Jeffers v. Clinton

730 F.Supp. 196 (1989)

M.C. JEFFERS, Al Porter, Evangeline Brown, Clyde Collins, O.C. Duffy, Earl Foster, the Rev. Elihue Gaylord, Shirley M. Harvell, Linda Shelby, J.C. Jeffries, Lavester McDonald, Joseph Perry, Clinton Richardson, T.E. Patterson, Earnest Simpson, Brian Smith, and Charlie Statewright, on Behalf of Themselves and All Others Similarly Situated, Plaintiffs,

v.

Bill CLINTON, in his Official Capacity as Governor of Arkansas and Chairman of the Arkansas Board of Apportionment; W.J. McCuen, in his Official Capacity as Secretary of State of Arkansas and Member of the Arkansas Board of Apportionment; and Steve Clark, in his Official Capacity as Attorney General of Arkansas and Member of the Arkansas Board of Apportionment, Defendants.

No. H-C-89-004.

United States District Court, E.D. Arkansas, E.D.

December 4, 1989.

First Dissenting Opinion December 7, 1989.

Second Concurring and Dissenting Opinion January 26, 1990.

*197 P.A. Hollingsworth, Little Rock, Ark., L.T. Simes, II, West Helena, Ark., Kathleen *198 Bell, Helena, Ark., Olly Neal, Jr., Marianna, Ark., Don E. Glover, Dermott, Ark., Penda D. Hair, Washington, D.C., Donna L. Dennis, New York City, Julius L. Chambers, Dayna L. Cunningham and Sherrilyn Ifill, NAACP Legal Defense Fund, Inc., New York City, and Sheila Y. Thomas, NAACP Legal Defense Inc., Washington, D.C., for plaintiffs.

Frank J. Wills, III, Atty. General's Office, Little Rock, Ark., for defendants.

Before ARNOLD, Circuit Judge, EISELE, Chief District Judge, and HOWARD, District Judge.

First Dissenting Opinion of Judge Eisele December 7, 1989.

Second Concurring and Dissenting Opinion of Judge Eisele January 26, 1990.

ARNOLD, Circuit Judge. [*]

In 1981 the Arkansas Board of Apportionment, consisting of the then Governor, Secretary of State, and Attorney General, placed into effect a plan of apportionment for the General Assembly. Seventeen black electors [1] bring this suit, claiming the plan violates Section 2 of the Voting Rights Act of 1965, as amended, 42 U.S.C. § 1973 et seq., and the Fourteenth and Fifteenth Amendments to the Constitution of the United States. They ask us to hold the existing arrangement of Senate and House districts unlawful, order a new plan into effect for the 1990 elections, and place the State of Arkansas under the preclearance procedure laid out in Section 3(c) of the Voting Rights Act, 42 U.S.C. § 1973a(c).

We heard evidence for twelve days. Numerous exhibits are before us. We have carefully considered the proof with due regard to the intensely practical nature of the political process. We now hold that the plaintiffs have demonstrated a violation of their rights under federal law. The 1981 apportionment plan created only five legislative positions, one in the Senate and four in the House, representing districts in which a majority of the voting-age population was black. We find that a total of 16 such districts, three in the Senate and 13 [2] in the House, could have been created, and that these districts would have been reasonably contiguous and compact. We further find that voting in the areas of the State in question is markedly polarized by race. Both black and white voters usually prefer candidates of their own race. Black voters are far from powerless. They exercise significant, sometimes decisive influence. But they can elect a candidate of their choice, in a district in which the voting-age population is majority white, only if that candidate is white. For the foreseeable future, the present location of legislative district lines will make it very difficult to elect more than six black legislators, out of a total in both houses of 135 members. And this is so even though black people make up about 16 per cent. of the total population of the State of Arkansas.

In this situation, black citizens have less opportunity than other members of the electorate to elect representatives of their choice. This is a violation of Section 2(b) of the Voting Rights Act, 42 U.S.C. § 1973 (b). For reasons we shall explain in this opinion, plaintiffs have proved a violation of the law in all of the areas of the State called in question by this suit, except for Pulaski County. We hold that no more elections may be held under the unlawful 1981 apportionment plan. The defendants will be enjoined from giving any further force or effect to that plan. A new, lawful plan must be drafted, in time to be in place for next year's elections; the filing period for these elections will begin on the third Tuesday in March — March 20, 1990. See Ark. Code Ann. § 7-7-203 (c) (1987). The parties are directed to submit plans for compliance to this Court on or before January 15, 1990. We especially emphasize the duty of the *199 defendants, the present members of the Board of Apportionment, to submit a plan. The responsibility of complying with the law is primarily theirs. Promptly after January 15, 1990, we will convene an evidentiary hearing on the remedy, if necessary, and thereafter enter an order embodying a new plan and directing that it be used for the 1990 legislative elections.

The opinion we file today decides only plaintiffs' statutory claim under Section 2 of the Voting Rights Act. Plaintiffs' constitutional claim and their request for preclearance as a remedy for constitutional violations remain under advisement. We will decide these questions in a second written opinion in due course. Our ruling on the statutory claim will require substantial adjustments in the existing apportionment plan. If relief is to be effective for next year's elections, time is of the essence. We therefore think it proper to file this opinion, which will start the process of re-drawing the lines, promptly. We need time, however, to consider the hard issues of intent and remedy that plaintiffs' constitutional claim raises, and we do not wish to delay the whole case while we take that time.

I.

The Board of Apportionment is created by Article VIII, Section 1 of the Constitution of Arkansas. Its present members are Governor Bill Clinton, Secretary of State Bill McCuen, and Attorney General Steve Clark. They are defendants in this case. The Board's members in 1981 were Governor Frank White, Secretary of State Paul Riviere, and Attorney General Clark. The plan presently in effect (except as modified as to Crittenden County by Smith v. Clinton, supra ) was adopted on June 28, 1981, by a vote of two to one. Governor White dissented. The plan created 35 single-member districts for the Senate. Of the 100 members of the House, the plan called for 74 to be elected from single-member districts and 26 from multi-member districts. [3] (We note, parenthetically, that only two of the multi-member districts in the House are in the area of the State towards which this suit is primarily addressed: the two-member district in Crittenden County, which no longer exists because of Smith, and a three-member district in Pulaski County. This three-member district has a majority-black voting-age population, and all three of its State Representatives are black. This lawsuit, however, is not primarily about single-member versus multi-member districts. It is about dilution of black voting strength. We see no reason why the multi-member districts in other parts of the State would be affected by the relief we are granting.)

Under the plan adopted in 1981, only two House districts had a voting-age population that was majority black — one of the three-member districts in Pulaski County and a single-member district in Jefferson County. (We use voting-age-population (VAP) percentages because they are the numbers relevant for purposes of voting, which is what this case is about. Under the one-person, one-vote principle, of course, which requires that legislative districts be substantially equal [4] in population, it is total population that counts, regardless of age or eligibility to vote. But this case is about effective use of the elective franchise, not equality of population. Black VAP numbers run consistently lower than total-population numbers in the areas affected by this *200 suit.) We insert at this point in the opinion a map of Arkansas (Map 1) showing each of the House districts in the affected areas, including black VAPs for districts located along the Mississippi River (an area referred to as the Delta).

The shaded areas indicate concentrations of black population. It is important to note that although these districts are located in the area of the State containing the greatest concentration of black people, none of them, as the 1981 plan was drawn, had a *201 majority-black VAP. The Senate plan follows the same pattern. Of the 35 single-member Senate districts, only one has a majority-black VAP. We insert at this point Map 2, which shows the Senate districts. No black person has ever been elected to the State Legislature (in this century, anyway) from a district that did not have a majority-black VAP.

II.

Before we get to the substance of the case, we must deal with defendants' argument that the suit is barred by laches, [5] that it was filed too late. The apportionment plan challenged was adopted in 1981, but suit was not filed until January 22, 1989. Only one election, that of 1990, remains *202 to be run before the State must be reapportioned anyway, because of the 1990 Census. So this case, defendants say, should be dismissed. If plaintiffs are still unhappy after the 1991 reapportionment, they can file suit then. This Court rejected a similar contention in Smith v. Clinton, 687 F.Supp. at 1312-13. But defendants earnestly press the argument and claim that Smith can be distinguished. We therefore address the point at some length.

Laches is an equitable defense. It means essentially this: if suit is unreasonably delayed, and this delay causes prejudice to the defendants, a court of equity may dismiss the complaint. The greater the delay, or the more unreasonable, the less prejudice need be shown; and vice versa. The Court must weigh the facts and interests on both sides, summon up the discretion of a chancellor, remember that it is a court of conscience and not of legal stricture, and come as close as it can to a fair result. Frequently there are some good arguments on both sides, and that is the case here. We start by quoting our discussion of the issue in Smith v. Clinton :

First, the injury alleged by the plaintiffs is continuing, suffered anew each time a State Representative election is held under the [illegal] structure. Second, there have been significant developments since the 1981 Arkansas reapportionment. The Voting Rights Act was amended in 1982, and the Supreme Court's interpretation of the statute in Thornburg v. Gingles, 478 U.S. 30 , 106 S.Ct. 2752 , 92 L.Ed.2d 25 (1986), lays down a rather uncompromising structure for the application of the law in vote-dilution cases. Third, we note that the plaintiffs are required to prove that, as a result of the challenged structure, a white majority bloc is usually able to defeat the preferred candidates of the minority. Of course, evidence of this circumstance would be unavailable unless the structure had been in place for some time.

687 F.Supp. at 1313.

All of these factors are present here, but defendants claim Smith was different. See Defendants' Post-Trial Brief, 2 n. 1. There, we dealt only with the dilutive effect of one two-member district. The relief affected only that district. There was no ripple effect. No other district's boundaries were disturbed. See Smith, 687 F.Supp. at 1313 n. 4. But here there would be some ripple effect. Plaintiffs claim that there should be as many as 16 (13 in the House, 3 in the Senate) majority-black districts. Relief cannot be accomplished simply by dividing up one or more existing multi-member districts. If the single-member districts requested, or any of them, are created, the boundaries of some adjacent districts will necessarily shift to some extent. Plaintiffs assert that the ripple effect will be confined to 26 of the State's 75 counties, all of them south and east of a diagonal line roughly splitting the State from northeast to southwest, but even as so limited the necessary changes would be substantially greater than those occasioned by Smith.

With this argument in mind, we compare plaintiffs' delay to defendants' prejudice. It is reasonable to start the clock running on June 30, 1986, when the Supreme Court handed down Thornburg v. Gingles, 478 U.S. 30 , 106 S.Ct. 2752 , 92 L.Ed.2d 25 (1986). This case is the pole star of the law in this area. Only with this decision can the law be said to have jelled. The present plaintiffs moved for leave to intervene in Smith v. Clinton on June 1, 1988, less than two years later. This motion was denied on July 1, 1988. The order denying the motion suggested plaintiffs could bring a separate suit, which they did about six months later. Suit was filed about 14 months before the filing deadline for the 1990 elections. (Compare Smith, 687 F.Supp. at 1312, in which candidates had filed and begun to campaign in the challenged district at the time of the hearing on plaintiffs' motion for preliminary injunction.) This sort of case takes an enormous amount of preparation, and it is to plaintiffs' credit that they took time to prepare it thoroughly before coming to court. This sequence of events does contain some delay, but most if not all of it was reasonable.

What about prejudice to defendants? As noted, the expense, trouble, and disruption *203 of compliance (assuming, as we must at this stage of the analysis, that the complaint has some merit) will be substantial, and more so than in Smith. But the expense, trouble, and disruption are not a consequence of plaintiffs' delay in filing. They would have occurred whenever the suit was filed — even if it had been filed, say, right after passage of the 1982 amendments to the Voting Rights Act. There is some additional increment of public confusion that will be caused by changing district lines less than two months before filing opens, and less than four months before the first primary election. It is also true that the census data relied on by both sides — and that presumably will be the basis of any remedial plan — become increasingly stale as time passes after the 1980 census. But it is by no means clear that these data, however inaccurate they may become in terms of absolute numbers, will also be inaccurate in relative terms. On the contrary, what evidence we have on the subject indicates that, although population as a whole has declined in the affected areas of the State, the proportions of blacks and whites have remained virtually constant. See Testimony of Dr. Engstrom, Dr. Morrison, Senator Benham, Representative McGinnis, and Representative Cunningham. And in any event the staleness of the 1980 census data cannot be escaped. For even under defendants' theory, which is that the case should be dismissed, the 1990 elections will be run in districts based on 1980 census data. The true comparison is between out-of-date districts that (by hypothesis) dilute the black vote, and out-of-date districts that do not.

The question is essentially one of judgment and degree. Logic cannot absolutely exclude either answer. In our judgment, the defense of laches must fail. In part, the expense and disruption that will undeniably occur are nothing but a consequence of the wrong that has been done. The illegality, the injury, extend beyond a single district, and so the remedy that must be applied will necessarily be more trouble than it was in Smith. To the extent that electoral confusion and disruption exceed what they would have been if the case had been filed earlier, we think that fairness and equal opportunity in voting are worth it. We will not say to these plaintiffs, "Wait for another census. The time is not yet ripe." They have heard these words too many times in the past.

III.

We turn now to the merits of plaintiffs' Section 2 claim. As amended in 1982, the statute now reads as follows:

§ 1973. Denial or abridgement of right to vote on account of race or color through voting qualifications or prerequisites; establishment of violation

(a) No voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner 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, or in contravention of the guarantees set forth in section 1973b(f)(2) of this title, as provided in subsection (b) of this section.

(b) A violation of subsection (a) of this section is established if, based on the totality of circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by members of a class of citizens protected by subsection (a) of this section in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered: Provided, That nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.

The statute embodies a "results" test. A voting practice or procedure (like an apportionment plan) violates the law if it results in denial or abridgement of the right to vote on account of race or color. *204 The law focuses on effects, not purpose or motivation. Subsection (b) adds a further direction to the courts: we are to consider "the totality of circumstances," which is lawyer's language for "all the facts," and decide whether members of a protected class "have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice."

A.

At the outset, defendants make two legal arguments that would bar the action altogether if successful. First, they say that under the plain language of Section 2(b) plaintiffs must show two separate things: (1) that they have less opportunity to participate in the political process; and (2) that they have less opportunity to elect representatives of their choice. Even if they have shown the second, the argument runs, they cannot win this case, because they cannot make the first showing. There are no presently existing legal barriers to voting by black citizens in Arkansas, and therefore they have just as much opportunity to participate in the political process as anyone else.

This argument fails to reckon with the present effects of past racial discrimination, much of it official and governmental. In Smith v. Clinton , we made the following findings:

(1) The Court takes judicial notice that there is a history of racial discrimination in the electoral process in Arkansas. See Perkins v. City of West Helena, 675 F.2d 201, 211 (8th Cir.), aff'd mem., 459 U.S. 801 , 103 S.Ct. 33 , 74 L.Ed.2d 47 (1982). We do not believe that this history of discrimination, which affects the exercise of the right to vote in all elections under state law, must be proved anew in each case under the Voting Rights Act.

(2) We further find that the history of discrimination has adversely affected opportunities for black citizens in health, education, and employment. The hangover from this history necessarily inhibits full participation in the political process.

687 F.Supp. at 1317 (footnote omitted). By order entered on October 2, 1989, the first day of the trial in this case, we made the same findings.

Moreover, the argument fails purely as a logical and linguistic matter. Even if plaintiffs failed to show less opportunity to participate in the political process, a showing that they have less opportunity to elect candidates of their choice would suffice to establish their claim. The right protected is the aggregate of these opportunities — the right to effective participation in the political system: "[t]he essence of a § 2 claim is that a certain electoral law, practice, or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by the black and white voters to elect their preferred representatives." Thornburg, 478 U.S. at 47 , 106 S.Ct. at 2764 . An example from a less exalted field of human endeavor will illustrate the point. Suppose you say that I have less ability to chip and putt than you do. If I am just as good a chipper as you are, but not so good at putting, this statement, as a matter of ordinary speech, is still a true one. It is a combination of qualities (play around the green) that we are discussing, and the comparison is between our respective totals or aggregates of these qualities. So in the context of Section 2: it is a combination of abilities (abilities to use the elective franchise) that we are comparing. If I can vote at will but never elect anyone, my political ability is less than yours. Elections, and winning them, are the whole point of voting. This is, at any rate, one reading of the statute that is grammatically available, and the statute should be construed liberally in favor of its object, which is to open up the electoral process to full participation.

Second, defendants also argue that Section 2 just doesn't apply at all to cases involving single-member districts. Both Thornburg and Smith, they say, were multi-member district cases. Single-member districts submerge no one. The majority wins, but that's inherent in any democratic system. We disagree with this reading of the law. Nothing in the words of the statute *205 supports it. Section 2 says nothing about any kind of districts. It is directed against certain described effects, regardless of the form of the political device or mechanism that produces them. Thornburg , to be sure, is a multi-member district case (in the Supreme Court, that is), [6] but it implies that a particular configuration of single-member districts can give rise to a vote-dilution claim, see 478 U.S. at 50 n. 16, 106 S.Ct. at 2766 n. 16, and in principle there is no reason why this should not be so. Accord, Neal v. Coleburn, 689 F.Supp. 1426 (E.D.Va.1988) (county single-member districts dilute black voting strength in violation of Section 2). Defendants cite Butts v. City of New York, 779 F.2d 141, 148 (2d Cir.1985), a pre- Thornburg opinion, for the proposition that vote-dilution claims won't lie with respect to single-member districts, but we do not so read the case. It holds that a particular device — the requirement of a run-off if no candidate gets 40% of the vote — does not violate Section 2 in a particular electoral unit choosing a single official. It does not hold that unlawful dilution is impossible when a multi-member body's members are chosen from single-member districts that have been drawn in such a way as to split a certain class of voters among different districts.

We agree that Thornburg and Smith cannot be automatically applied to the single-member context. Dilution may be much more obvious in a case like Smith, where a potential majority of black voters was submerged in a two-member district. But the basic principle is the same. If lines are drawn that limit the number of majority-black single-member districts, and reasonably compact and contiguous majority-black districts could have been drawn, and if racial cohesiveness in voting is so great that, as a practical matter, black voters' preferences for black candidates are frustrated by this system of apportionment, the outlines of a Section 2 theory are made out. Whether such a claim will succeed depends on the particular factual context, including all of the factors that Thornburg, Smith, and the legislative history of Section 2 say are relevant.

B.

We first address the major Thornburg factors, factors that must be present if the predicate of a vote-dilution claim is to be laid. In evaluating a Section 2 claim, the Court must first determine whether three preconditions to a legally substantial impairment of plaintiffs' ability to elect the candidates of their choice have been met. First, the claimant minority must establish that it is sufficiently large and geographically compact to constitute a majority in a single-member district. Second, the plaintiffs must show that the group to which they belong is politically cohesive. Thornburg, 478 U.S. at 50-51 , 106 S.Ct. at 2766-67 . Third, the minority voters must show that the "majority votes sufficiently as a block to enable it — in the absence of special circumstances, such as a minority candidate running unopposed ... to defeat the minority's preferred candidate." Id. at 51 , 106 S.Ct. at 2766 (footnote omitted). The latter two factors can be shown by proving that voting in the jurisdiction is highly racially polarized. Ibid. Accord, Smith, 687 F.Supp. at 1314-1315.

In the first place, we find that black communities in the areas of the State challenged by plaintiffs are sufficiently large and geographically compact to constitute a majority in single-member districts. This finding is based primarily on the testimony of plaintiffs' expert, Jerry Wilson. As already noted, this testimony, taken together with the details set out in Mr. Wilson's report, PX 5, establishes that the Board of Apportionment could have drawn, in 1981, thirteen majority-black House districts and three majority-black Senate districts. These districts could have been drawn in the areas of the State which are the focus *206 of this suit. For the convenience of the reader, we insert at this point in the opinion Maps 3 and 4, showing Mr. Wilson's alternative districts for the House and Senate, respectively. [7]

*207

As the maps indicate, some of the districts look rather strange, but we do not believe this is fatal to plaintiffs' position. Their alternative districts are not materially stranger in shape than at least some of the districts contained in the present apportionment plan. The one-person, one-vote rule inevitably requires that county lines and natural barriers be crossed in some instances, and that cities and other political and geographic units be split in others.

Defendants question the compactness and contiguity of plaintiffs' alternative districts, but, with perhaps one exception, all of these questions have been resolved to our satisfaction. Initially, defendants took the position that some of the alternative districts put forward by plaintiffs contained "islands," that is, areas completely surrounded by the alternative district, but not part of it. This criticism, in particular, was leveled against alternative House districts 1, which includes part of the City of *208 Marianna, 11, which includes part of the City of Magnolia, and 10, which includes part of the City of McGehee. An examination of the exhibits and study of the testimony of the opposing experts satisfy us that there are in fact no islands in House districts 1 and 11, as put forward by plaintiffs. We are not sure that we completely understand the geographical situation with respect to hypothetical district 10, but even if there is an island within this district, the problem can be easily cured simply by cutting off the "finger" referred to in Mr. Wilson's rebuttal testimony. See PX 91, 92. If this change is made, the problem of insularity clearly disappears, and does so without violating the one-person, one-vote rule, and without deviating from the goal of creating a majority-black district. We note, incidentally, that the existing apportionment plan indisputably contains islands — Senate district 27, for example, wholly surrounds Senate district 28.

We also have little difficulty in finding that voting patterns are highly racially polarized, in the sense that black and white voters prefer different candidates with a high degree of frequency. Furthermore, the white voting majority is powerful enough, and consistent enough, to defeat black voters' preferences for black candidates almost without exception.

In making this finding we rely primarily on actual events and practical politics. Since 1978 there have been ten races for membership in the State Legislature in which black candidates ran against white candidates in majority-white districts. The black candidate lost every one of these races. Complete election returns for seven of these ten races are in the record before us, and substantial evidence with respect to the other three was also introduced. In all but one of these races, that in which Earl Foster was the black candidate for State Representative in House district 88 in 1984, there was a high correlation between the race of the voter and the level of support for the black candidate. And even in Foster's case, he was the candidate preferred by blacks in his home county, and was not the candidate preferred by whites in any of the counties covered by the district in question. The fact is that there is a strong tendency for white voters to vote for white candidates when there is a black candidate in the race. Black voters behave in exactly the same fashion. This is not a particularly admirable commentary on the voting behavior of either race, but it is a fact in present-day Arkansas politics, and there is no reason to suppose that it will change substantially in the near future. Evidence concerning so-called exogenous elections, that is, elections for positions other than membership in the State Legislature, also supports this conclusion.

Both sides spent a good deal of time and energy debating the merits of various statistical methods. The plaintiffs' expert, Dr. Richard Engstrom, used three such methods: single regression, double regression, and homogeneous-precinct analysis. The defendants' expert, Dr. John Wildgen, objected particularly to double-regression analysis, claiming that it could show nothing about the behavior of individual voters. Yet, both the Supreme Court, Thornburg, 478 U.S. at 53-61 , 106 S.Ct. at 2767-72 , and this Court, Smith, 687 F.Supp. at 1316-17, have accepted testimony based on bivariate (that is, double) ecological regression and homogeneous-precinct analysis. We recognize, as Dr. Wildgen argues, that these statistical methods cannot predict the behavior of individuals, but we do not believe that plaintiffs make any such claim. Furthermore, exhibit after exhibit was introduced in the form of "scattergrams," in which the votes for black and white candidates are plotted for various precincts. To our untrained eye, the cumulative effect of these exhibits is overwhelming, whatever the technical merits or demerits of the various statistical theories. And our own experience as citizens of this State, which we are not required to lay aside, strongly confirms this conclusion.

There are some exceptions to this pattern of racially polarized voting. Judge Edwin Keaton, Municipal Judge in Camden, was re-elected in 1988 against a single white opponent, but he failed to win a majority of the white vote despite being an incumbent. In addition, black candidates for the House *209 have occasionally received a majority of the white vote, but only when running as incumbents in majority-black districts. And even in these races, the proportion of the black vote received by these candidates was much greater than that of the white vote.

Races not involving direct competition between white and black candidates do not lend themselves to such a clear pattern. Where two white candidates are running, for example, black voters can hold the balance of power. We do not wish to minimize this aspect of political reality, but we do not believe it has sufficient weight to negate the clear proof of polarization. The Supreme Court in Thornburg used a statistical analysis based solely on elections in which black and white candidates faced each other, and so did we in Smith. There, we stated that analysis of elections in which there are no black candidates proves only that "[c]andidates favored by blacks can win, but only if the candidates are white." 687 F.Supp. at 1318. White voters, in short, can elect white candidates against black opposition, but black voters cannot elect black candidates against white opposition, with insignificant exceptions. We hope the day will come when this is no longer true, when voters of both races will vote for the person and not for the color of his or her skin. Whatever distaste we may personally have for racial stereotypes in politics, the relevant question for present purposes is the preferences of voters in real life, and we believe they have been clearly established. It is true here, as it was in Smith, see 687 F.Supp. at 1317, that there is racially polarized voting in races for the Arkansas State Legislature, that black voters usually vote cohesively, as a unit, and that white voters have the strength under the present plan of apportionment (except in majority-black districts) to enable them to frustrate the choices made by black voters.

C.

What we have written so far suffices to establish that the plaintiffs have proved the essential predicate for a Section 2 violation. But it does no more than that. We must now examine all of the other relevant factors and decide whether, on balance, a diminution of black political opportunity, in violation of Section 2, has been shown. For this purpose we begin with the list of relevant factors contained in the report of the Senate Judiciary Committee on the bill that became the 1982 amendment to the Voting Rights Act. The relevance of this list is confirmed by the Supreme Court's opinion in Thornburg, 478 U.S. at 44-45 , 106 S.Ct. at 2763-64 . Accord, Smith, 687 F.Supp. at 1314. The list appears in S.Rep. No. 417, 97th Cong., 2d Sess. 28-29, reprinted in 1982 U.S.Code Cong. & Admin.News 177, 205-07, and reads as follows:

1. the extent of any history of official discrimination in the state or political subdivision that touched the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process;

2. the extent to which voting in the elections of the state or political subdivision is racially polarized;

3. the extent to which the state or political subdivision has used unusually large election districts, majority vote requirements, anti-single shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority;

4. if there is a candidate slating process, whether the members of the minority group have been denied access to that process;

5. the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment and health, which hinder their ability to participate effectively in the political process;

6. whether political campaigns have been characterized by overt or subtle racial appeals;

7. the extent to which members of the minority group have been elected to public office in the jurisdiction;

*210 Additional factors that in some cases have had probative value as part of plaintiffs' evidence to establish a violation are;

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;

whether the policy underlying the state or political subdivision's use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.

The second of these factors (racial polarization in voting) has already been fully discussed. We have taken judicial notice of the first (history of official discrimination in voting) and fifth (effects of past discrimination in education, employment, etc.) factors. Accord, Smith, 687 F.Supp. at 1317 & n. 7. The record made in this case, however, is much fuller than the one made in Smith, and we deem it appropriate to summarize some of the details of the proof on these points. We will try to keep the discussion within short compass. Much of the proof in this case was already obvious to any conscious Arkansan, and we do not wish to pile Pelion upon Ossa.

History of Official Discrimination in Voting. In addition to the findings made in Smith and in Perkins v. City of West Helena, 675 F.2d 201 (8th Cir.), aff'd mem., 459 U.S. 801 , 103 S.Ct. 33 , 74 L.Ed.2d 47 (1982) (maintenance of at-large elections for discriminatory purpose), we also call attention to Sherpell v. Humnoke School Dist. No. 5, 619 F.Supp. 670, 680-81 (E.D. Ark.1985), appeal dismissed, 814 F.2d 538 (8th Cir.1987). There, this Court found that the Humnoke School District No. 5 of Lonoke County, Arkansas, maintained an at-large system of electing school-board members for the discriminatory purpose of limiting black political opportunity.

Some of the history relied on by plaintiffs can be dismissed or minimized on the ground that it occurred so long ago that its effects have by now disappeared almost completely. In this category we would place the General Assembly's rejection of the Fourteenth Amendment, the white primary, and the poll tax (abolished by the voters in 1964). But other incidents cannot be so easily disposed of. A number of witnesses testified to the difficulties experienced by blacks in electoral politics in various of the areas affected by this litigation. Polling places have been moved on short notice; deputy voting registrars have, with isolated exceptions, been appointed only as a result of litigation; efforts have been made to intimidate black candidates. It is not necessary to decide at this point which (if any) of these barriers were motivated by invidious discrimination. It is sufficient for present purposes to note that these and similar practices clearly result in discouraging black participation in elections. Partly this is due to a higher level of illiteracy, poverty, economic dependence, even timidity, among the black population. And while defendants can hardly be blamed for creating these conditions, it is an inescapable fact that they are in large part the legacy of a history of discrimination, much of it governmental, beginning with the constitutionally sanctioned institution of human slavery.

We think it proper to refer to one additional instance of racial intimidation, occurring in 1986, in order to make the point that official discrimination designed to suppress black political activity is not wholly a thing of the past, at least not in the Delta. Roy Lewellen, a black lawyer in Marianna, Lee County, ran for the State Senate in 1986 against the white incumbent, Senator Paul Benham. At about the same time, the Sheriff and the Prosecuting Attorney instituted a well-publicized criminal prosecution against Mr. Lewellen for witness bribery. Mr. Lewellen testified in some detail before us. He gave a number of reasons for his belief that the prosecution was designed to discourage him in particular and black political activity in general. We find this testimony entirely credible. Defendants called no witnesses to rebut it. [8] Mr. Lewellen's *211 difficulties were to some degree of his own making. He allowed himself to be drawn into an equivocal situation with a prosecution witness in a criminal case in which he represented the defendant. But this is not the whole explanation for what happened. We do not think that a white lawyer, even one who opposed the political powers that be, would have been treated this way. This kind of intimidation no doubt had a powerful chilling effect. In making this finding, we are consistent with the provisional findings made by this Court, on motion for preliminary injunction, in Lewellen v. Raff, 649 F.Supp. 1229 (E.D. Ark.1986), aff'd, 843 F.2d 1103 (8th Cir.), opinion modified, 851 F.2d 1108 (8th Cir. 1988), cert. denied, ___ U.S. ___, 109 S.Ct. 1171 , 103 L.Ed.2d 229 (1989).

In short, there is a long history of official discrimination. It has a present effect. And some instances of it are still occurring.

Effects of Past Discrimination in Employment, etc. Much has already been said on this subject. See, e.g., Smith, 687 F.Supp. at 1317 n. 7 (income and education data for Crittenden County). Education is probably the key point here. Many more whites than blacks are high-school graduates, and many blacks were educated in schools that were both separate (by compulsion of law) and unequal. There is a tremendous amount of white poverty, especially in the Delta, but poverty among blacks is more nearly the rule than the exception. Blacks tend to have fewer telephones and fewer cars. If a person has no phone, cannot read, and does not own a car, the ability to do almost everything in the modern world, including vote, is severely curtailed. The point, again, is not that defendants produced these conditions. We are convinced that the two defendants who testified before us, Governor Clinton and Attorney General Clark, are determined to alleviate them. But as long as blacks, as a group, remain in a depressed socio-economic status, their political power will necessarily be less, and the impact on them of vote-diluting boundary lines will be greater.

We insert at this point a table setting out some of the economic facts of life for the two races in the 16 counties where this suit challenges district lines.

SOCIOECONOMIC FACTORS FOR BLACKS AND WHITES BY COUNTY [*]

HIGH SCHOOL —PER CAPITA— FAMILIES BELOW NO TELEPHONE [+] NO VEHICLE [+]

GRADUATES [†] INCOME POVERTY LEVEL AVAILABLE AVAILABLE

COUNTY (%) ($) (%) (%) (%)

WHITE BLACK WHITE BLACK WHITE BLACK WHITE BLACK WHITE BLACK

Ashley 60.2 25.2 6,023 2,987 12.2 36.3 8.5 28.4 7.6 30.8

Chicot 53.4 26.4 6,318 2,165 15.3 54.6 10.6 32.9 7.3 43.7

Columbia 60.6 32.6 7,014 2,739 8.2 37.9 6.9 33.0 7.5 28.7

Crittenden 61.4 19.6 6,970 2,293 8.6 51.6 7.3 28.5 4.7 36.3

Cross 48.1 24.4 6,070 2,429 12.1 38.2 13.1 28.1 8.9 30.4

Desha 58.5 25.1 6,909 2,594 10.0 42.0 8.9 32.3 9.2 37. 4

Phillips 58 .3 22.9 6,299 2,336 11.8 53.6 10.9 30.5 9.0 42.0

Pulaski 75.9 54.2 8,188 3,873 5.2 24.8 5.2 14.7 6.2 23. 4

Jefferson 65 .7 41.8 7,084 3,143 8.0 37.5 6.8 18.6 6.3 29.7

Lee 43.7 20.1 5,339 1,923 16.4 59.4 15.9 29.5 10.9 37.4

Lincoln 49.8 24.5 5,353 2,031 NOT AVAILABLE NOT AVAILABLE NOT AVAILABLE

Mississippi 52.3 27.7 5,685 2,426 13.4 45.0 13.5 30.5 9.6 33.6

Monroe 48.8 23.0 5,743 2,173 16.8 50.8 9.6 26.6 10.2 47.6

Nevada 55.1 27.7 5,916 2,859 9.7 43.1 11.7 35.6 7.5 20.5

*212

HIGH SCHOOL —PER CAPITA— FAMILIES BELOW NO TELEPHONE [+] NO VEHICLE [+]

GRADUATES [†] INCOME POVERTY LEVEL AVAILABLE AVAILABLE

COUNTY (%) ($) (%) (%) (%)

WHITE BLACK WHITE BLACK WHITE BLACK WHITE BLACK WHITE BLACK

Ouachita 60.9 32.5 6,648 3,037 6.8 36.5 6.8 16.6 7.6 27.2

St. Francis 57.4 26.0 6,056 2,402 14.7 47.4 13.1 29.2 7.6 33.0

Use of Majority-Vote Requirements and Other Devices. We do not believe that any of the challenged districts is unusually large. Races for the State Legislature require each candidate to run for a designated seat, so "single-shot voting" would have no practical significance. There is a requirement that candidates for the State Legislature get a majority of the vote in the primary to obtain a party nomination. And this majority-vote requirement, in four separate recent instances in which black candidates either won office by a plurality or were threatening to win, has been expanded to cover elections ( e.g., for municipal judge) to which it has traditionally not been applied. We do not stop to determine at this point whether the run-off primary, either in its origin or in its recent extensions, is the product of discriminatory intent. That issue will be addressed when we consider plaintiffs' constitutional claims and their request for preclearance under Section 3(c). For present purposes we simply note the existence of a majority-vote requirement affecting races for the General Assembly and many other public offices in Arkansas.

Candidate slating process. No evidence was introduced on this subject. As far as we know, the process of slating plays no part in races for the Arkansas Legislature. Nominations are made by primary, not (except in rare instances) by committee or convention. Anyone who wants to run simply files as an individual in the primary. If a candidate cannot pay a filing fee, petitions may be used to get on the ballot, and this method of ballot access is now fairly common.

Racial Appeals in Campaigns. Racial appeals, some quite offensive, are common in campaigns in which a white candidate is running against a black candidate. Sometimes simply informing the voters that one's opponent is black seems to be enough to do the trick. Some white candidates have bought newspaper ads or distributed political leaflets containing their black opponent's picture. Sometimes references to race are more explicit. In the Mayor's race in Pine Bluff in 1975, for example, a supporter of a white candidate publicly warned that if white voters didn't turn out, there would be a black mayor. Sometimes coarser words are spoken. We have in mind especially the testimony of Andrew James Willis of McGehee and his brother, Carol Willis, who is now an assistant to Governor Clinton. In 1976 Carol Willis ran for County Judge of Desha County, the first black person ever to do so, in this century anyway. Mr. Willis received obscene phone calls, including racial slurs. One night on his way home, he was run off the road by a group of individuals wearing hoods. And at a public rally his white opponent used profanity and a racial epithet — not in his actual speech, to be sure, but in open conversation. Both of the Messrs. Willis testified to this incident, and we believe them. The defendants offered no evidence to the contrary. After the election, the County (headed by the County Judge who had defeated Carol Willis) stopped or reduced its business with the Willis Funeral Home. The County had been paying the funeral home $100 to bury black paupers. The County even went so far as to pay a "white" funeral home $500 (a premium rate) to bury a black pauper — an event unheard of at the time and probably still unique. We find here a macabre echo of the testimony of R.C. Henry, the longtime (black) Chairman of the Lee County Republican Committee, from 1979 to 1989. To this day, the races live separately, he said, they go to church separately, and they even die separately. And as late as October *213 2 of this year, the City of Marianna was maintaining, at public expense, a cemetery for whites only.

The Extent to Which Blacks Have Been Elected. The statute expressly provides that members of a protected class have no right to be "elected in numbers equal to their proportion to the population." 42 U.S.C. § 1973 (b). But "[t]he extent to which members of a protected class have been elected to office in the State or political subdivision may be considered...." Ibid. This factor points strongly in plaintiffs' favor in the present case. Only in majority-black districts have black candidates been elected to the Arkansas General Assembly. All other black candidates have been defeated. It was not until 1972 that any blacks were elected to the Legislature in this century. Four black members, three in the House and one in the Senate, were elected at that time. Today there are only six black members, and one of them owes his seat to this Court's decree in Smith. No black person has ever won statewide office by election. And in all of the 75 counties of Arkansas, no black person has ever been elected, since Reconstruction, to any county-wide constitutional office. There are increasing numbers of black officials, even in the Delta, but in almost every instance they have been elected in majority-black single-member districts (we refer primarily to city councils, school boards, and quorum courts), and many of these districts owe their existence to civil-rights litigation. We believe, in addition, that many potential black politicians have simply not run. They know that as a practical matter their candidacy would probably be futile.

Lack of Responsiveness of Elected Officials. Plaintiffs claim that white legislators in the Delta are insensitive to the concerns of poor black people. Certainly there is a widespread feeling to this effect among black voters, and plaintiffs presented several witnesses who genuinely feel left out of the political process. These feelings are not without basis. If a person is blindfolded and taken to a given part of town, and the blindfold is then removed, he or she will almost certainly be able to tell whether the residential area is predominantly white or black. Housing is still largely segregated — not by law, but, just as effectively, by choice and economic necessity. Houses in black parts of town tend to be run down, streets and gutters are not as well kept, and there are more open ditches. Certainly this condition is due in part to the unresponsiveness of local government over the years, a situation that, however, is now beginning to change with the election of more black City Council members and Justices of the Peace.

We are not convinced, however, that the charge of unresponsiveness can be sustained as to the members of the State Legislature, and it is with them that we must be particularly concerned in this case. Members of the House like Representatives Cunningham, McGinnis, Flanagin, and Dawson are anything but unresponsive. They are well aware that a large proportion of their constituency is black, and they make assiduous and sincere efforts to represent these voters. There is bound to be some public dissatisfaction, among blacks as well as whites, with almost any office holder, but the charge that white legislators in the Delta are unresponsive to black needs has not been proved to our satisfaction on this record. No doubt some are more responsive than others, but such individual variations in political philosophy and conduct will always be found.

As examples of unresponsiveness, plaintiffs pointed particularly to two pieces of legislation, one that became law and one that did not. In 1987, a bill was passed to impose additional regulations on the use of the absentee ballot. The bill was sponsored by House members from the Delta, and plaintiffs claim that its purpose was to suppress the incipient effective use of the absentee ballot by black voters. There is some evidence to support this claim. Despite a long history of difficulty and fraud with the absentee ballot in Arkansas, it was not until 1987 that the bill in question was passed. (Among other things, the bill allows only close relatives to deliver absentee ballots from voters to the county clerk. In practice, plaintiffs say, this restriction *214 operates against elderly black people, who may not have relatives and are not able to get to the courthouse themselves. The option of mailing the ballot remains, but some poor people cannot afford the postage.) On balance, we find this claim unpersuasive. The legislation was passed overwhelmingly, with votes from all parts of the State, including all of the black members of the House of Representatives. It is facially neutral, and we find plausible the claim that there were real abuses in absentee balloting that needed to be addressed.

Plaintiffs also presented a considerable amount of testimony about H.B. 1130, which was introduced during the 1989 regular session of the General Assembly, but failed to pass in the Senate. This bill, sponsored by Governor Clinton, would have done two principal things: it would have removed from the State income-tax rolls a large number of low-income people, including many black people, and it would have imposed on taxpayers with incomes in excess of $100,000 a year a flat tax rate of 7% on their entire income. Plaintiffs claim that white legislators in the Delta opposed this bill, thus preferring the needs of a few rich constituents to those of thousands of low-income blacks. We do not agree with this claim. Representatives Cunningham and Schexnayder, for example, cosponsored H.B. 1130, and Representatives Dawson and Flanagin and Senator Benham voted for it. Some white legislators from the Delta did oppose the bill, but their motives in doing so, we think, were honestly political. Representative Nancy Balton, for example, testified that middle-class voters in her district asked her to oppose the bill. Nothing in the bill directly injured them, but some of them felt that poor people on welfare were living better than they were, and therefore they opposed anything that would improve the lot of these people. The reasoning is faulty, but the sentiment is all too familiar, and state legislators, like national ones, often vote on the basis of public perception of the effects of proposed legislation, as opposed to the effects in reality. This is a serious problem, but we do not believe it is a racial one.

There is some evidence of unresponsiveness on the part of the Legislature. Some white members, on being approached by black citizens in their own districts for help, referred these constituents to black legislators representing other areas. And black members have found it difficult to get white members to cosponsor some bills of interest to black voters — for example, the bill to create a holiday in honor of Dr. Martin Luther King, Jr. But on the record as a whole, the charge of unresponsiveness has not been proved.

The Strength of Policies Underlying the Plan of Apportionment. We are not quite sure how to apply this criterion in the present case. There was no policy, nor could there be a valid one, explicitly requiring the cracking or splitting of black-majority areas. The Board of Apportionment did profess allegiance to a number of other policies, however, including the undesirability of crossing natural barriers like rivers, the undesirability of splitting cities, a preference for continuity in representation (that is, avoiding situations where incumbents would have to run against each other), a desire to depart as little as possible from the 1971 apportionment plan, and a desire to cross political boundaries, like county lines, as little as possible. There was, in addition, the overriding policy of equality of population, which was probably the uppermost thought in the Board's mind at the time.

Some of these policies were not consistently applied. The cities of Little Rock, Pine Bluff, and El Dorado were split, but each of those cities is so large that it had to be split. It could not have been contained wholly within one House district. On the other hand, the City of Pine Bluff, which could have made up one Senate district, was split between two Senate districts, evidently to avoid the necessity of a political race between two senior senators. One reason given for preserving Crittenden County as a multi-member district was the desire to avoid splitting the City of West Memphis. There is a surface inconsistency here with the decision to split the City of El Dorado, but it is explained, perhaps, by the fact that Union County, of which El Dorado *215 is the county seat, is the largest county in the State in geographic area, and therefore a logical candidate for the creation of separate single-member districts. The policy against creating districts which are divided by a natural barrier, now urged upon us as a reason not to create a district that would contain territory in Pulaski County on either side of the Arkansas River, was violated in the drawing of House district 85, which includes territory on either side of the Arkansas River in Desha County, forcing citizens north of the river to drive more than 90 miles to reach the largest city in their district, where their state representative lives.

There are a number of crosscurrents here, and they point in various directions. On the whole, we are not persuaded that this factor has much weight. There were fairly strong policies underlying the apportionment plan, but they were not always consistently applied. To a certain extent, this is an inevitable part of line-drawing under the one-person, one-vote regime.

D.

Having fully reviewed the relevant factors, we must now balance them and come to a conclusion on the ultimate issue: whether black voters have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The areas of the State that are in issue include Pulaski County, Jefferson County and some adjacent territory, areas in South Arkansas including parts of Ouachita, Nevada, and Columbia Counties, and the area generally referred to as the Delta, including Mississippi, Lee, Phillips, Cross, St. Francis, Desha, Chicot, and Ashley Counties, together with some adjacent territory. (Crittenden County, which is also in the Delta, is not affected, except for one corner: it was dealt with in Smith, and neither side in this case asks for a change in House districts 48 and 49, which comprise most of the County.) To put it another way, the claim directly involves the following House districts: 56 through 70, 88, 80-83, 38, 47, 74, 75, 73, 85, and 100. It also directly involves the following Senate districts: 27, 28, 19, and 30.

We think the case is clear with respect to all these areas except Pulaski County (which includes House districts 56 through 70). The proof justifies consideration of the Delta, the Jefferson County area, and the Ouachita-Nevada Counties area as a group. There are some differences within the group, but as a whole it shares a significant number of common characteristics. Voting is markedly polarized between the races. With the exception of House district 82, in Jefferson County, there are no districts, House or Senate, with a majority-black voting-age percentage. A number of such districts, nine in the House and two in the Senate, could have been created. (The number of House districts, nine, is exclusive of the new majority-black House district in Crittenden County created by Smith. ) These districts would have been reasonably compact and contiguous. The history of official discrimination, the frequency of racial appeals in campaigning, the absence of successful black candidates, the effects of discrimination in education, employment, and health — all these factors point in favor of plaintiffs' claim. Also on the plaintiffs' side is the existence of a majority-vote requirement. The slating-process factor [9] is not in this case at all, and the state-policy factor does not point strongly either way. On the other hand, there are a number of responsive white legislators. On balance, a clear answer emerges. In these areas, black political opportunity is significantly lessened by the 1981 apportionment plan, and the plan violates Section 2 of the Voting Rights Act.

What about Pulaski County? Is it materially different from the areas we have just discussed? Many of the same factors are present. Voting is polarized, and four reasonably compact and contiguous [10] majority-black *216 single-member House districts could have been drawn, as opposed to the one three-member majority-black district that was drawn. (There are five single-member Senate districts, only one of them majority black, but the black population is not large enough to make possible more than one majority-black Senate district.) There has been no successful black candidate for the Legislature outside the majority-black areas. There have been some racial appeals in elections. The history of official discrimination affects voting in the entire State. Socio-economic disparities between the races are marked.

On the other hand, racial polarization in voting is less pronounced in Pulaski County. In 1976, Robert Johnston, an incumbent white House member, defeated a black opponent in a majority-black district. He had to have significant black support to do so. In 1984, a black candidate, Glenn Mahone, won the Democratic nomination for a House seat in a majority-white district against a white opponent in the runoff. He was defeated by the Republican nominee in the general election, but winning the Democratic nomination is better than any black candidate has done in legislative races in majority-white districts elsewhere in the State. In 1986, a black candidate for Congress, Thedford Collins, was defeated in the district as a whole, running fourth out of five in the Democratic Primary. But in Pulaski County he ran third and did well in some white areas, primarily among affluent voters. He was endorsed by the Arkansas Democrat, one of two state-wide newspapers, and shared with one other candidate the endorsement of the Arkansas Gazette, the other state-wide newspaper. Both newspapers are controlled by white people. And Marion Humphrey, elected municipal judge by a plurality in 1988, would not have won without significant white support. There have been racial appeals in elections, but they have not included the grosser features of some of the other incidents we have described.

Furthermore, the extent of socio-economic depression among black people is less pronounced in Pulaski County. As the table appearing at p. 211 above shows, among the 16 counties affected by the case, Pulaski County has the highest percentage of black high-school graduates, the highest black per capita income, the smallest proportion of black families below the poverty level, the smallest percentage of black people without a telephone, and the second smallest percentage of black people without a car. More importantly, the whole political atmosphere, with respect to black opportunity and participation, seems more open. Carol Willis, a man of wide political and governmental experience to whom we referred earlier, summed it up: Pulaski County is different.

Is it different enough? The question, in the end, is one of judgment, as to which reasonable politicians and judges may differ. We are instructed to approach such questions with "a searching practical evaluation of the past and present reality," and to take "a functional view of the political process." Thornburg, 478 U.S. at 45 , 106 S.Ct. at 2763 (citations and internal quotations omitted). Furthermore, the list of factors in the Senate Report "is neither comprehensive nor exclusive[;] ... other factors may also be relevant and may be considered." Ibid (footnote omitted). We think a factor not yet mentioned tips the balance against plaintiffs' claim in Pulaski County: the preference of the black community in Pulaski County, as expressed by two of its elected representatives, for the status quo. State Representatives Irma Hunter Brown and Grover Richardson, who in 1981 were two of the three black members serving in House district 62-64, both told the State Board of Apportionment that they preferred the existing multi-member arrangement to a system of 15 single-member districts in Pulaski County. See PX 14C (Minutes of the Public Hearing of May 12, 1981, before the Board of Apportionment). Others who appeared advocated *217 single-member districts in principle, but Representatives Brown (who is still in office) and Richardson, the only two black elected officials to speak, disagreed. Mr. Richardson seemed to favor single-member districts in Eastern Arkansas, but he thought "the three [black] representatives [in Pulaski County] favor[ed] the multi-member district." PX 14C, p. 11. Ms. Brown had mixed emotions on the subject but came down in the end to the view that the existing arrangement "was a little more palatable." Id. at 20. Both Representative Richardson and Representative Brown stuck to this view even under close questioning by Bill Bethea, who represented Governor White at the hearing. They seemed to find more strength and cohesion in the multi-member arrangement, especially for minority interests. We recognize that many political scientists disagree with this view, and that the opinions of two elected officials do not necessarily represent the views of all of their constituents, let alone the views of black citizens in other parts of the County. We note, however, the testimony of Governor White in the present case that no black citizen of Pulaski County asked for single-member districts.

After a careful consideration of all the factors, we resolve this close question in favor of the defendants. It would be unfair to fault the Board of Apportionment for acceding to the expressed wishes of the only two black legislators from Pulaski County who appeared before it.

In sum, we hold that no violation of Section 2 of the Voting Rights Act has been shown in Pulaski County. With respect to the other areas of the State challenged in this suit, we hold for the plaintiffs. Section 2 is being violated in those areas.

IV.

We must now fashion a remedy. A few comments may be helpful to the parties. In the first place, active participation at the remedial stage of this case on the part of the defendants, the members of the State Board of Apportionment, is vitally important. We are ordering them to submit a plan. They are the authorities charged by state law with the responsibility of drawing a plan of apportionment, and they should also bear the primary responsibility for conforming the present plan to the requirements of federal law. They are closer to and more familiar with all of the many factors, political, demographic, and social, that deserve to be considered. A suggested plan submitted by plaintiffs is also important. The alternative districts they have suggested already, at the liability stage of the case, may need to be adjusted, but plaintiffs have shown a considerable degree of skill in drawing them. We assume that defendants will consider the views of plaintiffs and all other interested and affected citizens in preparing their plan.

The task will not be simple. It is not just a matter of drawing districts that will have a majority-black voting-age population. There will inevitably be a ripple effect, necessary changes in the boundaries of adjacent districts. This effect should be limited as much as possible. It should be possible to limit it only to those districts which are directly adjacent to the new majority-black districts to be created. In this way, most of the legislative districts in the State will not be disturbed. We see no need for any changes in that part of the State which lies north and west of a diagonal line running from the southwest to the northeast. Nor, for reasons already given, need there be any changes in Pulaski County.

We are not holding that the law requires the creation of any particular number of majority-black districts. We know, and have found in this opinion, how many such districts can be created, and we also know that their lines can be drawn so as to make them reasonably compact and contiguous. There is, therefore, a sort of presumption that any plan adopted should contain that number of majority-black districts. There may be practical problems which we cannot foresee, though, and we are not foreclosing our duty to consider them after both sides submit their plans.

*218 We are requiring that plans be submitted by the parties on or before January 15, 1990. At the trial, defendants took the position that they would need 90 days within which to draft a plan, if we found liability on their part. We are giving them considerably less time, but we are doing so, we think, for good reason. First, the 90-day time period was mentioned as being necessary to re-draw the lines for the entire State, and that will not be necessary. Second, a good deal of the groundwork has already been laid by the detailed proof made in this case at the liability stage, including the maps of alternative districts introduced by the plaintiffs. Finally, if relief is to be effective at all, it must be in place sufficiently in advance of the 1990 elections to allow the public and prospective candidates to make necessary adjustments to the new lines. All of this will require a great deal of time, expense, and dislocation, but we believe the Voting Rights Act leaves us no alternative. The Act does not permit either this Court or the defendants to subject the plaintiffs and those they represent to another election conducted on an unlawful basis. The lines will have to be re-drawn again after the 1990 census, but fair and lawful representation in the Legislature to be elected in 1990 remains of great importance, especially in view of the fact that its members will draw new lines for congressional districts. Only if relief is afforded now will plaintiffs get in on the ground floor of this process to the extent required by law.

Judge Eisele has advised that he will dissent for the laches of the plaintiffs in filing this action and the inappropriateness of the requested injunctive relief so near to the next reapportionment and the 1990 elections. He further advises that he will concur with the majority in the finding that there was a Section 2 violation although he would exclude not only Pulaski County but also Jefferson County and House districts 74, 75, 82, and 100. Since he disagrees with certain of the Court's findings and conclusions, he will in due course file a dissent and concurring opinion in order to further set forth his views. He agrees with the majority, however, that its opinion on Section 2 of the Voting Rights Act should be filed at this time and that an appropriate order, or orders, be entered for the purpose of giving the defendant Board and all interested parties as much time as possible to redistrict in accordance with the Court's opinion. Judge Eisele also disagrees with the majority's opinion that, because the Court has found that the Board in 1981 could have created thirteen majority-black House districts and three majority-black Senate districts, there is some sort of a presumption that the present Board in 1990 should, in complying with this Court's order, create that number of House and Senate districts. He will also explain his views on this remedy issue in his opinion.

An order is being entered today to carry out the conclusions and directions expressed in this opinion.

CONTENTS

Dissenting Opinion, filed Dec. 7, 1989 (# 1) ........................... 219-226

Concurring and Dissenting Opinion, filed Jan. 19, 1990 ................. 226-284

OVERVIEW ............................................................... 227

I. SECTION 2 OF THE VOTING RIGHTS ACT: THE INTERPRETIVE

FRAMEWORK ................................................................ 228

A. A Walk Through the Statute ............................................. 229

B. To Make Out A Section 2 Violation, Must Plaintiffs Prove Both Of The

Following: That, As A Result of The 1981 District Lines, Blacks (1)

Have Less Opportunity To Participate In The Political Processes;

and (2) Have Less Opportunity To Elect Representatives Of Their

Choice? .............................................................. 230

C. Introducing The "Zimmer" and "Senate" Factors ............................ 231

II. CONSTITUTIONAL LIMITATIONS: USE OF "REPUBLICAN FORM OF

GOVERNMENT" GUARANTEE OF ARTICLE IV, SECTION 4, U.S.

CONSTITUTION ............................................................. 232

*219

III. HAVE PLAINTIFFS SHOWN THAT THE 1981 DISTRICTING PLAN

RESULTS IN BLACKS HAVING LESS OPPORTUNITY THAN OTHERS

TO PARTICIPATE IN THE POLITICAL PROCESS? ................................ 237

IV. HAVE PLAINTIFFS SHOWN THAT THE 1981 DISTRICTING PLAN

RESULTS IN BLACKS HAVING LESS OPPORTUNITY THAN OTHERS

TO ELECT CANDIDATES OF THEIR CHOICE? .................................... 239

V. THE THORNBURG FACTORS .................................................... 240

A. Political Cohesiveness ................................................ 241

1. Black Voting Behavior .............................................. 241

2. White Voting Behavior .............................................. 243

B. What is a "Minority?" ................................................. 250

VI. THE "ZIMMER" AND SENATE FACTORS .......................................... 252

A. Relevance and Effect .................................................. 252

B. Application of Senate Factors in This Case ............................ 255

1. Senate Factor # 1 ................................................... 255

2. Senate Factor # 5 ................................................... 257

3. Senate Factor # 3 ................................................... 257

4. Senate Factor # 6 ................................................... 259

5. Senate Factor # 7 ................................................... 261

6. Lack of Responsiveness of Elected Officials ......................... 262

7. The Strength or Tenuousness of Policies Underlying the Plan of

Apportionment ..................................................... 262

VII. THE MAJORITY'S DECISION TO EXCLUDE PULASKI COUNTY DISTRICTS:

AN ANALYSIS ............................................................. 262

VIII. DID PLAINTIFFS ESTABLISH ONE OR MORE SECTION 2 VIOLATIONS? ............... 263

A. Mississippi and Crittenden Counties ................................... 266

B. Phillips County ....................................................... 267

C. Chicot, Desha and Ashley Counties ..................................... 270

D. St. Francis and Lee Counties .......................................... 272

E. Jefferson County and Pine Bluff ....................................... 273

F. Lincoln and Cleveland Counties ........................................ 274

G. Ouachita, Nevada and Columbia Counties ................................ 275

H. The Senate Districts .................................................. 276

I. Summary ............................................................... 277

IX. CONCLUSION ............................................................... 278

ADDENDUM: THE EIGHTH CIRCUIT REVERSAL IN WHITFIELD: SAME

PROBLEM; SAME RESULT ............................................... 281

EISELE, Chief District Judge, dissenting. (Eisele # 1)

Filed Dec. 7, 1989.

The majority opinion dealing with Section 2 of the Voting Rights Act was filed herein on December 4, 1989. In that opinion, the majority concluded that the claims of the plaintiffs were not barred by the doctrine of laches. I dissent. [1]

I agree with Judge Arnold that laches is an equitable defense which means essentially that if there is unreasonable and unjustified delay in the filing of a lawsuit, and this delay causes prejudice, the court of equity may dismiss the complaint. As stated by Judge Arnold:

The greater the delay, or the more unreasonable, the less prejudice need be shown; and vice versa. The Court must weigh the facts and interests on both sides, summon up the discretion of a chancellor, remember that it is a court of conscience and not of legal stricture, and come as close as it can to a fair result. Frequently there are some good arguments *220 on both sides, and that is the case here.

Again I agree.

It is important to first note how legislative redistricting fits into our constitutional schemes for a representative democracy.

Regular, periodic legislative redistricting is indeed unique in the election law area. It is absolutely essential if we are to have fair and equal representation and if we are to give meaning to the one-person, one-vote concept. Yet, since our nation was formed, most legislative redistricting has, for practical reasons, been done on a decennial basis, keyed as it is to our national census procedures. The willingness to "let things alone" for such a long period (ten years) reflects the practical manner in which we carry out the democratic principle of proportional representation. [2] We are all aware that within such a span of years, there can be dramatic changes in the populations of the various legislative districts. It is generally conceded that the farther one gets from the last census the more unreliable are the population figures as a measure of the current weight actually reflected in each legislator's vote. And we know that the number of black and white residents may also change dramatically and not necessarily proportionately. For example, in Rybicki v. State Board of Elections, 574 F.Supp. 1082 (N.D.Ill.1982), evidence adduced at trial showed that between the 1970 and 1980 census, the number of senate districts in which blacks constituted a majority had increased from five to six. Id. at 1092 . In one district alone, the racial composition had changed from 21.6 percent black in 1970 to 57.7 percent black in 1980. Id. at 1109 . Still society accepts the utility of a decennial system and implicitly accepts that as time passes the apportionment may no longer comport with constitutional requirements. This necessarily detracts from the conclusion in Smith v. Clinton, 687 F.Supp. 1310 , (E.D.Ark.), aff'd, ___ U.S. ___, 109 S.Ct. 548 , 102 L.Ed.2d 576 (1988) that the recurring harm (assumedly occurring at each new election) justifies greater leniency in imposing time limits on the commencement of a voting rights challenge to redistricting plans. Population changes since the last regular redistricting may have corrected or mooted the original vice ( ergo, no recurring harm) or, indeed, may have exacerbated such vice or created new constitutional infirmities which will have to be corrected at the next decennial redistricting. We simply do not know. For the most part, we are left to guess at the realities of the current situation. Because of the seven-year delay, we are left standing in sand.

The 1980 census data shows that several of the House districts drawn by the defendant Board in 1981 (in addition to those in Pulaski and Jefferson Counties) had majority-black populations even though the voting-age-black populations were less than 50 percent. District 74 has 52 percent black total population but 47 percent black-voting-age population. District 75 had 51 percent black total population but only 45 percent black-voting-age population. And District 100 had 52 percent and 47 percent, respectively. Indeed, one Senate district — District 30 — had a 51 percent black total population but only 43 percent black-voting-age population. What inference can be drawn from these figures? One is that in 1981 blacks had a higher percentage of their total population in the "under-18" category than did whites. This means that, if all other factors (including in-and-out migrations and death rates) remained constant, more blacks than whites, proportionately, would reach the age of 18 during the period between 1980 and 1989. All persons nine years of age or older by a certain date in 1980 would be over 18 by that same date in 1989. The predictable consequence would be that the percentage of black-voting-age population would increase and the percentage of white-voting-age population would decrease in those districts during *221 that nine-year period. But we do not know the relative migration rates of blacks and whites during that period and we do not have a handle on the relative relevant mortality rates. Nevertheless, this exercise is instructive because it suggests how different might be the 1989 data (if it were available) from the 1980 census figures.

Assuming that plaintiffs knew, or should have known, of the factual and legal bases for their dilution claim back in 1981 when the reapportionment was adopted, or in 1982 when the Voting Rights Act was amended, and nevertheless allowed over seven years to pass before objecting, then it is not unreasonable to say that the strong practical reasons underlying our decennial reapportionment system also require the conclusion that they delayed entirely too long. Stability and continuity require redistricting plans, once in place for a substantial period of time, to remain in force and effect until the next decennial redistricting, unless a plaintiff can show some strong legitimate reason or excuse for postponing or delaying a challenge.

Was information available to plaintiffs and other citizens back in 1981 upon which to base their dilution claim? The answer is: Yes.

First, the 1981 reapportionment of Arkansas' General Assembly generated ample publicity and public debate before and after the Board of Apportionment adopted its plan for state House and Senate districts. Numerous public hearings were conducted throughout the state. Attorney General Clark testified that for a period of several months he received comments, suggestions and criticisms of various proposed districting plans almost daily.

It is interesting also to note that plaintiffs have in this lawsuit complained bitterly that a majority of the defendant Board members refused to grant even a 30-day extension back in late June of 1981 to enable interested parties to present information on adequate minority representation. However, this series of events is also relevant to the question of whether plaintiffs' delay in commencing the present action was reasonable. It is relevant because it shows that during this period of time in 1981 there existed a well-informed and highly-organized advocacy group representating the same interests as are championed by the plaintiffs here. Its work more than amply laid the groundwork for the present challenge. But that work, and the evidence developed then inexplicably lay unused for nearly eight years, only to be dusted off when this suit was filed near the close of the decade and only fifteen months before the commencement of the 1990 census.

As the apportionment process began, Ms. Brownie Ledbetter testified that she was appointed to serve as co-chairperson of the Committee for Fair Representation, a coalition of some eighteen national and local organizations concerned with various aspects of the reapportionment — including its effect on black voters. The coalition, which included the Arkansas Education Association, the Arkansas AFL-CIO, Common Cause, the state chapter of the National Organization of Women, the NAACP and the Urban League, organized public forums and called for additional public hearings in an effort to bring to the attention of the Board of Apportionment the need to insure adequate minority representation in both the House and Senate.

More importantly, the committee obtained copies of computer tapes which held census data for the state of Arkansas — the same tapes used by the Board of Apportionment. With the help of a technical support staff, the committee was able to analyze this data to determine the effect various districting proposals would have on blacks and other minorities, and then develop alternative proposals. And these tapes, or copies of them, were again used many years later by plaintiffs' expert Dr. Jerry Wilson to analyze the existing apportionment and develop alternate maps showing that a number of majority black House and Senate districts could have been drawn back in 1981. The information needed to mount an attack on the plan was as available to the plaintiffs here as it was to Ms. Ledbetter and the organizations involved with the Committee for Fair Representation. *222 And evidence concerning polarized voting and other issues was also available. This is obvious if one simply examines the evidence introduced in this case. Much of that evidence arose out of elections and events occurring before 1982. And almost all of the evidence used in this case was available by 1986. [3]

The decennial legislative redistricting process contemplated by Article VIII, Section 1 of the Constitution of Arkansas provides an opportunity for the Board of Apportionment to obtain and consider the views of all affected persons and groups before settling upon an apportionment plan. The whole process may be extended (as was done in 1981) [4] and there is little time pressure anyway because that process ordinarily occurs in off-election years. But the majority's order entered December 4, 1989, directs the defendants to come up with a new plan by January 15, 1990, that is, within a period of 41 days which embraces the Christmas and New Year holidays. The Court chose this early date because it recognized that the "election season" for 1990 will begin soon thereafter. (Indeed, it notes that the filing period will begin on March 20, 1990.) Once the new plan is adopted and filed with the Court, there should be an opportunity for interested parties to object to the plan. If objections are filed, a hearing will probably be required, and, after this Court acts, there is the possibility of appeals. The prospect that this Court may have to interfere with the state's election schedule is a real one. And if election sequences are not altered or enjoined, there is at least the probability that legislative races will be held under the cloud of one or more appeals. Under certain circumstances such prospects of difficulty, expense, and confusion could be accepted as reasonable and essentially unavoidable. But here?

The defendants have a Hobson's choice: (1) whether to spend a large amount of money in an effort to obtain updated current census figures, or (2) to simply rely on the stale, outdated 1981 census figures. Whether the first alternative is even available in the short period provided by the Court is questionable. If such information can be and is obtained, it is possible that it will send all parties "back to the drawing board," invite further controversy, and require new hearings. If the Board simply relies on the 1981 figures, it will have no assurance that its decisions will have any current validity. And, all for what purpose? To create legislative districts for use in only one election. Then, in 1991, the new Board will redistrict on the basis of the 1990 census figures. It is open to question whether it is fair or reasonable to require such immense expenditures of time and money to create, for one election, districts whose lines will probably be changed again within another year. And, if the Board uses the 1981 census data, it will be creating, for one election, essentially hypothetical legislative districts as to whose validity no one can even guess prior to the 1990 census.

*223 It may be said that the expenditure of such time and effort would have been required if the attack had been made in 1981 or 1982. But the data then would then have been relatively current and the resulting redistricting would have been in effect to control the elections for the following seven or eight years.

I note that the majority in rejecting the defendants' argument that the present lawsuit should be barred by the equitable doctrine of laches relies principally upon Smith . I have thought long and hard as to the effect to be accorded this ruling and the Supreme Court's affirmance. I conclude that the application of laches is not foreclosed by the ruling in Smith . It is clear that relief in voting rights cases is to be "fashioned in the light of well-known principles of equity." Baker v. Carr, 369 U.S. 186, 250 , 82 S.Ct. 691, 727 , 7 L.Ed.2d 663 (1962) (Douglas, J. concurring). These principles are by their nature case-specific. Thus, while Smith held that laches did not bar the plaintiffs in that case in their challenge to the election scheme for a single multi-member legislative district, its ruling should be limited to the specific facts of that case. It ought not be read to permanently enjoin a district court's exercise of equitable judgment in all redistricting voting rights cases forevermore.

As the majority points out, laches requires a finding of two key elements: inexcusable delay in commencing a suit resulting in undue prejudice to the defendant. However, the majority also takes the position that plaintiffs' delay in commencing their challenge to the 1981 apportionment should date from the decision in Thornburg v. Gingles, 478 U.S. 30 , 106 S.Ct. 2752 , 92 L.Ed.2d 25 (1986). In other words, the majority suggests that even if adequate evidence of a "Section 2" violation was available to the plaintiffs back in the 1981-1984 period, the pertinent law was so unclear and confusing that plaintiffs should be excused for any delay before Gingles cleared the air in 1986. And using this decision as the "pole star" of the law in this area, it is argued that plaintiffs waited only two years thereafter to commence this action, and that such a delay is not unreasonable. But I do not agree that this is the appropriate moment from which to measure plaintiffs' delay. To say that plaintiffs' "clock" began to run only after the Gingles decision had been rendered is to ignore a much broader sweep of recent history in the development of the law with respect to voting rights.

Specifically, the majority ignores that Congress amended Section 2 of the Voting Rights Act in 1982 by clearly prohibiting the use of any voting structure "which results in a denial or abridgment 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 applied). The amendment also added a new subdivision that stated that a violation of Section 2 could be established upon showing of such results "based on the totality of circumstances." Id. § 1973(b). The amendment was an unambiguous rebuke to the Supreme Court's decision in City of Mobile v. Bolden, 446 U.S. 55 , 100 S.Ct. 1490 , 64 L.Ed.2d 47 (1980), which held that a violation of the Voting Rights Act required a showing of discriminatory purpose.

Moreover, by amending the statute's language so that a violation of Section 2 could be established using a "results" analysis, Congress did not write on a clean slate or establish a new rule of law. It was reaffirming what it believed to have been the standard prior to Bolden . And note plaintiffs' contention in their pretrial brief that "[p]rior to the decision in Bolden , plaintiffs could prevail in a vote dilution case by showing either discriminatory results or intent." See Plaintiffs' Proposed Findings of Fact and Conclusions of Law at 53, citing, 1982 U.S.Code Cong. & Admin.News at 192-201.

The so-called Senate Report factors, which accompanied the 1982 amendment and which arguably must now be applied to the present case, were also not new. These factors were instead derived from the analytic framework first articulated in Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir.1973) ( en banc ), aff'd on other grounds sub nom., East Carroll Parish School Board v. Marshall, 424 U.S. 636 , 96 *224 S.Ct. 1083, 47 L.Ed.2d 296 (1976). Thornburg v. Gingles, supra, 478 U.S. at 36 n. 4, 106 S.Ct. at 2759 n. 4. And only days after passage of the amendments, the Supreme Court rejected a portion of the plurality's ruling in Bolden and held that the Zimmer factors are to be used to determine whether an election scheme purposefully discriminates against minority voters in violation of the Fourteenth Amendment. Rogers v. Lodge, 458 U.S. 613, 623-24 , 102 S.Ct. 3272, 3278-79 , 73 L.Ed.2d 1012 (1982). It is also interesting to note that these factors were known to the defendants back in 1981 as they prepared to develop a reapportionment plan.

After its initial meeting, the Board of Apportionment commissioned Professor Albert Witte of the University of Arkansas School of Law to prepare a lengthy legal memorandum on the current state of the law with rspect to legislative redistricting. Among its conclusions the memorandum noted that a challenge to an apportionment on vote dilution grounds could be supported with such evidence as "an historical pattern of underrepresentation, the current effects of past discrimination which hindered effective participation in the electoral process and the group's depressed economic and social status which further impeded their participation." Plaintiffs' Exhibit 14B, p. 27. The point is: The law was adequately clear to lawyers and legal scholars without Gingles .

While Gingles perhaps clarified the weight to be accorded the various Senate or Zimmer factors, it does not provide plaintiffs an excuse for waiting until January 1989 to commence this lawsuit. To say plaintiffs could do nothing without the benefit of Gingles is to say that the 1982 amendments to the Voting Rights Act were meaningless and without practical effect, indecipherable to anyone interested in voting rights issues. There simply is no basis for such a conclusion. Moreover, it is contrary to the experience of numerous other litigants, who in the absence of Gingles , still managed to challenge similar reapportionment schemes much earlier in this decade, and most immediately after the reapportionments took effect. See e.g., Ketchum v. Byrne, 740 F.2d 1398, 1401 (7th Cir.1984, cert. denied sub nom., City Council of Chicago v. Ketchum 471 U.S. 1135 , 105 S.Ct. 2673 , 86 L.Ed.2d 692 (1985)); Gingles v. Edmisten, 590 F.Supp. 345, 350 (E.D.N.C.1984), aff'd in part and rev'd in part, Thornburg v. Gingles, 478 U.S. 30 , 106 S.Ct. 2752 , 92 L.Ed.2d 25 (1986). (Complaint filed within two months after challenged apportionment adopted by the state; Major v. Treen, 574 F.Supp. 325, 327 (E.D. La.1983) (Trial in suit challenging 1981 congressional redistricting commenced in March 1983); Rybicki v. State Board of Elections of Illinois, 574 F.Supp. 1082, 1088 (N.D.Ill.1982) (State legislative redistricting plan adopted October 1981; district court opinion rendered four months later). Furthermore, prior challenges to Arkansas reapportionments were also brought soon after the adoption of the challenged plan. See Yancey v. Faubus, 251 F.Supp. 998 (E.D.Ark.1965), aff'd sub nom., Crawford County Bar Ass'n v. Faubus, 383 U.S. 271 , 86 S.Ct. 933 , 15 L.Ed.2d 750 (1966), Kelly v. Bumpers, 340 F.Supp. 568 (E.D.Ark.1972), aff'd 413 U.S. 901 , 93 S.Ct. 3047 , 37 L.Ed.2d 1019 (1973). The bases of these many attacks differ, of course, but all of the plaintiffs involved recognized the need to move promptly with their challenges.

Clearly then, plaintiffs should have been able to determine that a cause of action existed at least some time shortly after the 1982 amendment to the Voting Rights Act took effect, and also should have been aware of enough relevant evidence that could have been used in an effort to establish a violation of Section 2. Viewed in this context, plaintiffs' delay is more on the order of seven years rather than the two years suggested by the majority.

It is true, as Judge Arnold states, that this is a suit in equity. It is therefore not governed by any statute of limitations. Still, it is important to remember that in Owens v. Okure, ___ U.S. ___, ___, 109 S.Ct. 573, 583 , 102 L.Ed.2d 594 (1989), the Supreme Court held that a state's general or residual statute of limitations should be applied to Section 1983 claims. Arkansas' residual statute of limitations is three *225 years. Ark.Code.Ann. § 16-56-105. While this statute of limitations does not control here, it can provide a rough guide for the application of laches. See Dobbs, Remedies, § 2.3, p. 43-44 (1973).

Even if a case is not dismissed for laches, should not injunctive relief be denied where the time strictures caused by the tardy filing undermine the integrity or effectiveness of such relief? In Reynolds v. Sims, 377 U.S. 533 , 84 S.Ct. 1362 , 12 L.Ed.2d 506 (1964), Justice Warren, writing for the Court stated:

[O]nce a State's legislative apportionment scheme is found to be unconstitutional, it would be the unusual case in which a court would be justified in not taking appropriate action to insure that no further elections are conducted under the invalid plan. However, under certain circumstances such as where an impending elections is imminent and a State's election scheme is already in progress, equitable considerations might justify a court in withholding the granting of immediately effective relief in a legislative apportionment case even though the existing apportionment scheme was found invalid. In awarding or withholding immediate relief, a court is entitled to and should consider the proximity of a forthcoming election and the mechanics and complexities of state elections laws, and should act and rely upon general equitable principles. With respect to the timing of relief, a court can reasonably endeavor to avoid a disruption of the election process which might result from requiring precipitate changes that could make unreasonable or embarrassing demands on a state in adjusting to the requirements of the court's decree.

Id., 377 U.S. at 585 , 84 S.Ct. at 1393 . I recognize that the majority is trying to avoid "disruption of the election process" but it is nevertheless my view that the order entered herein on December 4, 1989, will, because of the foreseeable time limitations, make "unreasonable or embarrassing demands" upon the state of Arkansas.

In MacGovern v. Connolly, 637 F.Supp. 111 (D.Mass.1986) the three-judge district court panel dismissed plaintiffs' complaint attacking a 1977 state legislature's apportionment plan where the suit was not filed until 1986 just prior to the commencement of the next reapportionment. The court interpreted Reynolds as allowing dismissal of an apportionment claim as well as the staying of equitable relief. Id. at 114-16. The court also held that laches required dismissal because plaintiffs' delay in commencing the action was inexcusable, and because going forward would result in undue prejudice to the defendant. Id. at 116.

In Maryland Citizens for a Representative General Assembly v. Governor of Maryland, 429 F.2d 606 (4th Cir.1970), the court of appeals upheld a district court's refusal to convene a three-judge panel and subsequent dismissal of a suit challenging the state's apportionment of the general assembly. The court found that in passing upon the motion to convene a three-judge panel, the district court was required to determine whether the suit made out a "substantial claim" and further reasoned that Reynolds supported the conclusion that a claim could not meet this test where injunctive relief was unavailable. Id. at 610-11 .

Similarly, the court in Simkins v. Gressette, 631 F.2d 287 (4th Cir.1980), upheld the district court's dismissal of an apportionment challenge that was filed two days before the opening of the filing period for the 1980 election. Id. at 289 . Significantly, Simkins involved a challenge to the South Carolina Senate apportionment plan on the grounds that it diluted black voting strength in violation of Section 2 of the Voting Rights Act, as well as the First, Thirteenth, Fourteenth and Fifteenth Amendments. Id. at 289 . Nothing that the last election for the state senate had been held in 1976, the court stated:

Instead of bringing suit then, they chose to wait for more than three years until the eve of the 1980 elections. The record reflects no good reason for the delay. Such a delayed suit, if maintained, would clearly cause a major disruption in the election momentarily to begin in South Carolina. This disruption, coupled with the fact that 1980 is also the year of a *226 national census which will likely require reapportionment in South Carolina, places this case squarely within our holding in Maryland Citizens.

Simkins v. Gressette, supra, 631 F.2d at 296 .

Admittedly, dismissal of a voting rights challenge because of delay in filing represents a substantial step not necessarily supported by the language quoted above from Reynolds . However, these cases do make that very sound point that, although the judiciary may be called upon to enter the "political thicket," equity "demands that a federal court stay its hand when judicial relief makes no sense." MacGovern v. Connolly, supra, 637 F.Supp. at 116 .

The time frames here are such that I see no way that a reasonable and fair remedy — commensurate with the complexity and difficulty of the task — could be crafted and properly implemented before the next "election season" begins.

The proximity of the next election as well as the next reapportionment, the mechanics and cost of attempting a remedy that will last only one election, the staleness and unreliability of the essential data, the likelihood of confusion to voters and loss of continuity of representation caused by these events, firmly weigh in favor of dismissal or at least in favor of denying any remedy beyond declaratory relief. Relief beyond that in the circumstances of this case should be denied as a matter of law. And, on balance, I conclude that dismissal would be more appropriate under the facts of this case.

The majority concludes that any disruption caused by implementing a remedy is outweighed by the fundamental importance of vindicating the right of equal participation in the political process. Consequently, the majority believes it would be unreasonable to say to the plaintiffs that they should "[w]ait for another census. The time is not yet ripe." supra at 203. But the principle at stake here is not ripeness. If anything, this suit is overripe. The critical facts presented at trial, and which control much of the court's determination, are admittedly stale. The equitable principle implicit in withholding a grant of relief is predicated not upon the theory that the plaintiffs should wait, but upon the theory that they waited entirely too long.

EISELE, Chief District Judge, concurring and dissenting. (Eisele # 2)

Filed Jan. 19, 1990.

In a dissenting opinion filed on the 7th of December 1989, I set forth my belief that this entire proceeding should have been dismissed because the plaintiffs waited without excuse until 1989 to challenge the redistricting plan adopted in 1981 and because of the prejudice caused thereby to the defendants, the people in the affected areas, and, indeed, to the integrity and reliability of the Court's fact-finding function itself. Acknowledging the broad discretion given the chancellor in such equitable proceedings, I nevertheless concluded that it would be an abuse of that discretion — as a matter of law — to permit this case to go forward under its unique facts and circumstances. My brothers disagree; so the case proceeds to resolution on the merits.

The only "merits" issue dealt with by the majority to date is whether plaintiffs established one or more violations of Section 2 of the Voting Rights Act. The constitutional and "intent" issues have been reserved for later resolution. And this opinion, in like manner, will be confined to the Section 2 issues. I dissent from the majority's opinion on those Section 2 issues.

When the majority announced its findings and conclusions on the Section 2 issues, I indicated that I also believed that some Section 2 violations had been shown. However, as I further reviewed the law and the evidence, I have become convinced that plaintiffs failed to prove any Section 2 violations. This blanket conclusion is based upon my opinion that plaintiffs did not prove, in connection with any of the House or Senate districts, that the redistricting plan of 1981 resulted in blacks in those districts having less opportunity than others to participate in the political process. As a matter of law I have concluded that *227 plaintiffs must make that showing and that simply proving that blacks, as a result of the district lines, have less opportunity than others to elect candidates of their choice will not suffice. Upon the assumption that I am wrong in this conclusion, then I would agree that plaintiffs have shown a Section 2 violation with respect to at least one district and perhaps more. See Section VIII, infra.

The views I express here follow from, and are consistent with, opinions I have previously expressed in other voter rights cases coming out of Lee and Phillips Counties, the former involving justice of the peace districts in Lee County and ward districts in the city of Marianna and the latter involving a challenge to the state's run-off law in primary elections. See Campbell v. Lee County Election Commission, H-C-86-48 (E.D.Ark.1986) and Whitfield v. Democratic Party of Arkansas, 686 F.Supp. 1365 (E.D.Ark.1988), rev'd in part, 890 F.2d 1423 (8th Cir.1989) (Bright, J., dissenting in part). I have also dealt with a challenge to multi-member districts in Pulaski County, under Section 2 as it was interpreted before the 1982 amendments. See Leadership Roundtable v. City of Little Rock, 499 F.Supp. 579 (E.D. Ark.1980).

I concur in the majority's opinion that plaintiffs have not shown any Section 2 violations in connection with any of the district's created in 1981 in Pulaski County, Arkansas.

OVERVIEW.

It is my view that many Voting Rights cases, such as this one, are changing the political landscape of America in fundamental ways without legislative mandate and without the benefit of scholarly legal and political discourse. In so doing these cases are, in an almost inadvertent manner, redefining the nature of our democratic form of government, contrary, I believe, to the Constitution.

Do we really believe in the idea of one political society or should this be a nation of separate racial, ethnic, and language political enclaves? Surely such issues are worthy of serious, focussed debate and discussion. Ironically, there has been precious little of either. I attribute this to inadequate awareness on the part of the courts, the Congress, and the people of this Nation. And, strange as it seems, it appears that some of those who are most involved with the voting rights cases are the least aware of the larger, long-range issues and consequences which are directly implicated.

When the Voting Rights Act was passed in 1965, its single aim was "black enfranchisement in the South. Obstacles to registration and voting, that is, were the sole concern of those who framed the statute." A. Thernstrom, Whose Votes Count? Affirmative Action and Minority Voting Rights, 3 and 18 (1987) [Hereinafter "Thernstrom"]. See also 42 U.S.C. § 1973 et seq. With the Supreme Court's decision in Allen v. State Board of Elections, 393 U.S. 544 , 89 S.Ct. 817 , 22 L.Ed.2d 1 (1969), the emphasis began to shift as the Court "implicitly enlarged the definition of enfranchisement," making distinctions between "meaningless" and "meaningful" votes. Thernstrom at 4 and 22-24. Statutory amendments were enacted in 1982 empowering minorities to challenge any standard, practice, or procedure relating to elections on the ground of discriminatory "results." 42 U.S.C. § 1973 (1982). Court decisions interpreting Section 2 in an effort to deal with electoral discrimination have exposed questions relating to the nature and structure of our government and the efficacy of majoritarian democracy. See, e.g., Whitfield, supra . Where are we now in 1990?

We have arrived at a point no one envisioned in 1965. The right to vote no longer means simply the right to enter a polling booth and pull the lever. Yet the issue retains a simple Fifteenth Amendment aura — an aura that is pure camouflage. An alleged voting rights violation today is a districting plan that contains nine majority-black [or Hispanic] districts when a tenth could be drawn. The question is: how much special protection from white competition are black candidates entitled to? For instance, when a *228 different plan might give a seat to another black, should the interests of white incumbents give way to the goal of minority officeholding? The temptation to provide maximum protection (a maximum number of seats) is strong and has been only intermittently resisted. * * * Districting plans are torn up following every decennial census. * * * There is thus a need to devise new plans, to determine when black ballots "fully count." The phrase itself invites a definition that gives those ballots maximum weight, defined as officeholding; anything less suggests a compromised right. Yet maximum weight implies an entitlement to proportionate ethnic and racial representation — a concept that is no less controversial with respect to legislative bodies than with reference to schools and places of employment. Voting rights has become another immensely complex affirmative action issue, distinctive only in not being acknowledged as such.

The myth of moral simplicity has largely insulated the voting rights issue from debate, yet perhaps no other affirmative action question is more significant.

Thernstrom at 5-6. (Emphasis added)

My objective in formally stating my views in this dissent in such detail is not only to explain the law as I understand it but also to help open up the issues to legal, scholarly, popular and political debate, keeping in mind, as we must, that we are here dealing with the heart and soul of our democratic government. The issues raised in this case should be resolved before the 1990-1991 nationwide redistricting season begins.

I. SECTION 2 OF THE VOTING RIGHTS ACT: THE INTERPRETIVE FRAMEWORK.

The courts are thrust into what Justice Felix Frankfurter called a "political thicket" primarily as a result of Section 2 of the Voting Rights Act of 1965, as amended, 42 U.S.C. § 1973 . That section reads as follows:

(a) No voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner 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, or in contravention of the guarantees set forth in section 1973b(f)(2) of this title, as provided in subsection (b) of this section.

(b) A violation of subsection (a) is established if, based on the totality of the circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by members of a class of citizens protected by subsection (a) in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered: Provided, that nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.

I suggest that the federal courts have, in many cases, not followed the language of the Voting Rights Act or the U.S. Supreme Court's opinions interpreting that Act and, as a consequence, have unwittingly crossed constitutional lines. This is one of those cases. Claiming simply to be giving a liberal interpretation to the language of the statute to effectuate its purposes, "judicial decisions [have been] shaped less by the statute than the statute by the decisions." Thernstrom at 8-9. This approach has left judges without real guidance to, in effect, "do right" in this most sensitive of areas, where what is "right" is far from clear.

Congress and the courts have declared that members of minority groups are entitled to "an equal opportunity to elect representatives of their choice." What does that high-sounding phrase really *229 mean? Who counts as a "representative of their choice," and when are opportunities truly equal? If the ballots of white Democrats in an overwhelmingly Republican county "count," if Democrats who are never able to elect "candidates of their choice" are nevertheless fully enfranchised, what about black Democrats when no black ever wins? Are blacks and Hispanics ever properly represented by whites? Are black candidates who run as part of a white-dominated ticket consequently "white" candidates? Do socioeconomic disparities between whites and blacks mean that members of the two groups stand on unequal political footing, and that compensation in the form of special protection in the electoral process is appropriate?

Thernstrom at 6-7.

The grammar of Section 2 and the ordinary meaning of its words and phrases have taken second place to the felt need to remedy all perceived wrongs in the political processes through which we give expression to the "republican form of government" that is guaranteed to every state by the Constitution of the United States. See U.S. Const. art. IV, § 4. Although the language of Section 2(a) proscribes those things which result in a denial or an abridgement of the "right to vote," cases such as this carry us far beyond any possible issues concerning each citizen's "right to vote" to the new world of "group rights" to proportionate representation. Care must, therefore, be taken that in an effort to insure individual or even "group" rights, we do not undermine — perhaps unconstitutionally undermine — the basic principles of our democratic government.

So, while I agree with the majority that, "the statute should be construed liberally in favor of its object," I would not state that object quite as broadly as they do, to wit: "to open up the electoral process to full participation." Majority op. at 204. Rather, I would say the purpose of Section 2 in the context of this case is to prohibit states and other political subdivisions from imposing or applying any standard, practice or procedure which results in black citizens' having less opportunity than others to participate in the political process and to elect representatives of their choice. It is the "opportunity" of blacks "to participate" and "to elect" to the same extent as other members of the electorate that may not be denied or abridged by some state created or applied barrier. In my view, a "liberal construction" (or, for that matter, a "conservative construction") is only available where the language of the statute permits more than one reasonable interpretation. In connection with the issues we are called upon to resolve in this case, it is my opinion that the language of Section 2 and the decisions of the United States Supreme Court give adequate guidance. We need look no further for some liberal construction consistent with the Act's remedial purposes. We should simply rigorously apply the law. And, if after so applying the statute we are still left with constitutional problems, so be it. We must then simply face up to the need to deal with them.

I will in the course of this opinion attempt to point out where the majority has departed from this standard of rigorous statutory construction in its resolution of the issues.

A. A Walk Through the Statute.

By suggesting that the language of Section 2 is clear, I do not overlook the convoluted grammatical arrangement thereof which requires the reader to carefully follow the succession of phrases by which the statute "reinterprets" itself.

Subsection (b) of Section 2 makes it clear that the operative provisions of Section 2 are contained in subsection (a). In other words, the question always is whether the challenged standard, practice or procedure (hereinafter SPP) violates subsection (a). We know this because subsection (b) merely refers back to subsection (a) by use of the following introductory language: "A violation of subsection (a) is established if...." So the basic question which courts are asked to answer in applying Section 2 is whether the SPP is being imposed or applied "in a manner which results in a denial or abridgement of the right of any *230 citizen ... to vote on account of race...." If it is decided that the challenged SPP does not result in a denial of the right of any citizen to vote on account of race, then the court must determine whether that SPP results in an "abridgement of the right of any citizen to vote on account of race." In this context, it is assumed that "abridge" would mean to curtail, limit, or burden. In any event, we are dealing with the right to vote, either its denial or its curtailment or limitation.

As pointed out above, subsection (b) of Section 2 identifies what is necessary to establish a violation of subsection (a). Of course, it is assumed that there should be some logical nexus between the showing identified in subsection (b) and the "denial or abridgement" language of subsection (a). According to subsection (b), a denial or abridgement of the right to vote "is established if ... it is shown that the political processes leading to nomination or election ... are not equally open to participation by members of ..." a protected class.

It is easy to determine whether a SPP denies any citizen the right to vote. If not, subsection (b) states that a showing that the "political processes" are not "equally open to participation," because of the challenged SPP, would nevertheless be adequate to establish that the SPP abridged the right to vote. So we look at the SPP and determine if it has resulted in the political process "not being equally open to participation" by black citizens. If it did not have that effect, then, logically, one would conclude no violation.

But subsection (b) of Section 2 has another phrase that must be dealt with. It, in effect, states that a showing can be made that the challenged SPP has caused the political process "not to be equally open to participation" by black citizens if it can be shown that such black citizens "have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice."

So, if the challenged SPP does not cause black citizens to "have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice," then it cannot be found that such SPP causes the political processes "not to be equally open to participation" by black citizens and, therefore, finally, there would be no violation of the operative language in subsection (a), i.e., the "abridgement of the right of any citizen ... to vote on account of race."

Absent a showing that the challenged SPP resulted in the denial of a citizen's right to vote, the first inquiry, then, will be whether that SPP caused blacks to have "less opportunity than other members of the electorate to participate in the political processes and to elect representatives of their choice." [5]

B. To Make Out A Section 2 Violation, Must Plaintiffs Prove Both Of The Following: That, As A Result Of The 1981 District Lines, Blacks (1) Have Less Opportunity To Participate In The Political Processes; and (2) Have Less Opportunity to Elect Representatives of their Choice?

The defendants argue that, under the plain language of Section 2(b), plaintiffs must show two separate things: "(1) that they have less opportunity to participate in the political processes; and (2) that they have less opportunity to elect representatives of their choice." The majority opinion responds:

Moreover, the argument fails purely as a logical and linguistic matter. Even if plaintiffs failed to show less opportunity to participate in the political process, a showing that they have less opportunity *231 to elect candidates of their choice would suffice to establish their claim. The right protected is the aggregate of these opportunities — the right to effective participation in the political system....

I agree that statute would permit the interpretation placed on it by Judge Arnold if it were proper to speak of the "aggregate of these opportunities" as constituting the "right to effective participation in the political system." But the statute nowhere speaks of "the right to effective participation"; and the history of the act and the plain language of the statute strongly argue for the necessity of proving both.

If black citizens in the legislative district involved have as great an opportunity to vote and otherwise to participate in the political process as others, then their failure to elect candidates of their choice would be simply a function of the democratic process — no different from the fate of Republicans who consistently lose in a majority-Democratic district.

And, of course, if the black plaintiffs prove that the challenged SPP results in their having less opportunity to participate in the political process, then that would, in the usual case, automatically prove that that SPP also resulted in their having less opportunity to elect representatives of their choice. Ergo, a Section 2 violation. But the proof that the SPP results in blacks having less opportunity to elect candidates of their choice would not necessarily say anything about their opportunity to participate in the political process.

It is my view that Congress never intended that a simple showing that a challenged SPP resulted in a minority group's having less opportunity to elect representatives of their choice would suffice to establish a Section 2 claim in a districting case. Otherwise, it would be self-proving. If the district lines result in blacks being a minority voting age population (VAP), they clearly will have less opportunity than others to elect candidates. But it may be argued that the plaintiffs would at least have to show that, across the district lines, there were other blacks and that the lines could have been drawn to make the blacks the majority. But this is just judicial "fault" reasoning. How can blacks claim that the Board was at fault in drawing the lines so that they are a minority VAP in the district if the Board had no alternative? (This is no doubt the type of rationale that was responsible for the development of the Thornburg "preconditions." We lawyers and judges seek reasonable interpretations.) But, first, we must admit that under a "results" test any time district lines are drawn so that blacks are a minority VAP, those lines result in blacks having less opportunity than others to elect candidates of their choice. This is true whether or not the district lines could have been drawn to make blacks the majority VAP.

So I disagree with Judge Arnold. I believe that Congress intended to require proof in all cases that the challenged SPP resulted in blacks or Hispanics having less opportunity than others to participate in the political process. I do not agree that proof that the challenged SPP resulted in blacks having less opportunity to elect candidates of their choice will alone suffice.

C. Introducing the "Zimmer" and "Senate" Factors.

Having illustrated the difficulties with the grammar, language, and structure of Section 2, we are left to deal — preliminarily at this point — with the effect of the "Zimmer" or "Senate" factors in this interpretative process. As explained earlier, with respect to the Voting Rights Act's "results tests," the Senate purports to certify certain "typical factors" that may establish whether the challenged SPP has a discriminatory effect. Those factors are:

1. the extent of any history of official discrimination in the state or political subdivision that touched the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process;

2. the extent to which voting in the elections of the state or political subdivision is racially polarized;

3. the extent to which the state or political subdivision has used unusually large election districts, majority vote requirements, *232 anti-single shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority;

4. if there is a candidate slating process, whether the members of the minority group have been denied access to that process;

5. the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment and health, which hinder their ability to participate effectively in the political process;

6. whether political campaigns have been characterized by overt or subtle racial appeals;

7. the extent to which members of the minority group have been elected to public office in the jurisdiction;

Additional factors that in some cases have had probative value as part of plaintiffs' evidence to establish a violation are;

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;

whether the policy underlying the state or political subdivision's use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.

S.Rep., at 28-29, U.S.Code Cong. & Admin. News 1982, pp. 206-07. Since the 1982 amendments, federal courts have almost unanimously felt the need to make the findings suggested by these "factors" whether or not they had any real relevance to the issues at hand.

Substantive due process issues may arise when there is a lack of scientific or "true" relevance between such factors and the "fact" they are intended to establish. To make the point by an absurd example, suppose the Senate had added the following factor: "The extent to which members of the minority group drink coffee." Proof of this "factor" — one way or the other — could not give rise to any inference upon which it could properly be decided whether the minority group, because of the challenged SPP, had "less opportunity than other members of the electorate to participate in the political processes and to elect representatives of their choice." I suggest that some of the actual Senate factors may have no more relevance than this hypothetical "coffee drinking" factor. [6] It is my opinion that some federal courts have decided "Section 2 cases" on the basis of these factors without any but token reference to the precise language of that statute and, as a consequence, have reached conclusions at odds with the clear language thereof. See further discussion under Section V, infra.

Having identified some of the interpretive problems with which courts must attempt to deal in disposing of Section 2 claims, I turn to certain specific disagreements which I have with my brothers. But first: the outer limits.

II. CONSTITUTIONAL LIMITATIONS: USE OF "REPUBLICAN FORM OF GOVERNMENT" GUARANTEE OF ARTICLE IV, SECTION 4, U.S. CONSTITUTION.

Are there any constitutional provisions which protect certain voting or election standards, practices or procedures (SPPs) from effective challenge under Section 2, as amended? Stated otherwise, should federal courts in handling Section 2 challenges be sensitive to potential constitutional limitations or problems beyond those contained in the Fifteenth Amendment? It is my view that they should be. But rarely do we see evidence of such sensitivity. Even when enumerating a litany of "risks" involved in the application of Section 2, we do not find any suggestion of such constitutional limitations. Take the following language from the district court's opinion in Gingles v. Edmisten, supra at 345, 356:

*233 In making that political judgment, Congress necessarily took into account and rejected as unfounded, or assumed as outweighed, several risks to fundamental political values that opponents of the amendment urged in committee deliberations and floor debate. Among these were the risk that the judicial remedy might actually be at odds with the judgment of significant elements in the racial minority; the risk that creating "safe" black-majority single-member districts would perpetuate racial ghettos and racial polarization in voting behavior; the risk that reliance upon the judicial remedy would supplant the normal, more healthy processes of acquiring political power by registration, voting and coalition building; and the fundamental risk that the recognition of "group voting rights" and the imposing of affirmative obligation upon government to secure those rights by race-conscious electoral mechanisms was alien to the American political tradition.

The suggestion is, I suppose, that, if there are constitutional problems, they are only for the Congress — not the courts — to deal with. I disagree.

For example, consider the attack raised in Whitfield, supra, on the majority vote requirement. That attack is pertinent here too because the plaintiffs in this case also point to that requirement of the Constitution and various statutes of the state of Arkansas, in an effort to show not only discriminatory intent, but also that the majority vote runoff requirement, in fact, discriminates against blacks by virtue of its effect. I suggest that Senate Factor No. 3, unhappily, practically invites such an attack. That Senate Factor identifies "majority vote requirements" as one of those "voting practices or procedures that may enhance the opportunity for discrimination against the minority." The Senate Factor, in effect, asks courts to determine "the extent to which the state or political subdivision has used ... majority vote requirements...." The obvious suggestion is that if the state has used the requirement to a great extent, it should be assumed to have a discrimination-enhancing effect.

It is clear beyond question that the people of Arkansas have chosen, through their state constitution and through the acts of their state legislature, to use the "run-off," "majority vote" requirement in most primary and general elections.

Of course, in one sense, it can be stated that majoritarian democracy always discriminates against political minorities. But surely that is not the sense in which the Senate used its reference to "majority vote requirements." As explained in Whitfield, supra, run-off requirements do not have the effect of discriminating against blacks, even taken as a politically cohesive group. But let us assume that such requirements do have such an effect, as a panel of the Eighth Circuit has now decided in partially reversing this Court's Whitfield decision. See ADDENDUM, infra. It is fair to ask: are not some things "off limits" or "non-negotiable" in our political system? [7]

Mr. Justice Frankfurter cautioned the courts to steer clear of political questions. *234 His views are well described by Ms. Thernstrom:

What had actually been asked of this Court, he said, was nothing less than "to choose among competing bases of representation — ultimately, really, among competing theories of political philosophy."

Justice Frankfurter was convinced that courts and politics did not mix, and that political choices (beyond the competence of judges) were an inevitable component of every apportionment decision. That is, the means by which political influence was distributed inescapably suggested an end. To choose a particular distributive principle was to adopt a particular definition of democratic government. In every apportionment rule there lurked an implicit theory of representation, and no case of alleged malapportionment could be judged without such a theoretical framework. That is, the means not only suggested an end, but without a carefully delineated end, appropriate means remained uncertain. Lacking clearly articulated political values, Frankfurter asked, how could courts judge the equity of particular contested district lines? Without settled standards, how could judges weigh such competing considerations as the desirability of population equality, the value of an experienced incumbent, the need to honor community integrity or to provide a strong rural voice despite a sparse rural population?

Thernstrom at 64.

The Supreme Court nevertheless did get into the political thicket but only in a limited, non-comprehensive manner. The Court went far enough to decide that "one-person, one-vote" is required by the equal protection clause. This holding could be said to give real substance to the meaning of "Republican Form of Government," but the Court specifically eschewed Section 4 of Article IV, U.S. Constitution, as a predicate for its decision. The Court was able to announce this "one-person, one-vote" standard without precisely describing what practices are, and are not, essential to our democratic form of government. But, now, facing such attacks as the one upon the majority-vote requirement, can the courts still refuse to identify what is essential, and what is not essential, in our system of democratic government?

All of which leads me directly to a consideration of Article IV, Section 4 which provides, in part: "The United States shall guarantee to every State in this Union a Republican Form of Government...." Why has not this provision of our Constitution, which more clearly deals with the type of state government contemplated by the founders than any other, been the prolific source of great Supreme Court decisions, giving form and content to that which is essential in our democratic form of government? The answer will be found in a review of a few important decisions.

From Luther v. Borden 48 U.S. (7 How.) 1 , 12 L.Ed. 581 (1849) until Baker v. Carr, supra , it was accepted by the Justices of the Supreme Court that the Republican Form Clause can only be enforced by Congress. In Luther , the issue before the Court was which government of Rhode Island should be recognized as the legitimate state administration following the insurrection of 1841-42. The plaintiff brought a suit in trespass challenging the legality of a martial law decree under which the charter government's soldiers had invaded her home. She claimed the decree was invalid because the state government had been overthrown by popular convention. Justice Taney refused to review the case finding that only Congress had the power to decide questions of state legitimacy. Thus was born the political question doctrine.

Professor John Hart Ely opines that while Luther was correctly decided, it was "a gross mistake of logic to infer, as subsequent cases did, that all cases brought under the Republican Form Clause must therefore also present political questions." Ely, Democracy and Distrust p. 118 (1980).

The next large leap in the evolution of the political question doctrine came in Pacific States Telephone & Telegraph Company v. Oregon, 223 U.S. 118 , 32 S.Ct. 224 , 56 L.Ed. 377 (1912) in which Chief Justice *235 White found a utility company's challenge to a tax law passed by referendum to be a political question. Extending Justice Taney's reasoning, Chief Justice White declared that there was no judicial power to review any cases under the Republican Form Clause.

In Baker, supra, one of the seminal cases in the voting rights area, Justice Brennan concluded that state legislative apportionment was only justiciable under the Equal Protection Clause of the Fourteenth Amendment, the Republican Form of Government Clause being essentially a dead constitutional letter lacking standards to guide the Court's discretion. Justice Frankfurter in his dissent, commented that the Baker claim was, in essence, a "Guarantee Clause claim masquerading under a different label." He said:

The present case involves all of the elements that have made the Guarantee Clause cases non-justiciable. It is, in effect, a Guarantee Clause claim masquerading under a different label. But it cannot make the case more fit for judicial action that appellants invoke the Fourteenth Amendment rather than Art. IV, § 4, ...

Baker 369 U.S. at 297 , 82 S.Ct. at 754 .

In a rare post- Baker case, Kohler v. Tugwell, 292 F.Supp. 978, 985 (E.D.La.1968), a three judge panel reached the merits of the Republican Form Clause claim that a confusing ballot to adopt a state constitutional amendment was constitutionally offensive. While ultimately finding that the ballot was not constitutionally infirm Judge Minor Wisdom, writing for the majority, found that the Republican Form Clause was incorporated in the due process clause and noted in dictum: "Federal courts should be loathe to read out of the constitution as judicially nonenforceable a provision that the Founding Fathers considered essential to formulation of a workable federalism." (Emphasis in original)

Has not the time arrived for the courts to take up and interpret the meaning of "a Republican Form of Government?" Put otherwise, can we really avoid such an interpretive process in the light of attacks such as the ones being made in this case? If the federal courts order a ban on run-off elections, or order proportional representation, are they not sub silentio ruling that such orders are constitutional? (Surely they would not enter such orders if they believed their actions in so doing would offend the U.S. Constitution.)

The language of the clause itself urges that we challenge Congress' exclusive power to act under the Republican Form Clause. The clause states: "The United States shall guarantee to every State in this Union a Republican Form of Government." U.S. Const. art. IV, § 4. If the framers had intended to leave enforcement solely to the Congress, the clause should have specified "Congress" as the guarantor of such rights as it does in four other occasions in Article IV where Congress' power is exclusive. Instead, "the United States," presumably through all of its coordinate branches, is charged with this responsibility. Even Justice Frankfurter, while agreeing with what he viewed as a general prohibition against federal courts' becoming fora for "political debate," recognized that Article IV did not limit the enforcement thereof to the Congress. He states:

Art. IV, § 4, is not committed by express constitutional terms to Congress. It is the nature of the controversies arising under it, nothing else, which has made it judicially unenforceable. Of course, if a controversy falls within judicial power, it depends "on how he [the plaintiff] casts his action," ... whether he brings himself within a jurisdictional statute. But where judicial competence is wanting, it cannot be created by invoking one clause of the Constitution rather than another. When what was essentially a Guarantee Clause claim was sought to be laid, as well, under the Equal Protection Clause ... the Court had no difficulty in dispelling any mere confusion resulting from forms of expression, and considering the substance of things...."

Here appellants attack "the State as a State," precisely as it was perceived to be attacked in the Pacific States [ Telephone *236 & Telegraph Co. v. Oregon ] case, id. [ 223 U.S. 118 ], at 150, 32 S.Ct. [224] at 231 [ 56 L.Ed. 377 (1912)]. Their complaint is that the basis of representation of the Tennessee Legislature hurts them. They assert that a "minority now rules in Tennessee," that the apportionment statute results in a "distortion of the constitutional system," that the General Assembly is no longer "a body representative of the people of the State of Tennessee," all "contrary to the basic principle of representative government...." Accepting appellants' own formulation of the issue, one can know this handsaw from a hawk. Such a claim would be non-justiciable not merely under Art. IV, § 4, but under any clause of the Constitution, by virtue of the very fact that a federal court is not a forum for political debate. Massachusetts v. Mellon [ 262 U.S. 447, 485 , 43 S.Ct. 597, 600 , 67 L.Ed. 1078 (1923)], supra.

Baker 369 U.S. at 298 , 82 S.Ct. at 754 .

This last-stated, old view is weakening and should be reexamined. In any event, the idea that Section 4 of Article IV should be enforced only by the Congress is suspect. It should also be reexamined and, I suggest, abandoned.

The meaning of a "republican" government to the author of the clause, James Madison, indicates that the clause is the most logical source for judicial review of cases relating to voting. In The Federalist No. 39 at 100 (New Amer. Lib. ed. 1961), he stated that a republican government "derives all its powers directly or indirectly from the great body of the people." And the heated debates about the possible "tyranny of the majority" are known to all. The inference of majority election of representatives appears quite obvious. [8]

Justice Douglas, in his concurring opinion in Baker , also found the abdication of judicial power over all cases involving voting rights an absurd extension of the erroneous holding in Luther . In reviewing the Constitution for voting rights, Justice Douglas focussed on the Republican Form Clause. "So far as voting rights are concerned, there are large gaps in the Constitution. Yet the right to vote is inherent in the republican form of government envisaged by Article IV, Section 4 of the Constitution." Baker, supra at 242 , 82 S.Ct. at 723 , (Douglas, J., concurring) (Emphasis added).

Finally, the current awkward and unsatisfying analyses of reapportionment cases argue for a revival of the Republican Form Clause as a basis for review. Professor John Hart Ely writes:

In fact it seems likely that this unfortunate doctrine — that all Republican Form cases are necessarily cases involving political questions — will wholly pass from the scene one of these days. Friend and foe alike have come to recognize the obvious, that although the various state voting rights cases decided by the Warren and Burger Courts have been styled as equal protection decisions, they cannot comfortably be understood without a strong injection of the view that the right to vote in state elections is a rather special constitutional prerogative, a view that cannot be teased out of the language of equal protection alone and in textual terms is most naturally assignable to the Republican Form Clause.

Ely, supra, at 118. See also Bork, Neutral Principles and Some First Amendment Problems, 47 Ind.L.J. 1, 19 (1972).

I agree. And, although I believe that a proper interpretation of Section 2 as applied to the facts here avoids most constitutional problems, I am convinced that the majority's opinion does transgress constitutional limits by recognizing and enforcing "group affirmative political rights" and, in effect, requiring proportional representation *237 based upon race — all without an adequate legitimate factual predicate — in violation of the due process clause, the equal protection clause and Section 4 of Article IV.

III. HAVE PLAINTIFFS SHOWN THAT THE 1981 DISTRICTING PLAN RESULTS IN BLACKS HAVING LESS OPPORTUNITY THAN OTHERS TO PARTICIPATE IN THE POLITICAL PROCESS?

How does the majority deal with the defendants' argument that the plaintiffs cannot show that they "have less opportunity than other members of the electorate to participate in the political process." Judge Arnold, for the majority, appears to acknowledge that, "there are no presently existing legal barriers to voting by black citizens in Arkansas." Majority Op. at 204. But he nevertheless concludes that they do not have as much opportunity to participate in the political process as anyone else because of "the present effects of past racial discrimination, much of it official and governmental." Majority Op. at 204. The proof that he relies upon is principally supplied by the court's taking judicial notice of a finding in Smith v. Clinton, 687 F.Supp. 1310, 1317 (E.D.Ark. 1988) aff'd mem., ___ U.S. ___, 109 S.Ct. 548 , 102 L.Ed.2d 576 (1988), to wit:

We further find that the history of discrimination has adversely affected opportunities for black citizens in health, education and employment. The hangover from this history necessarily inhibits full participation in the political process.

I disagree with this type of "bootstrapping," but I also have more fundamental objections. First, Section 2(b) requires a cause and effect showing that the challenged SPP is responsible for blacks having "less opportunity than other members of the electorate to participate in the political process." The finding in Smith , means that neither present legal barriers nor the districting plan formulated by the Board of Apportionment in 1981 is responsible for blacks having "less opportunity than other members of the electorate to participate in the political process." Rather, it is the present effects of past discrimination in health, education and employment. Stated otherwise, the opportunity of blacks to participate in the political process would be the same regardless of the manner in which the Board drew the district lines. (This is different from the question whether the line drawing by the Board affected the opportunity of blacks "to elect candidates of their choice." I agree with the majority that in some instances the line drawing by the Board in 1981 did affect the opportunity of blacks to elect candidates of their choice. But it is very important analytically to deal separately with the issue of the opportunity "to participate" and the issue of the opportunity "to elect.")

The majority would probably reply, as certain other federal courts have, that we must look at the "interaction" of the SPP with impairments blacks suffer as a result of prior discrimination to determine if that SPP results in blacks having less opportunity to participate in the political process. This is linguistic legerdemain. Clearly it is not the line drawing by the board — the SPP here — which "results" in blacks having less such opportunity; rather, it is the diminished socio-economic status found to have resulted from prior discrimination. [9] And Section 2 does not purport to give a remedy solely on the latter basis. At the risk of argumentative overkill, assume the Board had drawn the district lines exactly as plaintiffs now request. Would the opportunity of blacks "to participate" be any different? *238 Clearly not. This error in analysis I identify as the source of so much unnecessary intellectual conflict in the "voting rights" cases and as the principal source of the misuse of certain of the Zimmer or Senate factors.

Is not all of this obvious? The socio-economic condition of blacks in any given area — whatever it is — will always be a "given" in any voting rights challenge. And, unfortunately in this country, from sea to sea, blacks, although making significant progress, still suffer adverse effects in education, the economic arena and health, and, on a statistical basis, are simply not as well off as non-blacks in our society. So, if one accepts that being poor, uneducated, unhealthy, etc., decreases one's "opportunity to participate" in the political process, then there will be no voting SPP which will be immune from attack. Elections per se could be as readily attacked! And, although no one has pushed that far yet, again note the plaintiffs' success in the Eighth Circuit Court of Appeals in attacking a "runoff, majority-vote" statute. Whitfield v. Democratic Party of State of Arkansas, supra .

This Court (the U.S. District Court) in its Whitfield opinion discussed this problem. After quoting Senate Factor No. 5, see supra, p. 1384, the Court stated:

This is also a "given" for the state of Arkansas and Phillips County and probably for every other political subdivision in the nation. But the effects are more devastating in Phillips County than in other places because of the dire economic circumstances that have developed in that area of the state over the past decade. Although the court is finding that blacks still bear the effects of discrimination in such areas as education, employment, and health, nevertheless, the Court also finds that those effects should not hinder their ability to participate effectively and equally in the political process. The Court also notes that typical factor # 5 refers to the "ability to participate" rather than "opportunity" to participate as stated in the statute. The statutory language, not being ambiguous, controls. The effects of discrimination referred to do not, in any legally significant way, hinder the "opportunity" or, indeed, the ability of blacks to participate effectively in the political process.

* * * * * *

And they can demonstrate that not having a telephone or an automobile makes it more difficult and less convenient for a citizen to qualify for, and to exercise, his or her voting rights. The tricky words are "difficult" and "inconvenient." .... But ordinary inconveniences such as one might experience if he wished to go to the doctor's office or to the post office or the general store should not be deemed to constitute legal barriers simply because the objective is to get to the voting place. The value one places on one's rights to vote will be reflected in the difficulties and inconveniences overcome in exercising that right.

Whitfield, supra, at 1384-85. See also Butts v. City of New York, 779 F.2d 141 , 149 and n. 4 (2d Cir.1985), cert. denied, 478 U.S. 1021 , 106 S.Ct. 3335 , 92 L.Ed.2d 740 (1986) (Section 2 does not employ a "difficulty of victory" test).

I conclude here, as I did in Whitfield, that the residual effects of past discrimination in such areas as education, employment and health do not in any legally significant way hinder the "opportunity" or, indeed, the "ability" of blacks to participate effectively in the political process.

So, since I agree with the defendants that plaintiffs, to succeed under Section 2(b), must, inter alia, prove the following two separate things, i.e., (1) that they have less opportunity to participate in the political process, and (2) that they have less opportunity to elect representatives of their choice, I conclude that the plaintiffs have failed to establish any Section 2 violations here. [10] However, if I agreed with the majority *239 that a simple showing that blacks have "less opportunity to elect candidates of their choice" would suffice, I would end up by concluding that plaintiffs have shown at least one Section 2 violation. See discussion, House District 100 in Section VIII C, infra. On the surface, this limited showing would appear easy to make, but the "preconditions" required by Thornburg alert us to the real difficulties. See Section V, infra. Before dealing with those preconditions, it is necessary to consider the "opportunity to elect" language in greater depth.

IV. HAVE PLAINTIFFS SHOWN THAT THE 1981 DISTRICTING PLAN RESULTS IN BLACKS HAVING LESS OPPORTUNITY THAN OTHERS TO ELECT CANDIDATES OF THEIR CHOICE?

This provision of Section 2 is comparatively straightforward. Of course, if race is used to define politics and we assume only two parties, blacks and whites, then if district lines are drawn so that blacks constitute a majority of the voting age population (hereafter "VAP"), we can say that the lines do not result in blacks having less opportunity to elect candidates of their choice. But if lines are drawn so that blacks are not a VAP majority, we can say that the lines as drawn result in blacks having less opportunity to elect candidates of their choice. Note that the socio-economic status of blacks is a constant. The variable is where the district lines are placed: (a) Majority Black or (b) Majority White. Does this mean that every district drawn so that blacks are a VAP minority violates Section 2? No, because before one gets to this question, he or she must ascertain whether the Thornburg preconditions are satisfied. See Section V, infra. But if the Thornburg preconditions are satisfied, then the result will follow automatically regardless of the socio-economic condition of the blacks or the whites in the area. There is nothing in Section 2 that states that if the blacks in the district are richer, better educated and healthier than whites, they will not be entitled to the benefits of the law.

While I have agreed with my brothers that the percentages of VAPs are critical in determining what is, or what is not, a "majority" or a "minority," I must nevertheless acknowledge that there is a certain lack of realism in that approach. In the areas of the state of Arkansas challenged in this lawsuit, as well as in practically all other areas of the United States, the number of persons who actually vote, black and white, is far below the number of those of voting age and also far below the number of those registered to vote. So, when we talk about political districts which have VAPs for blacks or whites ranging between 30 and 70 percent, we may be overlooking the reality that a much smaller percentage of black or white voters can determine the outcome of the usual election. We will take an example.

In Phillips County, Arkansas, according to the 1980 census figures, there were a total of 22,110 persons aged 18 and older, of which 11,542 were white and 10,393 were black. [11] So, the black VAP was 47 percent. Plaintiffs' Exhibit 31, Table 45.

In the Democratic preferential primary held on March 8, 1988, Mr. Whitfield, a black, ran against several whites and led *240 the ticket with 3,465 votes. The combined votes of his white opponents totaled 5,812. Plaintiffs' Exhibit 61pp. The evidence establishes strong racial voting polarization in Phillips County. Let us therefore assume that only blacks voted for Mr. Whitfield and only whites voted for his opponents. If that were the case, Mr. Whitfield received only 33.3 percent of the black VAP and the white candidates combined received 50.3 percent of the white VAP. Then, under Arkansas law, there was a runoff on March 22, 1988. At that runoff, Mr. Whitfield received 3,439 votes and his opponent, Mr. Stoner, a white, received 4,839 votes. Plaintiffs' Exhibit 61oo. On the assumption made above, Mr. Whitfield received 33.08 percent of the voting age black population and Mr. Stoner received 41.9 percent of the white voting age population. Thus, it will be seen that if Mr. Whitfield received 1,401 more black votes in the runoff, he would have won. That would have given him a total of 4,840 votes, which is only 46.57 percent of the voting age blacks in the county.

The point is: because of the dismally low participation of Americans, black and white, in exercising their voting rights, a small, but active, minority could easily win most local races over a much larger nonparticipating majority. However, if racial polarization is established, then one could not safely predict that an actively participating minority would consistently overcome the majority. The minority might "slip up" on the majority one or two times, but the suggestion is that the majority would soon respond in kind. This rationalization at least tends to justify our choice of VAPs as the critical figures because those figures tend to set "per possibility" outside limits. [12] But, when the VAPs of blacks or whites range from 45 to 55 percent, say, the fact that one or the other group might be the "minority" on a VAP basis will not alone explain the success or failure of the black or the white candidates who have opposed each other in elections past.

For the reasons I detail in Section VIII of this opinion, infra, I find that the plaintiffs have shown that the 1981 district lines did result in blacks having less opportunity than others to elect candidates of their choice with respect to seven of the House districts. However, since, as I explain elsewhere, that is not all plaintiffs have to prove to establish Section 2 claims, I conclude that plaintiffs have not met their burden even with respect to these seven House districts. See discussion, Section III, supra.

V. THE THORNBURG FACTORS.

The majority opinion herein identifies the Thornburg factors required to be established as a predicate for a vote-dilution claim:

In evaluating a Section 2 claim, the Court must first determine whether three preconditions to a legally substantial impairment of plaintiffs' ability to elect the candidates of their choice have been met. First, the claimant minority must establish that it is sufficiently large and geographically compact to constitute a majority in a single-member district. Second, the plaintiffs must show that the group to which they belong is politically cohesive. Thornburg, 478 U.S. at 50-51 , 106 S.Ct. at 2766-67 . Third, the minority voters must show that the "majority votes sufficiently as a bloc to enable it — in the absence of special circumstances, such as a minority candidate running unopposed ... to defeat the minority's preferred candidate." Id. at 51 , 106 S.Ct. at 2766 . The latter two factors can be shown by proving that voting in the jurisdiction is highly racially polarized. Ibid. Accord, Smith, 687 F.Supp. at 1314-1315 .

Although the majority aptly states the three critical factors, it does not, in my opinion, rigorously apply them to the facts of this case; nor does it adequately explore the interrelationships among those factors.

*241 A. Political Cohesiveness.

1. Black Voting Behavior.

It is clear that the majority opinion does not conduct the intensely fact-specific appraisal required by Gingles. It has not systematically studied the separate groups of black citizens projected for incorporation into the proposed new districts to determine if they, taken together, constitute a "politically cohesive" group. This, however, is understandable.

The Section 2 law in this "gerrymandering" (districting or redistricting) area developed as a negative proscription. The law prohibits "fragmenting" or "packing" cohesive political groups of black citizens with the effect of diluting their voting power. [13] The law did not develop in the context of affirmative obligations to reach out and include larger numbers of black citizens in order to enhance black political power. In other words, the mandate was "you shall not harm" rather than "you shall help." When dealing with the negative proscription, the issues were simple and straightforward. If blacks possessing majority political power in a single district were divided into two districts, in each of which they constituted a minority, everything was quite simple and straightforward. Or, if one, say, took two 60% black VAP districts and divided the black VAP so that one district had 80% and the other 40%, the case would also be clear. And the multi-member-at-large district is likewise a relatively simple case.

But here, plaintiffs made no direct effort to show that district lines were drawn in 1981 in a way which had the effect of packing or breaking up or fragmenting prior existing politically cohesive black groups. I waited in vain for evidence comparing the political status of blacks in the districts created in 1971 with their political status in the districts created in 1981. I asked myself: did the new lines drawn in 1981 decrease, limit, curtail or lessen, the effectiveness of black political participation by fragmenting or packing or otherwise? But plaintiffs eschewed this approach in favor of a theory that they need only show that the defendant Board in 1981 failed to take the opportunity to enhance black political effectiveness. The plaintiffs' view apparently is that it is immaterial whether they were better off politically under the 1981 plan than under the 1971 plan. Their point is that they were not better off under the 1981 plan than they could have been if the Board had taken advantage of every opportunity to enhance their political position. This is clearly an affirmative action theory. Affirmative action by definition carries implications of intent which may relate to plaintiffs' constitutional claims but have only tangential relevance to their Section 2 "results" claims.

The plaintiffs, in the main, simply had their experts identify the geographic locations of black citizens. Then they drew lines incorporating such black citizens into new districts in which the blacks would constitute a majority of the voting age population without any showing that those blacks represented a politically cohesive group. For the most part, the plaintiffs simply asked the court to assume that all black citizens incorporated into the new district would think alike and vote alike.

*242 But I steadfastly resist the notion that we should presume — or that the law permits us to assume — that all blacks, regardless of their differing geographic and political backgrounds, will vote as a racial bloc whenever a black candidate is opposing a white candidate; or, for that matter, that all white citizens (likewise brought together for the first time in the new district) will respond in kind. I suggest that such assumptions represent stereotyping at

This text is long and has been trimmed here. Open the source document for the complete record.