These preliminary legal questions are the foundation upon which any analysis of the district court's factual finding of dilution must be based. We will then apply these principles to the district court's factual findings on a county-by-county basis.

At trial, defendants maintained that partisan affiliation rather than race of either the voters or the candidates caused the consistent defeat of the minority-preferred candidate. Citing to Justice Brennan's plurality opinion in Gingles, the district court ruled after the trial that the cause of the voting patterns that led to the consistent defeat of the minority-preferred candidate was not relevant under Gingles.

Defendants maintain that this holding was erroneous. We agree. Both logic and the weight of authority make plain that the cause of racially polarized voting patterns can be relevant to the vote dilution inquiry. In particular, we hold that there is no illegal dilution of minority voting when polarized voting patterns are the product of partisan affiliation untainted by racial politics. We find in the 1982 amendments to the Voting Rights Act, their relation to earlier dilution jurisprudence, and the Supreme Court's decision in Gingles powerful support for the conclusion that § 2 does not prohibit electoral systems that allow the minority-preferred candidate to be defeated by partisan voting patterns unaffected by race. Proof of majority voting based on party affiliation prevents the showing of bloc voting required by Gingles.

Section 2 of the Voting Rights Act prohibits "denial[s] or abridgment[s] of the right of any citizen of the United States to vote on account of race or color...." 42 U.S.C. § 1973(a) (emphasis added). The provision by its terms requires a causal link between race and the electoral scheme's impediment to voters' equal participation in the political process. The Act does not reach inequality of participation not caused by ("on account of") race or color.

Our insistence that a nexus to race be maintained in applying § 2 does not rest alone on a parsing of statutory language, although the Act's plain language alone is sufficient. Rather, we insist that application of § 2 cannot escape its racial tether because that tether restrains the court from constitutional difficulty. The remedial power of Congress under the Civil War Amendments is for wrongs they prohibit. So the line between interest group politics and dilution of votes on account of race is both statutory and constitutional. Then-Justice Rehnquist summarized these limits. City of Rome v. United States, [446 U.S. 156, 206] 100 S. Ct. 1548, 1577 [64 L. Ed. 2d 119] (1980) (Rehnquist, J., dissenting); see also Oregon v. Mitchell, 400 U.S. 112, 152 [91 S. Ct. 260, 279, 27 L. Ed. 2d 272] (1970) (Harlan, J., concurring in part and dissenting in part).

Our task, then, is to locate the meaning of this link between race and inequality in political participation. We must decide whether this link is satisfied when the minority-preferred candidate consistently fails to win a majority of the white vote and consistently loses, if the losses were the result of partisan affiliation rather than racial politics. We conclude that such consistent defeat of the minority-preferred candidate cannot be "on account of" race, unless it is tied to racial bias in the electorate. This is the heart of § 2.

In Whitcomb v. Chavis, 403 U.S. 124 [91 S. Ct. 1858, 29 L. Ed. 2d 363] (1971), the Supreme Court held that the at-large election of state representatives and state senators from a multi-member district, Marion County, did not violate the Equal Protection Clause of the Fourteenth Amendment. Black residents in one part of Marion County, referred to as the "ghetto" by the Whitcomb Court, voted heavily for the Democratic Party's candidates; however, as the rest of the County voted for the Republican candidate, the "ghetto's" preferred candidate was defeated in four out of five elections between 1960 and 1968. The Whitcomb plaintiffs maintained that, if the county were redistricted, the preferred candidate of the minority voters would win more frequently.

Justice White, writing for five members of the Court, rejected this challenge to the multi-district system. The Whitcomb Court noted that the Democratic Party regularly slated candidates "satisfactory to the ghetto" but that the Democratic Party lacked the votes necessary to prevail frequently in the county's general elections. Whitcomb, 403 U.S. at 151-52 [91 S. Ct. at 1873]. The Court further noted that both political parties were open to minority voters and candidates and that there were no impediments to minority participation in primaries, voting registration, or other aspects of elections. The Court, therefore, concluded that

the failure of the ghetto to have legislative seats in proportion to its population emerges more as a function of losing elections than of built-in bias against poor Negroes. The voting power of ghetto residents may have been 'cancelled out' as the District Court held, but this seems a mere euphemism for political defeat at the polls.

Id. at 153 [91 S. Ct. at 1874]. According to the Whitcomb Court, "[t]he mere fact that one interest group or another ... has found itself outvoted and without legislative seats of its own provides no basis for invoking constitutional remedies where, as here, there is no indication that this segment of the population is being denied access to the political system." Id. at 154-55 [91 S. Ct. at 1875]. The Court, therefore, rejected the proposition that "invidious discrimination ... results when the ghetto, along with all other Democrats, suffers the disaster of losing too many elections." Id. at 153 [91 S. Ct. at 1874] (emphasis added).

It bears emphasis that in Marion County there was inevitably a high correlation between black votes and party votes. There was no suggestion in Whitcomb that such raw correlations supported the idea that votes for Democratic candidates were proxies for racial interests. Nor did the Whitcomb Court dispute the findings of the three-judge panel--that black residents within the "ghetto" suffered from high unemployment, high levels of welfare assistance, poor housing, and other indicia of poverty. Whitcomb, 403 U.S. at 132 [91 S. Ct. at 1863]; Chavis v. Whitcomb, 305 F. Supp. 1364, 1376-81 (S.D.Ind.1969). The Court nevertheless found no dilution. The clear implication of Whitcomb, therefore, is that consistent defeat of minority-preferred candidates as a result of polarized voting, even when combined with poverty and other lingering effects of past discrimination, does not suffice to show illegal dilution, when such defeat is the result of partisan competition between parties that are fully open to members of all races.

2. 1982 Amendments to § 2 of the Voting Rights Act

According to the Senate Judiciary Committee Report, the 1982 amendments to the Voting Rights Act enacted the "results test" contained in White v. Regester, 412 U.S. 755 [93 S. Ct. 2332, 37 L. Ed. 2d 314] (1973) and Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir.1973) (en banc), aff'd on other grounds sub nom., East Carroll Parish School Bd. v. Marshall, 424 U.S. 636 [96 S. Ct. 1083, 47 L. Ed. 2d 296] (1976). Congress understood Whitcomb to apply the "White/ Zimmer " results test. S.Rep. 417 at 19-21, 30 (referring to the results test in "Whitcomb, White, Zimmer, and their progeny"), reprinted in 1982 U.S.Code Cong. & Admin.News at 196-99, 207. The Judiciary Committee pointed to the at-large systems upheld in Whitcomb to refute contentions by opponents of the 1982 amendments that its results test would require the automatic invalidation of at-large election systems. S.Rep. 417 at 33, reprinted in 1982 U.S.Code Cong. & Admin.News at 211.

We are not persuaded that § 2, as amended by the 1982 amendment and read in the light of Whitcomb, prohibits at-large districts in which the minority-preferred candidate is defeated by the voters' partisan affiliation. Section 2 was intended to prohibit electoral systems that lead to "discriminatory results." The term is best understood in light of § 2's prohibition on electoral schemes that impede minority voters' political participation "on account of race or color." See Solomon v. Liberty County, 899 F.2d 1012, 1021-1037 (11th Cir.1990) (Tjoflat, J., concurring).

Under the results test, consistent defeat of minority-preferred candidates alone will not establish that minority voters are deprived of an equal opportunity to choose representatives on account of race. Such consistent defeat, by itself, is not the "discriminatory result" prohibited by § 2. If it were, the 1982 amendments to § 2 would have abandoned the law of Whitcomb.

The legislative history of the 1982 amendments indicates that voters' race-conscious voting is relevant to the determination of the statute's required nexus to "race or color." The Judiciary Committee's report repeatedly refers to the need to prevent "racial politics" and "racial political considerations" from interfering with minority voters' equal opportunity for political participation. See S.Rep. 417 at 33, reprinted in 1982 U.S.Code Cong. & Admin.News at 211; Solomon, 899 F.2d at 1027-32 (minority voters are denied "equal opportunity to participate meaningfully in elections" in "communities where racial politics ... dominate the electoral process"); United States v. Marengo County Comm'n, 731 F.2d 1546, 1567 (11th Cir.1984) ("section 2 is intended not to create race-conscious politics, but to remedy it where it already exists") (emphasis added). "Racial politics" implies racially conscious politics. It does not include politics in which minority voters only fail to achieve maximum feasible success because they were outvoted by other interest groups.

The 1982 amendment responded to City of Mobile v. Bolden, 446 U.S. 55, 100 S. Ct. 1490 [64 L. Ed. 2d 47] (1980). Its overarching purpose was to eliminate Bolden 's requirement that plaintiffs prove that legislators intended to discriminate when they enacted or maintained a challenged electoral system. With the 1982 amendment, an electoral system cannot escape § 2 by innocent birth if its result is to dilute voting power "on account of race or color," but the "on account of race or color" language remained a part of the Act.

It is suggested that this smuggles the banned intent into our analysis. We disagree. Rather, we are persuaded that this is the meaning of the plain language of the Act as amended. Under the 1982 amendments the state's intent in creating or maintaining an electoral system is not controlling. Jones v. City of Lubbock, 727 F.2d 364, 378 (5th Cir.1984) (purpose of 1982 amendments was to "prohibit electoral practices and procedures that created discriminatory results even though the responsible government body had not installed or maintained the electoral practice or procedure in order to discriminate") (emphasis added); see also S.Rep. 417 at 27 (1982 amendment designed to "make clear that plaintiffs need not prove a discriminatory purpose in the adoption or maintenance of the challenged system or practice in order to establish a violation" (emphasis added)), reprinted in 1982 U.S.Code Cong. & Admin.News at 205. Nowhere in the statute or its legislative history is there any indication that voters' racial prejudice is not relevant to voting patterns. Indeed, the testimony by supporters of the "results" test suggest that "private discrimination" is central to the issue of whether there is illegal dilution under § 2. See, e.g., Hearings on the Voting Rights Act Before the Subcomm. on the Constitution of the Senate Comm. on the Judiciary, 97th Cong., 2nd Sess. 1367-68 (1982) (statement of Drew Days, Associate Professor of Law, Yale University) (§ 2's result test would operate "where a combination of public activity and private discrimination have joined to make it virtually impossible for minorities to play a meaningful role in the electoral process") (emphasis added).

The Zimmer factors, incorporated into the dilution analysis by the Senate Judiciary Committee's report, are directly concerned with the cause of voting patterns. One of the Zimmer factors concerns "whether political campaigns have been characterized by overt or subtle racial appeals"; another is the "extent to which members of the minority group have been elected to public office." Both are relevant because they shed light on the political climate of the community--that is, the likely causes of voters' behavior.

Success of minority candidates goes directly to whether voters are influenced by race. As the facts of this case disclose, the minority candidate is not always the preferred candidate of the minority group. Therefore, the success of minority candidates must be relevant beyond showing that minority voters elect candidates of their choice. The success of minority candidates is significant in part because it indicates that white voters' racial bias is not excluding a candidate even when the minority-preferred candidate loses. Significantly, this Zimmer factor, directly relevant to the "on account of race" limits of the statute, is the only factor actually mentioned as probative in the text of the statute itself.

Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752 [92 L. Ed. 2d 25] (1986), is the Supreme Court's most detailed interpretation of § 2 as amended in 1982. Plaintiffs maintain that voters' motivations are not relevant to the vote dilution inquiry under Gingles. This asks too much of Gingles. Of the four opinions in Gingles, only Justice Brennan's plurality opinion concluded that causation was not relevant to the ultimate question of illegal racial vote dilution. Gingles, [478 U.S. at 62-74] 106 S. Ct. at 2772-78. Five Justices, writing separately, refused to join this part of Justice Brennan's opinion.

Justice White's brief opinion illuminates the relevance of partisan affiliation. Justice White notes that, in a district that is 60% white and 40% black, if six white and two black Republicans run against six white and two black Democrats, and all of the Republicans win, then, under Justice Brennan's analysis, there would be a likely § 2 violation, if black voters voted Democratic and were geographically compact. However, Justice White states that:

[t]his is interest-group politics rather than a rule hedging against racial discrimination. I doubt that this is what Congress had in mind in amending § 2 as it did, and it seems at odds with the discussion in Whitcomb v. Chavis, 403 U.S. 124, 149-160, 91 S. Ct. 1858, 1872-1878, 29 L. Ed. 2d 363 (1971).

Id. [478 U.S. at 83, 106 S.Ct.] at 2783. Justice O'Connor, writing for herself and three other Justices, agreed with Justice White that the race of the candidate is relevant to the finding of racial vote dilution. Id. [478 U.S. at 100-01, 106 S.Ct.] at 2792. Justice O'Connor stated that the presence or absence of racial hostility among voters is relevant to the vote dilution inquiry, because such bias was relevant to determining whether minority voters could influence officials for whom they did not vote. Crucial to Justice O'Connor's analysis is the premise that minority voters can have influence even if their preferred candidates lose, if "other indirect avenues of political influence" are not barred by racial animosity in the community. Id.

The majority of Justices in Gingles, in short, took the position that both the race of the candidate and the voters' motivations were relevant to the inquiry into racial vote dilution. As Justice O'Connor points out, evidence that white voters reject minority-preferred candidates for reasons other than race precludes a showing of racial bloc voting by whites. Id. Moreover, Justice White's opinion implies that, where the minority-preferred candidate loses the white vote because of partisan affiliation, and the white voters give their support to black candidates of the white voters' preferred party, then the defeat of the minority-preferred candidate is not the result of illegal vote dilution but rather is the result of interest-group politics. This division of the Court cuts deep, reflecting quite different visions of voting rights and their statutory treatment. The division also reflects fundamentally different political views of factions and our constitutional arrangement for accommodating their simultaneous demands for fluidity and fixity.

We find that the law before the 1982 amendments of § 2, the amendments themselves, and Gingles support the conclusion that racial vote dilution does not occur when the minority-preferred candidate is defeated by partisan affiliation and not "on account of race or color." In the structure erected by Gingles, we find that proof of majority voting based on partisan voting patterns unaffected by race precludes a finding of the third prerequisite: white bloc voting to defeat minority-preferred candidates. The black voter who supports a losing Democratic candidate in a county where the majority of voters are Republicans is in precisely the same position as the white Democratic voter in such a county. If the white Democratic voter is not the victim of a racially discriminatory result, then it is impossible to see why the black voter is. Both suffer an identical "political defeat at the polls," not "built-in bias" against a racial group. Whitcomb, 403 U.S. at 153 [91 S.Ct. at 1874-75]. Both are defeated because they belong to the weaker political party. Creating safe districts to protect the black Democrat but not the white Democrat subverts the very idea of political equality that § 2 is supposed to protect. See Alan Howard & Bruce Howard, The Dilemma of the Voting Rights Act--Recognizing the Emerging Political Equality Norm, 83 Colum.L.Rev. 1615, 1618-19 (1983) ("Giving some groups safe districts and proportional representation and not others thus necessarily treats the groups, and individual voters, unequally."); see also Whitcomb, 403 U.S. at 154, 91 S. Ct. at 1875 (questioning whether "poor Negroes of the ghetto [are] any more underrepresented than poor ghetto whites who also voted Democratic and lost"). "The mere fact that one interest group or another ... has found itself outvoted and without legislative seats of its own" provides no basis for finding illegal racial dilution where "there is no indication that this segment of the population is being denied access to the political system." Whitcomb, 403 U.S. at 154-55 [91 S.Ct. at 1874-75].

Section 2 was not enacted to create safe districts for political parties. It was intended to prevent minority voters from being shut out of interest group politics by race or color. Racially influenced voting presents a unique obstacle to coalition-building, different in kind from disagreement over ideology or interest. As one commentator has observed, "prejudice is a lens that distorts reality. We are a nation of minorities and our system thus depends on the ability and willingness of various groups to apprehend those overlapping interests that can bind them into a majority on a given issue; prejudice blinds us to overlapping interests that in fact exist." John Hart Ely, Democracy and Distrust 153 (1980). Where the normal interplay of contending factions is untainted by "racial politics," and there are no barriers preventing black voters from joining and fully participating in all major political parties, forming coalitions, and taking their chances at the polls, there is no "discriminatory result" within the meaning of § 2.

We do not ignore the concerns of Justice Brennan's plurality opinion in Gingles that apparently race-neutral reasons for the minority-preferred candidates' defeat may be a proxy for race. Political party can become a proxy for race when voters support or oppose a political party because of their bias against racial or ethnic groups. See Thomas B. Edsall & Mary D. Edsall, Chain Reaction: The Impact of Race, Rights, and Taxes on American Politics 137-53 (1991). The Act is violated when a minority-preferred candidate is defeated by the bias of white voters directed toward the candidate or his constituency. Describing the election in ostensibly race-neutral terms of partisan affiliation will not change the result.

Such a proxy, however, is not signaled by raw statistical correlation of partisan affiliation to race. Of course, socio-economic characteristics may have some correlation to race. Gingles, [478 U.S. at 69-70] 106 S. Ct. at 2776. This correlation may lead members of minority groups to give stronger support to certain political platforms than non-members give. As a result, a majority of minority voters may support a political platform opposed by a majority of non-minority voters, and the minority-preferred candidate may be consistently defeated.

The partial correlation of differences in political interest or opinion with race, however, does not demonstrate that parties are proxies for racial groups. Partisan affiliation may only imperfectly map membership in racial groups. For instance, in several of the counties, the Democratic Party has the support of a large majority of black voters, while the Republican Party has the support of most white voters. However, about 30%-40% of the white voters also support the Democratic Party. The Democratic Party, therefore, does not represent interests uniquely or even mainly shared by black voters, and it is not a proxy for a distinctively black agenda. At best, this correlation is evidence only that the Democratic Party here represents a coalition supported by black voters, among others.

Opposition to such an interest group coalition is a far cry from opposition to a candidate on account of race or color. See Terrazas v. Clements, 581 F. Supp. 1329, 1351-52 (N.D.Tex.1984) (three-judge panel) (noting that "[t]he testimony ... suggested that partisan affiliation has, to some extent, eclipsed the importance of [racial] bloc voting in Texas politics"). Such opposition is no more than the coalition building that has been part of the politics of this nation since its founding. In sum, proof that the minority-preferred candidate is defeated because he adopts a political platform opposed by a majority of voters is not proof that the defeat was the result of racial vote dilution prohibited by § 2.

But, it is urged, plaintiffs may face insuperable barriers in trying to prove the cause of voting patterns. Gingles, [478 U.S. at 66-67], 106 S. Ct. at 2774. As Justice Brennan noted in Gingles, the 1982 amendments were enacted in large part to alleviate the difficulty of proof faced by plaintiff alleging vote dilution under Bolden. Congress made some things clear, including the fact that plaintiffs need not produce a "smoking gun" of racial bias in the original adoption or subsequent maintenance of an electoral system for their prima facie case of illegal vote dilution. See Abigail M. Thernstrom, Whose Votes Count? Affirmative Action and Minority Voting Rights 120-22 (1987).

We impose no new burden of producing a "smoking gun" on the plaintiffs. As always, the district court may infer illegal racial dilution with some circumstantial evidence of racially influenced voting by the electorate. Such circumstantial evidence may include (but is not limited to) proof of the relevant Zimmer factors--racial appeals, non-responsiveness of elected officials, history of discrimination, and lack of success of minority candidates. The consistent and predictable refusal of white voters to support minority candidates would also be relevant to an inference of race-conscious politics, as would extremely close correlation of partisan affiliation with race of voters.

Where the factfinder infers racial politics from some combination of these facts and also finds that such race-conscious voting interacts with some electoral practice to prevent minority voters from having an equal opportunity to elect representatives of their choice, we will not second-guess the district court's inference of racial vote dilution. Likewise, the consistent and predictable willingness of white voters to give to minority candidates of the voters' preferred party support equal or greater than the white voters give to white candidates of the same party is powerful evidence that party rather than race determines voting patterns.

B. Texas' Interest in Defining the Judicial Office

By making coterminous the elective and jurisdictional bases of trial courts, Texas serves uniquely judicial values of fairness and independence. At-large elections insure that trial court judges are each electorally accountable to all of the residents within the area of their primary jurisdiction. This linkage of jurisdictional and elective bases avoids the appearance, if not the fact, of favoritism--protection of their voting constituents--to the prejudice of the other litigants who frequently appear before them. LULAC II, 914 F.2d at 649-51 (Higginbotham, J. concurring). By insuring that judges are accountable to a broad base of all people within their primary jurisdiction rather than one narrow subsection, at-large elections protect "the fact and appearance of independence and fairness [which] are so central to the judicial task." Id. at 646. Presumably for that reason the overwhelming majority of states which select trial judges by election share this structure and electoral scheme. By contrast, the systemic incentives of subdistricting "diminish the appearance if not the fact of ... judicial independence--a core element of a judicial office." Id. at 650.

In Houston Lawyers', the Supreme Court agreed. Justice Stevens noted that Texas' linkage interest is "a legitimate factor to be considered by courts among the 'totality of circumstances' in determining whether a § 2 violation has occurred." Houston Lawyers', [--- U.S. at ----] 111 S. Ct. at 2381. Houston Lawyers' disagreed that this "linkage" interest should defeat liability "automatically, and in every case." Rather, the Court held that the interest is one consideration that must be weighed against other relevant factors to ascertain whether the interest "outweigh[s] proof of racial vote dilution." Id. See also Nipper v. Chiles, 795 F. Supp. 1525, 1548 (M.D.Fla.1992) (holding that "a state's interest in maintaining an electoral system is a legitimate factor to be considered ... in the liability phase of a section two case").

Therefore, while the Supreme Court rejected our contention that the linkage interest in all cases defeated liability under § 2 as a matter of law, the Supreme Court endorsed our position that the linkage interest is relevant to a determination of liability. Indeed, by noting that the linkage interest does not "automatically, and in every case, outweigh proof of racial vote dilution," the Court held that the state interest could outweigh what would otherwise be proof of illegal dilution and thus foreclose liability.

The issue we face is determining when the linkage interest will outweigh other factors and defeat liability under § 2. In resolving this issue, we reject the polar extremes of the parties. The State of Texas and at least one defendant maintain that the interest defeats liability in every case, regardless of the other circumstances in the totality. The Supreme Court rejected this position when it held that the linkage interest does not "automatically, and in every case, outweigh proof of racial vote dilution." Houston Lawyers', [--- U.S. at ----] 111 S. Ct. at 2381.

We also reject the position of plaintiffs that the linkage interest can never defeat liability under the totality of the circumstances if "illegal" dilution is otherwise established. The plaintiffs maintain that only the absence of a compelling state interest in an electoral scheme is relevant to liability. According to plaintiffs, such an absence "is an optional factor" that plaintiffs can use to support a finding of illegal dilution; however, according to plaintiffs, the existence of a compelling interest can never defeat liability that is otherwise established under the totality of the circumstances. This position was also rejected by the Supreme Court. This state interest is to be weighed as part of the totality of the circumstances. Houston Lawyers', [--- U.S. at ----] 111 S. Ct. at 2381.

Plaintiffs urge that the Zimmer factor of non-tenuous state policy is among the least important of the factors for determining dilution and cite Jones v. City of Lubbock, 727 F.2d 364, 383 (5th Cir.1984), and United States v. Marengo County Comm'n, 731 F.2d 1546, 1571 (11th Cir.1984), in support of this contention. These decisions state only that defendants cannot defeat liability by using the non-tenuous policy justification of an electoral scheme to prove that scheme "does not have a discriminatory intent." Marengo County, 731 F.2d at 1571. See also Terrazas v. Clements, 581 F. Supp. at 1345 n. 24 ("In the case of tenuousness, the lesser weight is consistent with the change in emphasis from intent to results. The principal probative weight of a tenuous state policy is its propensity to show pretext" (emphasis added)).

The plaintiffs' argument misses the point. The linkage interest is not urged as a non-tenuous policy to prove there was no discriminatory intent in adopting or maintaining the at-large election system. Rather, linking electoral to jurisdictional base is relevant because it serves objectively substantial interests. The plaintiffs confuse the inquiry into whether an interest is substantial with an inquiry into whether an interest is nontenuous. Proof of a merely non-tenuous state interest discounts one Zimmer factor but cannot defeat liability. It does not follow, however, that proof of a substantial state interest cannot defeat liability. As we have explained, the Voting Rights Act largely codifies Fourteenth Amendment jurisprudence embodied in White v. Regester, 412 U.S. 755 [93 S. Ct. 2332, 37 L. Ed. 2d 314] (1972). Jones, 727 F.2d at 379-80. The substantiality of the state's interest has long been the centerpiece of the inquiry into the interpretation of the Civil War Amendments and their interplay with the civil rights statutes.

Having rejected both poles of the argument--that the linkage interest either always or never defeats § 2 liability, we turn to when the linkage interest precludes a § 2 vote dilution violation. The weight of Texas' interest is virtually assigned by a Supreme Court decision handed down on the same day that the Supreme Court decided Houston Lawyers'. In Gregory v. Ashcroft, [--- U.S. ----, ----] 111 S. Ct. 2395, 2404 [115 L. Ed. 2d 410] (1991), the Supreme Court held that the Age Discrimination in Employment Act does not apply to judicial offices in Missouri. The Court noted that "the authority of the people of the States to determine the qualifications of their most important government officials ... ' "lies at the heart of representative government." ' " Id. [--- U.S. at ----, 111 S.Ct.] at 2402 (quoting Bernal v. Fainter, 67 U.S. 216, 221 [104 S. Ct. 2312, 2316, 81 L. Ed. 2d 175] (1984)). Gregory noted that "the States' power to define the qualifications of their officeholders has force even as against the proscriptions of the Fourteenth Amendment." Id. [--- U.S. at ----, 111 S.Ct.] at 2405. To protect this power to define the judicial office, Gregory required a clear statement from Congress for an override of qualifications imposed by the State for important state government offices. Id. [--- U.S. at ----, 111 S.Ct.] at 2406.

"The people of Missouri have a legitimate, indeed compelling, interest in maintaining a judiciary fully capable of performing the demanding tasks that judges must perform." Id. [--- U.S. at ----, 111 S.Ct.] at 2407. If that interest is compelling, the people of Texas have at the very least a substantial interest in defining the structure and qualifications of their judiciary. Linking elective and jurisdictional bases is a key component of this defining effort. That the Texas interest in linkage of electoral and jurisdictional base is substantial cannot then be gainsaid.

Our confidence in this conclusion is bolstered by the recognition and pursuit of the linkage interest in other states. Courts have recognized the legitimacy and substance of similar linkage interests in Florida and Alabama. See Nipper v. Chiles, 795 F. Supp. 1525, 1548 (M.D.Fla.1992); Southern Christian Leadership Conf. of Ala. v. Evans, 785 F. Supp. 1469, 1478 (M.D.Ala.1992). Of the twenty-nine states which elect their principal trial court judges, including Texas, Alabama, and Florida, twenty-five employ district-wide elections.2 Two others, Mississippi and Louisiana, only recently abandoned the link between jurisdiction and electoral base in order to settle prolonged litigation. The overwhelming preservation of linkage in states which elect their trial court judges demonstrates that district-wide elections are integral to the judicial office, and not simply another electoral alternative.

Thus we note now, as we noted in LULAC II, that the decision to make jurisdictional and elective bases coterminous is more than a decision about how to elect state judges. It is a decision of what constitutes a state court judge. Such a decision is as much a decision about the structure of the judicial office as the office's explicit qualifications such as bar membership or the age of judges. This decision preserves several values, among them a careful balance of judicial independence and judicial accountability.

On the other hand, plaintiffs' interest in full value of minority votes does not translate easily into single member districts for electing trial judges with county-wide jurisdiction. Measured by the concerns of the Act, single-member districts can create perverse results. After subdistricting, a handful of judges would be elected from subdistricts with a majority of minority voters. The price of such re-districting would be that all but a few sub-districts would be stripped of virtually all minority members. The great majority of judges would be elected entirely by white voters. Minority litigants would not necessarily have their cases assigned to one of the few judges elected by minority voters. Rather, the overwhelming probability would be that the minority litigant would appear "before a judge who has little direct political interest in being responsive to minority concerns." LULAC II, 914 F.2d at 650 (Higginbotham, J., concurring).

This concern does not rest on paternalism. It recognizes Texas' historic interest in having district judges remain accountable to all voters in their district. Regardless of the race or residency of particular litigants, judges make choices which affect all county residents. Texas has insisted that they answer to all county voters at the ballot box.

Unlike officers in legislatures or even judges on collegial benches who make decisions in groups, each district judge holds a single-member office and acts alone. While subdistricting in multi-member offices can enhance minority influence because members from minority subdistricts would participate in and influence all the members and decisions of the larger body, subdistricting for single-member district court judgeships would leave minorities with no electoral influence over the majority of judges in each county.

By contrast, under the present regime, minority voters participate in all judicial elections in each county. This participation gives minority voters the opportunity to influence all elections. As Justice O'Connor noted in her plurality opinion in Gingles, voters can wield influence over elections even when those votes are cast for losing candidates. Gingles, [478 U.S. at 98-99] 106 S. Ct. at 2791. Denying importance to this ability to influence asks that all measures of success be found in the win-loss column. This mandates proportional representation as the measure of dilution, contrary to the explicit terms of § 2. Indisputably, subdistricting would assure the absence of minority influence over the judicial process. See LULAC II, 914 F.2d at 649-50 (Higginbotham, J., concurring); Southern Christian Leadership Conf. of Ala. v. Evans, 785 F. Supp. 1469, 1478 (M.D.Ala.1992) (Hobbs, J.) (by subdistricting judicial positions, "black voters ... will ... be sacrificing [an] extremely valuable political right--the right to vote for all of the judges who will be serving as judges in the circuit wherein they live").

Plaintiffs contend that linking jurisdictional and elective bases does not, in fact, protect these uniquely judicial interests. All of the plaintiffs' arguments reduce to the single contention that Texas does not consistently apply the policy of linking jurisdictional and elective bases. The plaintiffs contend, for instance, that Texas district court judges often adjudicate controversies involving litigants who are not residents of the county and that Texas allows the residents of counties to 'opt out' of the linkage by selecting judges from regions smaller than a county.

We have rejected these arguments in LULAC II and need pause only briefly to address them here. See LULAC II, 914 F.2d at 651. Doing so, we note that in assessing the relationship between the end pursued and the means employed, "our scrutiny will not be so demanding where we deal with matters resting firmly within a State's constitutional prerogatives." Sugarman v. Dougall, 413 U.S. 634, 648 [93 S. Ct. 2842, 2850, 37 L. Ed. 2d 853] (1973). As both Sugarman and Gregory make clear, such "matters" include "the establishment and operation of its own government, as well as the qualifications of an appropriately designated class of public office holders." Sugarman, 413 U.S. at 648 [93 S. Ct. at 2851]. Examining Texas' linking of electoral and jurisdictional bases in light of these considerations, we find that it serves the substantial interests we described.

By drawing attention to venue, plaintiffs only remind us of concerns unique to the trial judge's office. Venue rules preserve judicial fairness by preventing forum-shopping and diminishing the chances of biased adjudication. At the same time, the rules keep most local matters in local courts, where judges are accountable to voters for the legal and policy choices they make. Even though district courts try some cases involving non-resident litigants, Texas' venue rules ensure that they principally handle cases related to the counties comprising their districts. This is particularly true in criminal matters, when venue is based on events related to the offense and not by reference to a defendant's domicile. Tex.Code Crim.Proc.Ann. ch. 13 (Vernon 1977). Similarly, family law matters will almost always be handled by the local district court. See, e.g., Tex.Fam.Code Ann. §§ 3.21, 11.04 (Vernon 1986) (concerning venue in divorce and parent-child relationship suits). Of course quintessentially local matters such as suits against counties or disputes involving title to real property must be tried in the district court of the same county. Tex.Civ.Prac. & Rem.Code Ann. §§ 15.001, 15.015 (Vernon 1986). The argument that Texas' venue rules somehow abrogate its interest in linking jurisdiction and electoral base is illusory.

Likewise, the contention that Texas abandoned its interest in linkage by allowing voters to approve the creation of districts smaller than counties is without merit. As Chief Justice Phillips explained at trial, this provision was merely part of a constitutional and statutory scheme designed to equalize court dockets by allowing the realignment of judicial districts. The 1985 amendment states that a district smaller than a county may not be created unless approved by a majority of county voters. Tex.Const. art. V, § 7(a)(i) [7a(i) ]. In none of Texas' 254 counties have voters voted to break the link between jurisdiction and electoral base. In no district has the linkage been abandoned.

Moreover, even if one county were to subdivide, the interest in linkage would not be lost in the state as a whole. In Mahan v. Howell, 410 U.S. 315 [93 S. Ct. 979, 35 L. Ed. 2d 320], modified 411 U.S. 922 [93 S. Ct. 1475, 36 L. Ed. 2d 316] (1973), the Court recognized that although Virginia divided one county when reapportioning its state legislature, it retained its interest in preserving boundaries of all other political subdivisions. Id. at 327 [93 S. Ct. at 986]. Texas' interest in preserving the structure of its judiciary by linking jurisdictional and electoral boundaries is even greater than a state's interest in observing boundaries in legislative reapportioning.

In finding that Texas' interest is substantial, we recognize that it will not always defeat § 2 liability. Substantiality is not quantifiable, and we translate its force in the practical world of trials to the burden required to overcome it. As we see it, plaintiffs cannot overcome a substantial state interest by proving insubstantial dilution. We hold that the totality of all circumstances besides the state interest must sum to substantial proof of dilution and do so convincingly if they would overcome the state's linkage interest. As a matter of law, Texas' interest cannot be overridden by a totality of circumstances that sum to a marginal case. It will take more to create a fact issue for trial.

We do not now attempt to define in detail what sort of proof of dilution would be substantial enough to override the state's linkage interest. We do not change the nature or usual means of proof. The Zimmer factors remain relevant. In particular, proof of racial appeals in elections, non-responsiveness of elected officials to minority voters, and persistent lack of electoral success by minority candidates must all be considered. We also look to the degree and nature of racial bloc voting to see if such voting patterns suggest racial bias in the electorate.

Two facts are especially relevant to assessing the substantiality of the plaintiff's proof of dilution: first, the willingness of the racial or ethnic majority (in this case, white voters) to give a majority of their votes to minority candidates, and, second, the ability of the racial or ethnic minority to elect candidates of their choice even when opposed by a majority of votes from the racial or ethnic majority. Where the dominant racial or ethnic group repeatedly casts most of its votes for members of racial or ethnic minorities, the likelihood is that racial bias (as opposed to ideological difference) is not an insurmountable obstacle to coalition-building. Likewise, where the racial or ethnic minority is large enough repeatedly to elect candidates of its choice even when opposed by the racial or ethnic majority, one cannot conclude that the minority is incapable of influencing elections because it is submerged in a sea of hostile voters. In either case, the plaintiff's proof of dilution may, as a matter of law, be too insubstantial to overcome the state's substantial interest, even if it might be sufficient to establish liability absent that interest.

2. Weight of Texas' interest is a legal question, not a question of fact

The plaintiffs urge that whether Texas' linkage interest is substantial is a matter of fact for the district court to decide in the first instance. The plaintiffs contend that we may review such a factual determination only for clear error. We disagree.

The Supreme Court has held that the ultimate finding of dilution is a factual matter reviewable only for clear error. A substantial state interest is not inherently preclusive of dilution and is not raised to disprove the existence of dilution. Rather, the state's interest is weighed against proven dilution to assess whether such dilution creates § 2 liability. Houston Lawyers', [--- U.S. at ----] 111 S. Ct. at 2381 (weighing of linkage interest on remand goes to determination of whether interests "outweigh[s] proof of racial vote dilution").

The weight to be assigned to a state's interest under the Fourteenth Amendment has always been a legal question, not a factual question. The determination of the substantiality of Texas' linkage interest under the Voting Rights Act, a statute enacted to enforce the guarantees of the Civil War Amendments, is analogous. We hold that the determination of substantiality of Texas' interest under § 2 is a question of law for this court to determine de novo and not a question of fact that somehow will be described on a county-by-county basis.

3. Absence of less restrictive means to accommodate the linkage interest

Plaintiffs urge that the linkage interest can be accommodated through a scheme of single-member districts, by making the district court judges' area of primary jurisdiction co-extensive with the single-member district from which the district court judge is elected. They provide no evidence that such a radical re-working of the venue of Texas courts would be administratively feasible. The district court likewise simply asserted that such an arrangement of venue limited to a single-member district could accommodate Texas' interests, without a glance at the feasibility of such an arrangement. A glance at Harris County cut into a grid of fifty-nine venue squares is enough to show the bizarre nature of this proposal. Enough time and money is already wasted arguing over which county's courts should try a suit, without the venue depending upon in which voting precinct an event occurred.

We cannot conclude that Texas' interests could be adequately accommodated by such a radical re-working of Texas venue rules. The proposal illustrates how different the judicial offices' at-large election scheme is from legislative and executive at-large elections. Plaintiffs propose essentially that the district court eliminate Texas venue rules and replace them with rules of the district court's own creation. The very necessity for the proposal is a powerful testament to the reality that linkage is an essential part of the structure of the judicial office, much more than the method by which the holder of the office is elected.

C. Relevance of Elections with Hispanic Candidates To

Black-Preferred Candidates' Success

In determining the success of the black-preferred candidate as required by Gingles, the plaintiffs examined only those elections in which black candidates participated. Thus, in Harris, Tarrant, Jefferson, and Dallas counties, the plaintiffs ignored the success of the minority-preferred candidates in elections in which Hispanics had run. The district court seems also to have discounted the importance of these white-Hispanic elections.

The district court also found, however, that black and Hispanic voters were a single cohesive minority group in three counties (Midland, Lubbock, and Ector), apparently because the members of the two minority groups generally voted for the same candidates. Taebel studied a number of elections in each county and determined the percentage of black and Hispanic votes cast for the minority/winning candidate. In Midland County, of the 8 elections Taebel studied, the black and Hispanic voting percentages were within 10% of each other 4 times. In Lubbock County, 6 out of 7 elections showed blacks and Hispanics voting together within 10%. In Ector County, blacks and Hispanics voted within 10% of each other in only 2 out of 10 elections.

The district court made no findings as to whether black and Hispanic voters were also a cohesive group in other counties. However, nothing in the record suggests that black and Hispanic voters were not equally cohesive outside Ector, Midland, and Lubbock counties. Indeed, in Harris and Tarrant Counties, the undisputed data in Taebel's studies show that a large majority of both Hispanics and blacks usually supported the same candidates. In Harris County, Taebel studied 46 elections. In 35 of these, the black and Hispanic vote percentages were within 10% of each other. Out of 17 elections in Tarrant County, 13 involved blacks and Hispanics voting for the same candidate within 10%. Thus, the record shows that blacks and Hispanics were actually more cohesive in Harris and Tarrant County than in Midland and Ector, two out of three counties in which the district court made a finding of cohesion.

If black and Hispanic voters form a cohesive minority group in counties outside Ector, Midland, and Lubbock, then the district court clearly erred in ignoring races with Hispanic candidates in assessing the success of the black-preferred candidate. This court has held that elections between white candidates alone are generally less probative in determining the success of the minority-preferred candidates, because such elections do not give minority voters the choice of a viable minority candidate. Campos v. City of Baytown, 840 F.2d 1240, 1245 (5th Cir.1988) ("The district court was warranted in its focus on those races that had a minority member as a candidate"); Citizens for a Better Gretna v. City of Gretna, 834 F.2d 496, 503 (5th Cir.1987). As we noted in Gretna, " Gingles is properly interpreted to hold that the race of the candidate is in general of less significance than the race of the voter--but only within the context of an election that offers voters the choice of supporting a viable minority candidate." Gretna, 834 F.2d at 503.

Gretna does not explain which elections are relevant in determining the success of the black-preferred candidate in communities where black and Hispanic voters vote cohesively. However, if black and Hispanic voters are truly a cohesive minority group, then logic requires that the district court consider elections with Hispanic candidates, because Hispanic candidates would count as members of this bi-minority "group." If black and Hispanic voters are, in effect, a single minority group with common "social, economic, and political interests," League of United Latin American Citizens v. Midland Indep. School Dist., 812 F.2d 1494, 1501 (5th Cir.1987), then elections with a viable candidate from this single minority group constitute elections with a "viable minority candidate" within the meaning of Gretna.

This panel is not free to examine afresh whether Hispanic and black voters ought to be treated as a single minority group under § 2 of the Voting Rights Act. Both judicial opinions and commentators have suggested that such treatment of different ethnic groups as fungible for the purposes of § 2 is contrary to the purposes of the Voting Rights Act and flies in the face of political reality. See, e.g., Midland I.S.D., 812 F.2d at 1505-07 (Higginbotham, J., dissenting); Katherine I. Butler & Richard Murray, Minority Vote Dilution Suits and the Problem of Two Minority Groups: Can a "Rainbow Coalition" Claim the Protection of the Voting Rights Act?, 21 Pacific L.J. 619, 641-57 (1990).

The law of this circuit, however, is that black and Hispanic voters can form a cohesive minority group under Gingles. Midland I.S.D., 812 F.2d at 1501. We hold that, where black and Hispanic voters are cohesive, elections with Hispanic candidates present black voters with the option of supporting a viable minority candidate within the meaning of Gretna. Therefore, where Hispanics and blacks vote cohesively, it is error not to examine elections with candidates of either group in assessing the success of either black- or Hispanic-preferred candidates.

Plaintiffs contended at oral argument that blacks and Hispanics are cohesive only in those counties where the plaintiffs proceeded on behalf of the combined minority voters in the county. If blacks and Hispanics are cohesive for the purposes of Gingles, however, they are cohesive regardless of the plaintiffs' litigation strategy. Cohesion is a political and social fact, not a matter determined by the plaintiffs' pleadings. Plaintiffs essentially urge that Hispanic and black voters are cohesive only when such cohesion is beneficial to the plaintiffs' proof. Such an approach is illogical and insupportable, if understandable.

The undisputed facts show that blacks and Hispanics are just as cohesive in Harris and Tarrant Counties as in Midland, Ector, and Lubbock Counties: the voting patterns in all five counties were not different in any relevant respects. The district court, therefore, clearly erred in ignoring elections with Hispanic candidates in determining the success of the black-preferred candidate in Harris and Tarrant Counties.

D. Relevance of Small Number of Minority Lawyers

Undisputed evidence shows that in all of the counties, the percentage of minority lawyers was much smaller than the percentage of minority voters. In Ector County, for example, there appear to be six minority lawyers eligible for district judgeships.3 It is true that we have refused "to preclude vote dilution claims where few or no [minority] candidates have sought offices in the challenged electoral system." Westwego Citizens for Better Gov't v. City of Westwego, 872 F.2d 1201, 1208 n. 9 (5th Cir.1989). That holding is a far cry from the conclusion that the number of minority candidates eligible to run has no relevance. Section 2 and the Senate Report instruct us to consider the number of minority candidates elected to office. At the same time, we are instructed to look at the totality of the circumstances, which includes the fact that few minority citizens can run for and be elected to judicial office. Our functional analysis of the electoral system must recognize the impact of limited pools of eligible candidates on the number of minority judges which have resulted. See Southern Christian Leadership Conf. of Ala. v. Evans, 785 F. Supp. 1469, 1476-77 (M.D.Ala.1992).

District judges in Texas must be licensed attorneys. Tex. Const. art. V, § 7. As a result, relatively few minority citizens are eligible to run for this office. Consider Tarrant County, where black lawyers comprised only 2.4% of eligible attorneys in 1989. From 1985 to 1989, however, between 8% and 13% of Tarrant County district judges were black. In four counties, Dallas, Harris, Tarrant, and Bexar, the percentage of minority judges exceeds the percentage of minority lawyers. The absence of eligible candidates goes a long way in explaining the absence of minority judges. Plaintiffs here cannot emphasize the scarcity of successful minority candidates to support their dilution claims and simultaneously urge that the number of minorities eligible to run is not relevant. The district court erred by dismissing the significance of the paucity of minority lawyers.

We now turn to the application of these principles of law in each county. The district court's findings of dilution were based upon erroneous legal principles and cannot be relied upon. Considered in light of the controlling standards, the evidence admits of only one conclusion in each county: no violation of § 2. Therefore, we reverse. We reach our conclusions, however, by different routes in different counties. In six counties (Dallas, Tarrant, Travis, Midland, Ector, and Lubbock), we find that the district court erred in finding proof of dilution, regardless of Texas' linkage interest. In four of those counties (Dallas, Ector, Midland, and Lubbock), the overwhelming evidence showed that partisan affiliation, not race, explained the results. In the three remaining counties (Harris, Bexar, and Jefferson), we assume that the evidence may have sufficed to show dilution absent the state's substantial interest. The proof of dilution, however, was so insubstantial that it was outweighed by Texas' linkage interest as a matter of law.

One thread runs throughout the plaintiffs' case in all of the counties--the insubstantiality of the plaintiffs' proof that the minority-preferred candidate lost "on account of race." Except in Dallas County, the district court's finding of such racial dilution was based on only three findings. These were according to the district court, (1) proof of the Gingles factors; (2) general history of discrimination in the counties; and (3) underrepresentation of minorities in the state judiciary of each county. This evidence was supplemented by proof of some "enhancing" Zimmer factors (large districts, anti-single shot voting, etc.). In Dallas County, the district court also found two instances of racial appeals.

As we explain in more detail below, this evidence of dilution by itself is slender, barely amounting to a proof that any minority-preferred candidate lost on account of race. We are persuaded that it is outweighed by Texas' linkage interest.

The voting age population of Dallas County is 1,106,757. Of this number, 180,294 (16.3%) are black, and 90,966 (8.2%) are Hispanic. Plaintiffs proceed on behalf of the black voters in Dallas County. In 1989 there were 36 district court judges in Dallas County. Until 1987, there were no black district court judges. In 1987 and 1988, there were three black district court judges or 8.3% of the total and in 1989, there were two or 5.5% of the total. There were also two black county court judges elected at-large in Dallas County. The defendants' undisputed expert testimony and surveys showed that black lawyers made up at most 2.2% of the lawyers residing in Dallas County.

The undisputed expert evidence demonstrated that 99%-100% of black Dallas voters support the Democratic candidate in every judicial election. The evidence also indicated that the majority of the white voters always voted for the Republican, and thus for candidates other than the black-preferred Democratic candidate.

As a result of these voting patterns, the black-preferred Democratic candidate always lost in judicial elections, regardless of the year of the election in Dallas County. The Republican Party dominated every analyzed judicial race. Defendants understandably contend that the defeat of black-preferred candidates is the result of partisan affiliation rather candidates' race. According to defendants, elections are determined by straight-party voting in which voters support their party's ticket regardless of the race of the candidates. The undisputed facts overwhelmingly support this assertion.

Taebel and Engstrom analyzed seven district court general elections with black candidates. The following is a summary of the election results as they appear in the record. The underlined candidate is the black candidate. The first figure represents the non-black vote as estimated by plaintiffs' expert. The range is defined by the homogenous precinct and bi-variate regression analysis performed by Engstrom. The second figure, in parentheses, represents the white vote as estimated by Taebel. Taebel did not analyze the 1984 Tinsley-Maloney race.

Year Election NB vote (white vote) 1980 Winn (Dem.) 38.6"39.7% (36%) Howell (Rep.) -------------------------------------------- 1984 Baraka (Rep.) 60.6"61.8% (61%) Metcalfe (Dem.) -------------------------------------------- " " Tinsley (Dem.) 28.7"30% Maloney (Rep.) -------------------------------------------- " " White (Dem.) 30.6"31.9% (31%) O'Donnell (Rep.) -------------------------------------------- 1986 Tinsley (Dem.) 36.6"37.5% (31%) Kendall (Rep.) -------------------------------------------- " " Wright (Rep.) 70.6"71.7% (77%) Brin (Dem.) -------------------------------------------- 1988 Oliver (Dem.) 36.9"37.9% (38%) Brown (Rep.) --------------------------------------------

Roughly 61%-77% of white voters consistently support Republicans, even when black Republicans run against white Democrats. Virtually all black voters supported the Democratic candidate, even when the Democratic candidate was white, running against black Republicans.

Moreover, white support for Republicans remained constant, even when the Republican was black. Black Republicans won in two of the seven analyzed district court races. According to Taebel's study, one of these Republicans, Carolyn Wright, did better among white voters than any other Republican, white or black, winning 77% of the white vote. Even Engstrom's study shows Wright doing better among white voters than most white Republicans, receiving between 70% and 72% of the white vote. Other black Republicans received percentages of the white vote comparable to those received by white Republicans. Judge Baraka, the other black Republican district court candidate, took about 61% of the white vote against a white Democrat. County Judge Brashear, a black Republican, took 66% of the white vote in his successful race for a county court judgeship against a white Democrat.

Just as black Republicans did as well as or better than white Republicans, so too, black Democrats won as large a percentage of the white vote as white Democrats. The white vote for Democratic candidates ranged between 23% to 39%. According to plaintiffs' exhibits, black Democrat Oliver won about 38%--a larger than average share of the white vote for a Democrat. Winn, another black Democrat, received as much as 39% of the white vote. By comparison, white Democrat Brin received no more than 29% of the white vote when running against Wright, a black Republican.

Republican candidates lost the black vote and won the white vote regardless of their public positions on matters related to race. Judge Carolyn Wright, for instance, had been a member of the Dallas Chapter of the Coalition of 100 Black Women, served as a legal intern for the Lawyer's Committee on Civil Rights, a project related to civil rights in South Africa, and was a charter member and past vice-chair of the National Political Congress of Black Women. By contrast, the record is silent regarding the record of Wright's opponent, Brin. He nevertheless won the black vote handily in the general election.

Professor Anthony Champagne, defendants' expert witness, testified that this voting pattern was the result of straight-ticket voting. According to Dr. Champagne, judicial elections are low-profile elections in which the voters know little more about the candidates than what they read on the ballot. The voters, therefore, will make their choice based on the information that the ballot contains--which means party affiliation. Because a majority of voters in Dallas are Republican, Republicans tend to prevail in most of the races.

Engstrom did not control for partisan affiliation in his study of Dallas county voting patterns. The district court also made no finding that partisan affiliation did not determine election outcomes, but instead held that the cause of voting patterns was irrelevant under Gingles.

We are persuaded that the undisputed facts indicate that the defeat of the black-preferred candidate was the result of the voters' partisan affiliation, not the result of race. The black-preferred candidate was always the Democratic candidate, while the majority of white voters always supported the Republican candidate. Although between about 30%-40% of the white voters supported the Democratic candidate, the combination of black and white Democratic votes was insufficient to cause the Democratic candidate to prevail. Black and white Democratic voters alike were, therefore, unable to elect any Democratic judicial candidates. Plaintiffs have therefore failed to establish racial bloc voting as required by Gingles.

As we noted above, the race of the candidate did not affect the pattern. White voters' support of black Republican candidates was equal to or greater than the support that white voters gave to white Republicans. Likewise, black and white Democratic candidates received equal percentages of the white vote. Given these facts, we cannot see how the minority-preferred candidate was defeated "on account of race or color." Rather, the minority-preferred candidates were consistently defeated because they ran as members of the weaker of two partisan organizations.

The plaintiffs contend that the Democratic Party better represents the political views of black voters in Dallas County. This contention, if true, is not relevant to whether the minority-preferred candidate is defeated on account of race. To the extent that black voters' preferred candidates are consistently defeated because of their substantive political positions, they are simply the casualties of interest group politics, not racial considerations. This is not the harm against which § 2 protects. Section 2 protects black voters against defeat of the minority-preferred candidate on account of race or color, not on account of political platform. See Whitcomb, 403 U.S. at 154-55 [91 S.Ct. at 1874-75]. We do not deny that political positions can be proxies for racial prejudice. See Gingles, [478 U.S. at 66-67] 106 S. Ct. at 2774. However, where white voters support black candidates of a particular party in larger percentages than they support white candidates of the same party, there is no basis, without more, for us to conclude that the parties' political positions are proxies for racial bias.

Even assuming arguendo that plaintiffs have met the Gingles threshold by showing bloc voting, the totality of circumstances in the record cannot support a § 2 violation. Plaintiff-intervenors Oliver, White, and Tinsley contend that "race considerations pervade elections in Dallas County." They support this proposition with the district court's finding that there were two instances of overt or subtle racial appeals in Dallas County elections. In one, Larry Baraka was called a "Black Muslim" by his opponent. In another, Vance, a candidate for district attorney, printed his opponent's picture in campaign literature, thus informing the electorate that Vance was running against a black opponent. Nothing in the district court's opinion indicates that these racial appeals were anything more than isolated incidents. In the only judicial election affected by a racial appeal, the black candidate, Baraka, won both the Republican primary and the general election, winning a majority of the white vote in both elections.

Oliver, Tinsley, and White also contended at trial that voting patterns in non-partisan elections show that partisan affiliation could not explain the defeats of black-preferred candidates. They produced data from seven Dallas City Council elections, one Democratic Presidential primary, and four ballot issues to support this contention. The district court made no findings about this data, and the parties do not discuss it in their briefs on appeal.

Assuming arguendo that these high-profile elections had any relevance to voting patterns in low-profile judicial elections, we find that the data presented does not support the plaintiffs' argument. The data shows that the candidate favored in predominantly black areas of Dallas prevails in four out of seven City elections and wins the highest plurality (46%) in the Presidential primary. If this data proves anything, it shows that, when one controls for partisan affiliation, the minority-preferred candidate is not defeated and wins the majority of the white vote in most elections.

All other evidence indicated that racial politics played no role in the minority-preferred candidates' defeat. The district court rejected the suggestion that the Republican Party was a white slating organization. The undisputed testimony shows that any eligible candidate could run as a Republican, and plaintiff-intervenors themselves testified at trial that they had been heavily lobbied by the Republican Party leadership to do so. The plaintiffs presented general evidence of the lingering effects of past discrimination, but no specific evidence of depressed levels of black political participation such as low black voter registration or turn-out.

On the contrary, the minority-preferred candidates ran professional, well-financed campaigns backed by the Democratic Party, a party that, until the late 1970's, had dominated Dallas County judicial races just as completely as the Republican Party now dominates those races. These Democratic candidates lost because Dallas County had shifted from being a county of predominantly Democratic straight-ticket voters to a county of mostly Republican straight-ticket voters.

Plaintiffs made no substantial factual riposte to the overwhelming evidence that election outcomes were the product of partisan affiliation. Rather, plaintiffs' answer was the legal assertion that the effect of partisan affiliation, virtually admitted, was not relevant. Plaintiffs' own Dr. Engstrom conceded that there is "a stronger association between partisan affiliation and success than there is between the race of the candidate and success." Dr. Engstrom asserted that partisan affiliation did not explain all of the voting patterns in Dallas County. However, he conceded that he had no data that black Democrats generally did worse than white Democrats. In fact, the undisputed facts show that, when one controls for party, black candidates did as well as, or better than, white candidates in winning the white vote and elections. Plaintiff-intervenor Judge White conceded that partisan affiliation determined her electoral defeat. She admitted that "if I ran as a Republican ... the likelihood is that I would win."

In short, the undisputed facts show that the defeat of the minority-preferred candidate was the result of the voters' partisan affiliation. The district court erred in finding racial vote dilution.

Harris County elects 59 district court judges at-large. At the time of trial, three were black, three were Hispanic, and the rest were Anglo. Defendants produced uncontested expert testimony and surveys which established that black lawyers make up at most 3.8% of the lawyers residing in Harris County but comprised 5.1% of Harris County's district judges. There was also one black county court judge in Harris County who was elected at-large. According to plaintiffs' evidence, 1,685,024 people of voting age reside in Harris County; 305,986, 18.2%, are black, and 222,662, 13.2%, are Hispanic. Plaintiffs claim to represent all black voters in Harris County.

Defendants' expert, Taebel, studied Harris County judicial elections between 1980 and 1988 with either a black or Hispanic candidate. Taebel's study covered 24 district court elections, 9 county court elections, one court of appeals election, one Supreme Court election and ten primary elections. Plaintiffs' expert, Engstrom, studied 17 district court elections with a black candidate in Harris County. All but two of Engstrom's elections were included in Taebel's study. Neither Taebel nor Engstrom analyzed all of the judicial races in which blacks or Hispanics ran. There was evidence in defendant Judge Sharolyn Wood's exhibits that minority candidates participated in 42 races between 1980 and 1988, including six omitted by Taebel and Engstrom.

Engstrom studied 17 judicial races with black candidates between 1980 and 1988. The black-preferred candidate won three, about 17.65%. County court judges also run in county-wide races. The jurisdiction of county court is also county-wide but is limited to smaller cases than the listed courts. If we include races for county court judgeships with black candidates, races that Engstrom apparently ignored, the black-preferred candidate prevailed in five out of twenty-three races--21.74%.

Curiously, plaintiffs' study ignores judicial races with Hispanic candidates. Taebel's study includes some of these races, and Judge Sharolyn Wood's exhibits include all of them. In the judicial races with black or Hispanic candidates, the black-preferred candidate won 15 of 37 races or about 40.5% of the time. Eliminating the two exogenous races for Supreme Court posts by Gonzalez in 1986 and 1988, the minority-preferred candidate still prevailed 37.143% of the time.

These percentages do not fully reflect the success of minority-preferred candidates. The black-preferred candidate was sometimes neither black nor Hispanic. Partisan affiliation always trumped race in predicting which candidates would be supported by white voters. For instance, when black Republican Mamie Proctor was defeated by white Democrat Schuble in her 1986 race for a district court judgeship, Proctor won the majority of the white vote, but lost the vast majority of the black vote to Schuble, the black-preferred candidate. Likewise, when Cheryl Irvin, another black Republican, ran against Duncan, a white Democrat, for a county court judgeship, she won the white vote, but Duncan received virtually all the black vote. Judge Kenneth Hoyt, now a United States District Court Judge, won the white vote and the election (as well as the endorsement of the Houston Lawyers' Association) but lost virtually all of the black vote in his race as a black Republican against a white Democratic opponent for a position on the state appellate bench.

In 1980, 1984, and 1988, years in which popular Republican presidential candidates ran, Republicans defeated virtually all Democrats and, thus, black-preferred candidates. In 1982 and 1986, however, the black-preferred candidates won eight out of the twelve races in which black candidates participated. In races with a Hispanic candidate, the black-preferred candidate won 22 out of 36 races. The black-preferred candidates enjoyed consistent success of 20% or 40% victory, looking respectively to races with black candidates only and elections with both Hispanic and black candidates. The black-preferred candidate was the Democratic candidate in every election in every county. Defendants buttressed these facts with expert testimony that the defeat of both white and black Democrats was the result of straight-ticket voting. According to defendants, both white and black judicial candidates, whether Democratic or Republican, fared according to the popularity of ticket leaders. Defendants contended that the race of the candidate was irrelevant to the outcome of judicial elections. The black-preferred, that is Democratic, judicial candidate would win when the more visible Democratic candidates were strong and would do poorly when the Republican ticket was headed by a popular presidential candidate.

Undisputed data in the parties' studies provides substantial support for defendants' theory. The black-preferred Democratic candidates' success varied with the year of the election. According to the data, in 1980, 1984, and 1988, when the Republican ticket was headed by popular presidential candidates, all but one Republican judicial candidate, including one black Republican, won, and Democrats lost. In 1982 and 1986, when either Governor Mark White or Senator Lloyd Bentsen headed the Democratic ticket, the success at the top of the ballot carried down to judicial races more marked by anonymity than name identity.

Defendants further note that the Republican candidate always won the white vote and the Democratic candidate won the black vote, regardless of that candidate's race or the race of his opponent. Republicans always won a majority of the white vote, generally taking between 55% and 65% of the white vote regardless of whether the Republican candidate was black, Hispanic, or white. Similarly, Democratic candidates always took virtually all of the black and Hispanic vote, even when a white Democratic candidate ran against a black or Hispanic Republican.

It does not follow from this data, however, that black candidates did as well as white candidates, even if one controls for partisan affiliation. According to Judge Sharolyn Wood's exhibits, of 22 black Democrats only three were elected, about 13.64%. By contrast, white Democrats won four out of five, 80%, of their races. Indeed, again according to Judge Wood's exhibits, four out of six black Democratic candidates, Berry, Fitch, Fisher, and Lee, lost in 1986, a year in which all other Democrats won. The undisputed testimony also indicated that Francis Williams, a black Democrat not included in Judge Wood's exhibits, also lost his race in 1986. All of the defeated black Democrats were incumbent judges. In short, black Democratic candidates did not fair as well as white Democratic candidates in the elections studied.

The district court made no findings about the cause of the defeat of the black or black-preferred candidates for judicial posts. Instead, relying on Justice Brennan's opinion in Gingles, the district court incorrectly held that the actual cause of the defeats of the black-preferred and black candidates was legally irrelevant in a § 2 vote dilution case.

Nonetheless, we will assume arguendo that the district court could infer dilution from the evidence of elections in Harris County. We will also assume that, absent Texas' substantial interest in its at-large system, such dilution might create liability under § 2. The question remains whether such assumed proof of dilution is sufficient to overcome Texas' interest in linking electoral and jurisdictional bases of trial court judges.

The district court found dilution based from the three Gingles factors, two primary Zimmer factors, and three "enhancing" Zimmer factors. The primary Zimmer factors were (1) the general history and lingering socio-economic effects of past discrimination and (2) the underrepresentation of minorities in the judiciary. The enhancing Zimmer factors were (1) the large size of Harris County; (2) the prevention of single shot voting by numbered post election; and (3) the majority runoff requirement in primary elections.

Assuming that this proof sufficed to show dilution absent Texas' substantial interest, we find the proof to be marginal. The undisputed facts show that a majority of white voters invariably supported black Republican candidates and that black voters could repeatedly elect their preferred candidates even when opposed by a majority of white voters. In the absence of any substantial evidence of racial politics, any possible inference that the minority-preferred candidate lost on account of race would be tenuous. As a matter of law, Texas' linkage interest must outweigh such a tenuous case.

First and most significantly, white voters repeatedly support black Republican candidates, showing that the defeat of minority-preferred candidates was substantially, although perhaps not entirely, caused by partisan affiliation. Kenneth Hoyt, Mamie Proctor, and Cheryl Irvin, black Republicans, all won a majority of the white vote; Proctor and Irwin lost elections because black voters gave virtually all of their votes to the white Democratic candidates. In the Democratic primaries where party affiliation plays no part, blacks actually did better than white candidates, winning 70% of their contested races.

In addition to evidence of partisan affiliation, minority-preferred candidates had substantial success in winning elections, even when they were opposed by a majority of white voters. The undisputed facts showed that minority voters elected the candidate of their choice in 13 out of 37 analyzed non-exogenous elections in which either black or Hispanic candidates participated--37.14% of the time.

Minority voters could, therefore, repeatedly elect candidates of their choice, even when opposed by a majority of white voters. Far from being submerged in a white majority, black voters were a potent electoral force that could form coalitions with minorities of white voters to elect their preferred candidates. This ability to form coalitions and influence the elections of all judges in Harris County would be lost in the system of single-member districts proposed by the plaintiffs. Instead, black voters would receive the right to elect nine judges, abdicate any right to vote for the balance of fifty, and thus radically reduce the chances of having their disputes decided by a judge over whom they had any influence. As we stated, absent clear evidence that the at-large vote is meaningless because of racial politics, such a result is perverse.

The remaining evidence adds little to plaintiffs' claims of illegal vote dilution. The district court cited these Zimmer factors: (1) the general history and lingering effects of discrimination in Harris County, (2) the underrepresentation of blacks on the Harris County bench, and (3) the consistent defeat of the minority-preferred candidate by the majority of white voters.

The lingering effects of past discrimination can "exacerbate the effects of a racially or ethnically polarized electoral process" by reducing minority registration, voting, and general political participation. Terrazas, 581 F. Supp. at 1346 n. 26. See also White v. Regester, 412 U.S. 755, 768 [93 S. Ct. 2332, 2340, 37 L. Ed. 2d 314] (1973); Kirksey v. Board of Supervisors, 554 F.2d 139, 145 (5th Cir.1977) (en banc). However, while past discrimination exacerbates racial politics otherwise proven, it is at best weak circumstantial evidence of racial politics. The legislative history of the 1982 amendments indicates that such past discrimination by itself does not support an inference of dilution. See S.Rep. 417 at 34 ("mere existence of underrepresentation plus a history of dual schools" is insufficient to show dilution under White/ Zimmer test), reprinted in 1982 U.S.Code Cong. & Admin.News at 212. Whitcomb also suggests that evidence of black poverty alone, even when combined with polarized voting, does not prove dilution: in Whitcomb, the Supreme Court did not dispute the three-judge panel's finding that black residents in the so-called "ghetto" area suffered from poor housing, high unemployment, and other burdens of poverty. Whitcomb, 403 U.S. at 132 [91 S. Ct. at 1863].

Moreover, if lingering effects of past discrimination in combination with polarized voting could prove dilution, then proof of the three Gingles factors would virtually always suffice to establish § 2 liability, because there are few if any communities in the United States that do not still suffer from the economic effects of racism. Allowing proof of the effects of past discrimination alone to prove dilution under the totality of circumstances would make a mockery of the "totality of the circumstances" inquiry. Far from being intensely local, such inquiry would involve no more than judicial notice of the obvious.

In Harris County, the plaintiffs presented no evidence that general black poverty has reduced levels of black registration or black voters' turnout. Indeed, there is no record evidence that black voters' registration or turnout is lower than white registration or turnout. We do not presume that past discrimination causes lower turnout or registration of minority voters. S.Rep. 417 at 29 n. 114 ("The courts have recognized that disproportionate educational, employment, income level and living conditions arising from past discrimination tend to depress minority political participation. Where these conditions are shown, and where the level of black participation in politics is depressed, plaintiffs need not prove any further causal nexus between their disparate socio-economic status and the depressed level of political participation" (emphasis added)), reprinted in 1982 U.S.Code Cong. & Admin.News at 207 n. 114; see also McIntosh Cty. NAACP v. City of Darien, 605 F.2d 753, 759 (5th Cir.1979). Given the lack of any evidence that a smaller percentage of black voters register and vote than white voters, plaintiffs have failed to show even that past discrimination has deprived black voters of equal access to the political system.

Likewise, the representation of blacks on the Harris County bench cannot support an inference of racial politics. Three blacks were district court judges in 1989--5.1% of the total. By contrast, black attorneys make up at most only 3.8% of the attorneys in Harris County. The fact that blacks occupied a smaller percentage of the bench than the black voters' percentage of Harris County's population, therefore, is not surprising: if judges were chosen at random from the pool of eligible candidates, there would be fewer black judges on the Harris County bench.

Aside from the number of blacks on the Harris County bench and the general history of discrimination, the district court found three Zimmer enhancing factors. See Nevett v. Sides, 571 F.2d 209, 218 (5th Cir.1978). Such factors enhance the opportunity of a white majority to engage in racial politics. However, they do not "meaningfully advance the inquiry into whether race is at issue," Terrazas, 581 F. Supp. at 1346 n. 26, and therefore cannot support an inference of racial politics.

The circumstantial evidence of a relation between defeats and race is, at best, tenuous, given the willingness of white voters to support black Republican candidates and substantial success enjoyed by minority-preferred candidates. We express no opinion as to whether this minimal proof of dilution shows a violation of § 2 absent Texas' linkage interest. Assuming without deciding that it does, we find that plaintiffs' proof at best produces only a marginal case in Harris County too insubstantial to outweigh Texas' linkage interest, as a matter of law.

There were 23 district courts in Tarrant County in 1989. In 1985, 1986, 1987, and 1988, three of these judges (over 13%) were black. In 1989, two district court judges were black (8.7%). The defendants' undisputed testimony and surveys indicated that only 2.4% of the Tarrant County Bar is black. There are 613,698 residents of voting age in Tarrant County. Of this number, 63,851 (10.4%) are black. Plaintiffs proceed on behalf of black voters in Tarrant County.

The evidence indicated that blacks voted cohesively for the Democratic candidate. Brischetto, plaintiffs' expert, analyzed four elections, three of which were general elections for district judgeships and one of which was the Democratic presidential primary of 1988. In all four elections, the regression estimates show that from 85% to 100% of Tarrant County blacks voted for the black-preferred and Democratic candidate. Taebel's analysis is substantially the same.

Taebel analyzed nine general elections, including three exogenous elections, in which a black or Hispanic had participated. Brischetto analyzed only those elections in which a black candidate had participated--three general elections for state district court judge and one Presidential primary. As in all other counties, the evidence showed that black voters voted cohesively for Democratic candidates. Unlike other counties, however, black judges made up more than 13% of the Tarrant County bench for four out of five years--a proportion of the bench that is greater than the proportion of black voters in the County's population.

The success of the black-preferred candidate was also greater in Tarrant County than in other counties. In those general elections with black candidates, the black-preferred candidate, always the Democratic candidate, won only one out of three general elections--33.3% of the studied races. However, in nine general elections with Hispanic candidates included in Taebel's study, the black-preferred candidate won four out of nine, or 44.4% of the elections. In the six non-exogenous district and county court elections included in Taebel's study, the black-preferred candidate won three out of six, or 50% of the elections.

Brischetto includes the 1988 Democratic presidential primary in which Jesse Jackson won virtually all of the black vote in Tarrant County but received between 14% to 16% of the white vote.4 Taebel's study also includes a 1986 Democratic primary for state district court in which Ross, the black-preferred candidate, received 57% of the black vote but lost the white vote and thus came in third out of a field of four candidates. Brischetto ignores the two Democratic primaries in which black-preferred Hispanic candidates prevailed.5

In short, the evidence shows that the black-preferred candidate won 40% of the Democratic primaries and half of the non-exogenous judicial elections, including elections with Hispanic candidates. The record also shows that black judges have consistently made up a greater proportion of the Tarrant County bench than the proportion of black voters in the Tarrant County's population.

Finally, the undisputed evidence showed that black candidates won as great a share of white votes as white candidates, if we control for party affiliation. For instance, Sturns, a black Republican with a long history of involvement in civil rights and black community organizations, won 57% of the white vote to beat a white Democrat. Wayne Salvant, another black Republican, also won a majority of the white vote, although he lost his race for a district court judgeship to a white Democrat supported by a combination of black voters, Hispanic voters, and white Democrats. Black Republicans also won the same share (50%) of their elections as white Republicans.

A prima facie case of vote dilution requires proof that the white bloc majority is "usually ... able to defeat candidates supported by a politically cohesive, geographically insular minority group." Gingles, [478 U.S. at 49] 106 S. Ct. at 2766 (emphasis in original). The plaintiffs failed to prove that the minority-preferred candidate was usually defeated in Tarrant County and therefore, failed to prove a prima facie case of vote dilution.

The district court, by contrast, found that the black-preferred candidate was consistently defeated in Tarrant County. The district court reached this conclusion by ignoring elections in which Hispanics had participated. This rejection of white-Hispanic elections was erroneous. The undisputed facts as reflected by Taebel's exhibits are that a majority of Hispanic voters always supported the candidate favored by black voters in every general election. The district court found that Hispanic and black voters were cohesive in Midland, Lubbock, and Ector counties on similar evidence. With virtually identical proof in Tarrant County the same conclusion must follow, and we hold that it does.

Given that black and Hispanic voters cohesively shared political interests, Hispanic candidates were capable of representing these shared interests. Hispanic candidates, in short, represented "viable minority candidates" within the meaning of Gretna. It follows that white-Hispanic elections were relevant in determining the success of black-preferred candidates. Considering such elections, it is clear that the black-preferred candidates were not consistently defeated. Considering white-Hispanic races, the black-preferred candidate won four out of nine elections--44.4% of the time. Excluding the three state-wide exogenous elections for Supreme Court and Attorney General, the black-preferred candidate won three out of six elections--or 50% of the time. Exogenous elections are generally less probative of the local communities' political culture and are therefore less probative of polarized voting than non-exogenous elections. 834 F.2d at 502. This is not the "consistent defeat" required by Gingles.

Moreover, blacks were not underrepresented on the Tarrant County bench. Plaintiffs' own exhibit indicates that, for four of the five years studied, three of Tarrant County's district court judges were black; for these four years, while blacks made up only 11.8% of the population of Tarrant County, more than 13% of the Tarrant County bench was black. Given this persistent and substantial black presence on the Tarrant County bench, the consistent and substantial success of the minority-preferred candidate, and the absence of any evidence of racial politics in Tarrant County, we find that, even if the plaintiffs had proven the three Gingles factors, the district court clearly erred in finding illegal vote dilution under the totality of the circumstances. There is no case as a matter of law in Tarrant County.

There are 13 district judges elected in Travis County. From 1985 to 1988, there was one Hispanic district judge, or 7.7% of the total number. This judge was defeated in 1988. The undisputed expert testimony and surveys of defendants showed that Hispanic lawyers made up at most 3.9% of the lawyers in Travis County. There are 312,392 voting age residents in Travis County. 44,847 have Spanish surnames. Only 29,067 are black. Plaintiffs proceed on behalf of Hispanic voters in Travis County.

Plaintiffs' witnesses stated that the Republican Party is insignificant in Travis County and the proper testing ground for candidates is the Democratic primary. Plaintiffs analyzed three Democratic primary elections for district court positions. Defendants analyzed four exogenous general elections, one for state senator, two for Supreme Court justice, and one for Attorney General; four exogenous primary elections, one for state senator, one for Supreme Court justice, and two for appellate court justice; and the three indigenous elections analyzed by plaintiffs.

By Taebel's analysis, the Hispanic-preferred candidate won all four exogenous general elections. In three of the four, the Hispanic-preferred candidate won a majority of the Anglo vote. The Hispanic-preferred candidate also won two exogenous primaries, for Supreme Court and state senator, out of the seven primaries. The Hispanic-preferred candidate, therefore, won 54.55% of both the primaries and general elections studied, giving the exogenous primary and general election races for intermediate appellate courts and the Supreme Court the same weight given to the three entirely indigenous primary elections studied by plaintiffs.

The district court however, found that the three indigenous primary elections for district court positions were "closer in nature to District Court elections" and were sufficient to show a pattern of racial bloc voting sufficient to defeat the Hispanic-preferred candidate. The district court therefore relied solely on the three elections analyzed by both Taebel and Brischetto to find that the Hispanic-preferred candidate lost 100% of the time.

In each of the three district or county court primary elections analyzed by the plaintiffs, the Hispanic and Hispanic-preferred candidate was defeated by a white majority. In one of these races, however, white voters gave their support to a black candidate and thereby defeated both a Hispanic and a white candidate. The black candidate, Brenda Kennedy, had the overwhelming support of black as well as white voters, so it is difficult to conclude that Celia Castro and Robert Hughes, the Hispanic and white candidates respectively, were defeated by a white bloc. They were defeated by a black-white coalition. Castro's defeat is not evidence of the white majority's ability "usually to defeat the minority's preferred candidate." Gingles, [478 U.S. at 51], 106 S. Ct. at 2766-67.

The two remaining primary elections are too meager a proof to support a finding of liability, even if we disregard Texas' linkage interest. It was clear error for the district court to find otherwise. Plaintiffs' proof reduces to three facts: (1) two Hispanic and Hispanic-preferred candidates lost in 1988 in their Democratic primary races for district and county court judgeships; (2) only one Hispanic has served on the district court bench between 1985 and 1988, while no Hispanic served in 1989; and (3) Hispanics had suffered from past discrimination in Travis County.

In finding clear error, we repeat Justice Brennan's admonition that "the usual predictability of the majority's success distinguishes structural dilution from the mere loss of an occasional election." Gingles, [478 U.S. at 51], 106 S. Ct. at 2767. It defies common sense to believe that the loss of two races in one year constitutes usual and predictable defeat by a white bloc, rather than simply "loss of an occasional election." However, assuming arguendo that these two elections constituted sufficient proof of the third Gingles factor, they are too meager to show dilution under the totality of the circumstances, as a matter of law. The plaintiffs contend that Hispanics are underrepresented on the Travis County bench. Hispanics make up 7.7% of the Travis County bench in four out of five years, while they make up, at most, 3.9% of the Travis County bar. Given such a small pool of eligible candidates, it is not surprising that Hispanics make up such a small proportion of the Travis County bench. One need not assume racial bias among voters to explain such a result.

The cause of the low number of Hispanic district judges in Travis County need not be attributed to the interaction of racial bias with the at-large system. Rather, it is equally likely such low numbers were the result of a dearth of eligible Hispanic candidates. The plaintiffs can point to only one race in which a Hispanic candidate lost an election for district court--Gallardo's 1988 defeat by Scott McCown. However, even if Gallardo had won this race, the Hispanic share of Travis County's bench would be only 2 out of 13--15.4%, or 2% less than the number of Hispanics in the population. Gallardo's victory, of course, would not affect Hispanic representation in 1985, 1986, and 1987. It is pure speculation to attribute the low number of Hispanic judges to racial discrimination, given the low number of eligible candidates. To find dilution on the basis of these low numbers is clear error.

Far from signaling a submerging of minority voting strength by an intervention of elective processes and bias, the undisputed facts indicate that Travis County's political system is open to Hispanic and white candidates alike. Hispanics won half of the four exogenous primary elections for Supreme Court, Court of Appeals, and State Senator produced by the defendants. The Hispanic-preferred candidate also won all four of the general elections produced by the defendants. The City of Austin contains most of Travis County's population. As this court noted in Overton v. City of Austin, 871 F.2d 529, 540 (5th Cir.1989),

Austin has repeatedly elected black and Mexican-American council members during the past 17 years.... [T]he winning minority candidates frequently received well over fifty percent (50%) of the Anglo vote and were also the preferred candidates of the minorities. Minority candidates have routinely been elected to other posts in Austin and the surrounding Travis County.

The defendants produced uncontradicted evidence that Hispanic County Commissioners had been elected from predominantly Anglo districts and won Anglo precincts and that John Trevino, a Hispanic City Council member, had been elected in city-wide elections to the Austin City Council. Against this background of minority success, plaintiffs' minimal case, based on two defeats of the Hispanic-preferred candidate by a white majority, is plainly insufficient to prove illegal vote dilution.

There was no evidence that racial politics defeated the two Hispanic-preferred candidates, Juan Gallardo and Alberto Garcia. The district court found no Zimmer factors indicating that race played any part in any election, such as racial appeals or unresponsiveness of elected officials to minority constituents. The only Zimmer factors found by the district court concerned the absence of single-shot voting and majority runoff requirement. Their only relevance is a tendency to enhance the effect of racial politics, not to prove racial politics' existence.

672,220 voting age residents reside in Bexar County. Of these, 46,767 (7.0%) are black, and 278,577 (41.4%) are Hispanic. Nineteen district judges are elected from Bexar County. Of this number, five (26.3%) were Hispanic. Defendants' undisputed evidence shows that 11.7% of the lawyers in Bexar County were Hispanic. Plaintiffs proceed on behalf of Hispanic voters in Bexar County.

Plaintiffs and defendants analyzed six district court general elections with Hispanic candidates. Defendants also analyzed two appellate court and three county court elections with either Hispanic or black candidates. As in every other county, Hispanics voted cohesively for the Democratic candidate while Anglos supported the Republican candidate.

In the twelve studied elections, the Hispanic-preferred Democratic candidate won four times, 33.3%. The Republican candidate usually won the general election, and always won the Anglo vote, regardless of the candidate. There were four exceptions to this pattern: (1) the 1980 appellate court race between Murry and Esquivel; (2) the 1980 district court race between Prado and Priest; (3) the 1988 district court race between Bowles and Mireles; and (4) the 1988 county court race between Patterson and Canales. In these races, Priest, an Anglo Democrat, beat Prado, a Hispanic Republican, while Esquivel, Mireles, and Canales, Hispanic Democrats, beat their Anglo Republican opponents.

We find, however, that partisan affiliation does not completely explain the voting patterns in the Democratic primary elections. By defendants' own evidence of Democratic primaries in Bexar County, the Hispanic-preferred candidate lost in nine of fourteen elections, prevailing only 35.7% of the time, when Anglo voters voted for the Hispanic candidate's Anglo opponent. White support for the Hispanic candidate was seldom above 30% and as low as 1%--whereas the Hispanic vote for the Hispanic-preferred, and always Hispanic, candidate was above 70% for all but four of the unsuccessful candidates.

Although the evidence may create a fact-issue as to whether there was dilution, there is no fact-issue as to whether the proof of dilution was tenuous. As in Harris County, the undisputed facts indicated Anglo voters would invariably support minority candidates of their preferred party and that minority-preferred candidates could prevail even when a majority of Anglo voters opposed them. Given these facts and a complete absence of substantial evidence of racial politics, any proof of dilution was meager at best and therefore could not overcome Texas' linkage interest as a matter of law.

The undisputed facts indicate that partisan affiliation accounts for much of the voting patterns analyzed by the parties. Most white voters are Republicans, most Hispanic voters are Democrats; the Republican candidate generally wins because the Republican party has more supporters. As in Harris County, white voters give a majority of their votes to Republicans and Hispanic voters give a majority of their votes to Democrats even when Hispanic Republicans are opposed by white Democrats. Prado and Barrera, Hispanic Republicans, won 70% and 84% of the estimated white vote respectively when running against white Democratic opponents, who received the overwhelming majority of the Hispanic vote.

Because Hispanic voters make up 41% of the population, they can elect Democratic candidates with minimal white support and do so repeatedly. The minority-preferred candidate won four out of twelve elections in which a Hispanic candidate participated--33% of the time--with as little as 17% of the white vote. As in Harris County, Hispanic voters were plainly a potent political force that could elect candidates by forming coalitions with small percentages of white voters. If Bexar County were sub-districted, Hispanic voters might elect a few more of their preferred candidates, but only at the price of losing their influence over the majority of Bexar County judges. The perversity of such a result is self-evident.

Finally, the evidence that elections were affected by racial politics preventing the formation of such coalitions is thin. As in Harris County, it consisted solely of (1) the general history of past discrimination; (2) the usual enhancing factors present in every Texas county--anti-single shot voting and majority runoff requirement; and (3) the fact that Hispanic judges occupied only five out of nineteen district judgeships--26% of the total--when Hispanics made up 41% of Bexar County's population. Again, we note that Hispanic attorneys made up only 11% of the bar, so that the representation of attorneys on the bench is actually higher than would be produced by random choice from the pool of eligible candidates. This evidence, even if probative of racial vote dilution, is as meager as the evidence in Harris County.

The lingering effects of past discrimination, including evidence of low Hispanic voting registration, would exacerbate any racial politics in Bexar County. However, as we noted regarding Harris County, such evidence is, at best, weak circumstantial evidence that the minority-preferred candidate was defeated on account of race-conscious politics. Noting the effects of past discrimination, therefore, does little to strengthen plaintiffs' proof in the area where it is weakest.

The undisputed facts compel the conclusion that whatever dilution could have been found by the district court was marginal at best and cannot as a matter of law outweigh Texas' substantial state interest. If Texas' linkage interest does not outweigh this proof of dilution, then the interest would be a nullity. We hold that plaintiffs' proof fails in Bexar County as a matter of law.

There are 8 district judges elected in Jefferson County. No black Judge has been elected there between 1985 and 1989. Defendants' evidence showed that 3.1% of the attorneys in Jefferson County are black. Depending on the survey examined, the evidence showed that between 14 and 17 black lawyers reside in Jefferson County. The Jefferson County district court judges have filed an amicus brief requesting judicial notice that John Paul Davis, a black, was elected to the county court in 1990. The amicus brief also notes that Morris Overstreet, a black Democrat, and Dan Morales, a Hispanic Democrat, won a majority of votes in Jefferson County for state appellate court judge and Attorney General respectively in 1990. 179,708 people reside in Jefferson County who are of voting age. Of this number, 44,283 (24.6%) are black. Plaintiffs proceed on behalf of black voters in Jefferson County.

Brischetto analyzed eight primary and run-off elections including the 1988 Democratic Presidential Primary. Taebel analyzed four primaries and two general elections involving either blacks or Hispanics, all exogenous elections. Unlike their other studies, Brischetto and Taebel analyze totally different elections.

In all but one of the elections analyzed by Brischetto, the black vote was cohesive. In one case, the black-preferred candidate won a high plurality (47%) of the black vote. A majority of white voters always opposed the minority-preferred candidate in the primary elections.

Whether the black-preferred candidate was consistently defeated by the white bloc is a close question. The answer varies with the elections counted and how they are counted. Defendants point to two exogenous primaries (one for Supreme Court, and one for state appellate court) and two primaries for state representative in which black or Hispanics participated. In three of these, the black-preferred candidate prevailed. Defendants also rely on two exogenous general elections, for Supreme Court and Attorney General, in which Hispanic candidates participated in which both of the black-preferred candidates prevailed. Plaintiffs offer six indigenous primaries, four for justice of the peace,6 one for county court, and one for U.S. President, in which black candidates participated. The black-preferred candidate prevailed only once, when Jesse Jackson won a plurality in the Democratic Presidential primary of 1988.

Unlike Tarrant County, defendants' exhibits do not include estimates of how Hispanic residents in Jefferson County voted. We cannot find from the undisputed facts that Hispanic and black voters were politically cohesive in Jefferson County. It is possible, therefore, that white-Hispanic elections are entitled to less weight in a determination of the minority-preferred candidate's success than black-white races.

However, even if we confine our consideration to the elections analyzed by the parties in which black candidates participated, we find that the plaintiffs' evidence is inadequate to prove that black voters were denied an equal opportunity to participate in the political process so as to overcome the state's linkage interest. The plaintiffs and defendants together produced evidence of eight primary elections in which black and black-preferred candidates had participated. The black-preferred candidate won 3 primaries out of these eight elections--a victory rate of about 38%. All three of the black-preferred candidates' victories were exogenous: Jesse Jackson won the 1988 Democratic Presidential primary in Jefferson County, while Price won two Democratic primaries for state representative.

As in every other county but Dallas, the district court found no sign of racial appeals or non-responsiveness on the part of elected officials to the concerns of black constituents. The statistics indicated that the white vote did not monolithically throw its power against the black-preferred candidate. Price won the Democratic nomination for state representative with over 40% of the white vote in his two primary races. The minority-preferred candidate also prevailed in every general election submitted by the parties.

The plaintiffs' case was further weakened by their use of extremely dated statistics: three of their five elections came from elections held in 1972, 1974, and 1978. This is hardly a practical and searching appraisal of contemporary conditions in Jefferson County. See Nipper v. Chiles, 795 F. Supp. 1525, 1540 (M.D.Fla.1992) (noting limited probative force of "stale" elections).

There are three district judges in Midland County; none are Hispanic or black. Defendants' undisputed survey evidence shows that 10 black or Hispanic lawyers reside in Midland County. The County contains 82,636 voting age residents, 6,893 (11.9%) of whom have Spanish surnames and 4,484 (7.8%) of whom are black. Plaintiffs proceed on behalf of both Hispanic and black voters in Midland County.

Plaintiffs analyze three general elections in Midland County, two of which are the exogenous Gonzalez races for Supreme Court and one of which is an indigenous race by a black candidate for a Justice of the Peace position in 1986. Defendants also analyze Gonzalez's two bids for the Supreme Court in 1986 and 1988, as well as two Democratic and two Republican primaries in which either a black or Hispanic candidate participated. Defendants also analyze the Mattox-Barrera race for Texas Attorney General.

The analysis of both parties shows that the majority of whites always opposed the candidate preferred by the geographically compact and cohesive combined minority population in the general elections. The minority-preferred candidate was always defeated by this white majority.

We find that the district court clearly erred in finding dilution. The undisputed facts indicate that partisan affiliation, not race, caused the defeat of the minority-preferred candidate. The majority of minority voters always cast their votes in favor of the Democratic candidate. The white voters cast the majority of their votes for the Republican, regardless of the race of the Republican candidate. Indeed, Barrera, the Hispanic Republican candidate for Attorney General, won 76% of the white vote when running against Mattox, a white Democrat--the second highest vote received by any of the four Republicans who ran in the analyzed general elections. Because Republican voters outnumbered Democratic voters, the minority-preferred Democratic candidate consistently lost. The plaintiffs have not met the third requirement of Gingles.

Even if plaintiffs could meet the Gingles threshold, the totality of circumstances does not add up to dilution. The plaintiffs can show only a general history of discrimination and a lack of minority judges. The latter fact proves little: in Midland County, only one minority lawyer has ever run for local office, according to the plaintiffs' exhibits, and none has ever run for a district court position. These low numbers are a reflection of low numbers of eligible candidates. According to the defendants' undisputed surveys, there were only ten minority lawyers residing in Midland County in 1989. As we have noted above, a general history of discrimination, without more, cannot show dilution.

Because the undisputed facts show that partisan affiliation uninfected by racial politics caused the minority-preferred candidates' defeat, we hold that the district court erred in finding dilution.H. Lubbock Counties

There are five district judges elected from Lubbock County. None are black or Hispanic. The surveys introduced by the defendants indicate that 22 or 23 black or Hispanic lawyers reside in Lubbock County. The total voting age population is 150,714 people. Of this number, 22,934 (15.2%) have Spanish surnames, and 9,5090 (6.4%) are black, yielding a combined minority percentage of 21.6%. Plaintiffs proceed on behalf of the combined Hispanic and black voters in Lubbock County.

None of the parties analyze indigenous elections in Lubbock County; no minority has ever run for a position on the district court. Plaintiffs analyze two exogenous primaries and two exogenous general elections, for the Supreme Court and for the court of criminal appeals. Defendants analyze the same two general elections, adding a further exogenous general election for Attorney General of Texas.

Plaintiffs' and defendants' evidence showed that blacks and Hispanics tend to vote cohesively. There is also no dispute that the majority of white voters vote against the candidate favored by the minority voters in Lubbock County in all the elections studied.

As in Midland County, however, the undisputed facts show that, in general elections, partisan affiliation and not racial politics caused the consistent defeat of the minority-preferred candidate. The data indicated that, in both counties, 60%-65% of the white voters supported the Republican candidate, while the majority of minority voters supported the Democratic candidate. As a result of this voting pattern, the minority-preferred, and always Democratic, candidate consistently lost to his Republican opponent, regardless of the ethnicity of the candidates.

In the 1986 and 1988 races for the Texas Supreme Court, Gonzales, Hispanic Democrat, lost Lubbock County's vote to Bates and Howell, white Republican opponents. However, in his race for the position of Texas Attorney General, Mattox, a white Democrat, lost to Barrera, a Hispanic Republican. Like Howell and Bates, Barrera took a majority of the white votes, while his white opponent, Mattox, took a majority of the minority votes. In short, the defendants established as a matter of law that voting patterns in these two counties were unaffected by the race of the candidates but rather were the result of partisan loyalty. Therefore, plaintiffs have not met the third Gingles factor.

Additionally, the totality of circumstances cannot support dilution. As in Midland County, the countervailing evidence of racial politics was too slender to create a fact issue. Although there were no minority judges on the Lubbock County bench, no minority candidate ever ran for such a position. The dearth of minority candidates, in turn, is not probative of race-conscious politics in light of the defendants' undisputed estimate that only 23 minority lawyers reside in Lubbock County.

The plaintiff also relied on two exogenous Democratic primary elections for state appellate and Supreme Court positions.7 However, the minority-preferred candidate won a majority of the votes cast in one of the two elections included in the plaintiffs' exhibits. Balancing Martinez's defeat, Gonzales won a majority of the votes cast in the Lubbock County Democratic primary for state Supreme Court. These primary races, therefore, do not indicate that the minority-preferred candidate was consistently defeated within the meaning of Gingles, and they cannot establish dilution.

There are four district judges in Ector County. All are Anglo. Defendants' surveys estimated that six black and Hispanic attorneys reside in Ector County. Ector County has 79,516 voting age residents. 14,147 (17.8%) are Hispanic, while 3,255 (4.1%) are black. Plaintiffs proceed on behalf of the combined minority population in Ector County.

The parties rely on the same exogenous races in Ector County that they produced in Lubbock. The plaintiffs rely on two primary and two general elections of Gonzalez and Martinez for Supreme and appellate court respectively. The defendants add the Mattox-Barrera race for Attorney General.

The undisputed facts indicate that the minority-preferred, Democratic candidates were consistently defeated in general elections by a white majority voting for their Republican opponents. The undisputed facts also indicate that the minority-preferred candidate won half of the Democratic primary races and therefore was not consistently defeated in the primaries: Martinez won a majority of the votes cast in the Democratic primary.

As in Lubbock County on virtually identical facts, we find that the district court clearly erred in finding racial vote dilution. The undisputed facts indicate that partisan affiliation controlled the outcomes of the general elections: as in Lubbock County, while Hispanic Democratic candidates lost the white vote, Barrera, a Hispanic Republican, won a majority of the white vote running against Mattox, his white Democratic opponent.

While partisan affiliation would not explain polarization in the Ector County Democratic primaries, the undisputed facts indicate that the minority-preferred candidate was not consistently defeated by racial polarization in the primary elections, but rather won half of the two races analyzed. Plaintiffs have failed to meet the third threshold requirement of Gingles.

We find that the district court erred by holding that the at-large method of electing judges violated § 2 in any of the nine counties. Defendants prevail by two routes. Either the evidence was insufficient to support a conclusion of vote dilution, or the proof of dilution was so meager as to be outweighed by the linkage interest as a matter of law.

There was no evidence that white voters refused to support black candidates. In all counties, white voters supported the minority candidate of their preferred party. There was no evidence that minority candidates could not be elected in any county. In all counties, minority candidates were elected with support from the white community. In the one alleged racial incident in a judicial race, the minority-preferred candidate, a black, won with 61% of the white vote. Any racial appeal was rejected.

In short, in the totality of the circumstances, plaintiffs produced no substantial evidence to override the state's substantial interest. Plaintiffs failed to prove a dysfunction traceable to race or color in the political climate of the contested counties. Absent such dysfunction, there is little indication that minority voters could not influence elections through the normal mechanisms of interest-group democracy despite the defeats of the minority-preferred candidate. Ending the county-wide election of district judges would only reduce minority influence by denying minority voters any say in the election of a judge that by large odds will preside over any dispute of their own or their family. This perversity coupled with the problematic proof of any dilution on account of race leaves the weighing of totality of the circumstances so plain as to permit only one outcome.

REVERSED.

ORDER

Feb. 11, 1993.

Before POLITZ, Chief Judge, KING, GARWOOD, JOLLY, HIGGINBOTHAM, DAVIS, JONES, SMITH, DUHE, WIENER, BARKSDALE, and DeMOSS, Circuit Judges.*

BY THE COURT:

A majority of the Judges in active service, on the Court's own motion, having determined to have this case reheard en banc,

IT IS ORDERED that this cause shall be reheard by the Court en banc with oral argument on a date hereafter to be fixed.