Hays v. State of La.

MEMORANDUM OPINION

WIENER, Circuit Judge:

Plaintiffs Ray Hays, et al. (Plaintiffs) challenge the congressional redistrieting plan (the Plan) adopted by the Louisiana State Legislature (the Legislature) when, following the 1990 census, it enacted Act 42 of 1992 (Act 42).¹ In its present posture, this case considers the constitutionality of the Plan, admittedly designed inter alia to increase the number of black representatives in Louisiana’s congressional delegation from one out of eight to two out of seven. The Plaintiffs insist that the Plan accomplishes this result by employing impermissible racial gerrymandering to create a new majority-black voting district, thereby violating the Equal Protection Clause of the United States Constitution.

In simplest form, this case poses the question, “Does a state have the right to create a racial majority-minority congressional district by racial gerrymandering?” In simplest form, the answer — largely supplied by the United States Supreme Court’s opinion in Shaw v. Reno², rendered during the pendency of this ease — is “Yes, but only if the state does it right.”

Finding that the Plan in general and Louisiana’s Congressional District 4 in particular are products of racial gerrymandering and are not narrowly tailored to further any compelling governmental interest, we conclude that the Legislature did not “do it right.” We hold, therefore, that Plaintiffs’ right to equal protection as guaranteed by the United States Constitution is violated by the Plan. Consequently,, we declare Act 42 of 1992 to be unconstitutional and the redistricting plan embodied therein to be null and void; enjoin the State of Louisiana from holding any future congressional elections based on the Plan; and, although we do not invalidate the 1992 congressional elections held thereunder, hold that the term of office of each member of the United States House of Representatives from Louisiana who represents a district created under the Plan shall expire, ipso facto, at noon on the 3rd day of January, 1995,³ such terms of office not to be extended or carried over into the next Congress in any manner whatsoever.

I

STATEMENT OF THE CASE

The Plan comprises five majority-white districts (Districts 1, 3, 5, 6 & 7) and two majority-black districts (Districts 2 and 4).⁴ District 2 has a minority voting age population of 56%, while District 4 has a minority voting age population of 63%. New District 2 covers essentially the same geographic area as did old District 2 in the previous plan: almost the entire population of Orleans Parish and roughly one-third of the population of Jefferson Parish. District 2 is not challenged in the instant lawsuit, nor elsewhere to our knowledge. Rather, District 4, appearing for the first time in Act 42, is the primary focus of this constitutional challenge.

Shortly after the enactment of Act 42, the Plaintiffs filed this suit seeking to have 1) the Plan declared unlawful, 2) the Defendants enjoined from using the Plan in the impending congressional elections, 3) the Legislature ordered to create a new plan that would not segregate state residents into voting districts on the basis of race, and 4) the Defendants. preliminarily enjoined “from taking any action in preparation for the primary or general elections for the U.S. House of Representatives____” In their complaint the Plaintiffs — who are black, white, and Asian residents of either District 4 or District 5— allege that the Plan violates Section 3 of the Louisiana State Constitution, Section 2 of the Voting Rights Act, and the Fifth, Fourteenth and Fifteenth Amendments of the United States Constitution.

As required by the Voting Rights Act, the chief judge of the United States Court of Appeals for the Fifth Judicial Circuit appointed the instant panel to hear this case. It was tried on August 26 and 27, 1992 (the Trial). At the conclusion, of the Trial, we issued an interlocutory Memorandum Ruling and Order (the 1992 Order) in which we denied Plaintiffs’ request for an injunction, allowed the 1992 Congressional elections to go forward under the Plan, refused to consider Plaintiffs’ state.and federal constitutional claims, and took Plaintiffs’ Voting Rights Act claims under advisement. We also requested post-trial briefing on the question whether the Plan dilutes, the voting strength of either blacks or whites in contravention of the Voting Rights Act.

Being aware of an essentially identical case originating in North Carolina — now known as Shaw v. Reno — which had progressed further than had the instant case, we held this one under submission until the results of Shaw became known. On June 29, 1993, the Supreme Court rendered its decision in Shaw, profoundly affecting this case and similar ones pending in other states. We requested the parties and invited amici curiae to file supplemental briefs, generally discussing the implications of Shaw and specifically addressing whether — in the terminology of Shaw — the Plan was “narrowly tailored to further a compelling government interest.” In August 1993, we held two additional days of trial (the Evidentiary Hearing) to assist us in determining whether the Plan is in fact the product of racial gerrymandering, and, if so, whether it should nonetheless be sustained because it is narrowly tailored to further a compelling state interest.

II

PRELIMINARY LEGAL MATTERS

A. Standing

Early in this case Defendants suggested that the white Plaintiffs intrinsically lack standing to challenge the Plan. The Defendants evidently believe that only historically disadvantaged minorities have standing to attack state laws that segregate citizens on the basis of race. But that is not the case.

In Shaw, the Supreme Court reaffirmed the important principle “that equal protection analysis is not dependant on the race of those benefited or burdened by a particular classification.”⁵ “The guarantees of the Fourteenth Amendment extend to all persons. Its language is explicit: ‘No state shall ... deny to any person within its jurisdiction the equal protection of the laws.’ The guarantee of equal protection cannot mean one thing when applied to one individual and something else when applied to a person of another color. If both are not accorded the same protection, then it is not equal.”⁶ White citizens thus clearly have standing to challenge redistricting plans under the Equal Protection Clause of the Fourteenth Amendment, just as do black citizens, Hispanic citizens, Asian citizens, Native American citizens, and citizens of any other race. We reject out of hand the implication that, although all are equal under the law, “some ... are more equal than others.”⁷

B. Survival of Plaintiffs’ Equal Protection Claim

The Defendants assert that the Plaintiffs’ equal protection claim does not conform to the structure of the argument approved by the Court in Shaw and, consequently, that the Plaintiffs fail to state an equal protection claim under Shaw. The Defendants also argue that our 1992 Order, in which we denied Plaintiffs’ state and federal constitutional claims, effectively disposed of the Plaintiffs’ Equal Protection Clause claim with finality; and that we therefore may only consider the Plaintiffs’ claim that the Plan violates the Voting Rights Act. We disagree with both of these assertions.

Although Shaw had not yet been decided at the time the Plaintiffs initiated the instant suit, the allegations in the Plaintiffs’ complaint are clearly sufficient to make out a cognizable claim of racial gerrymandering under Shaw. The narrow holding of Shaw is that a citizen states a claim under the Equal Protection Clause by alleging that the reapportionment scheme adopted by his state is so irrational on its face, so bizarrely shaped and convoluted, “that it can only be understood as an effort to segregate voters into separate voting districts because of their race____”⁸ There is no question that the Plaintiffs make sueh allegations in their complaint.

For example, the Plaintiffs allege that “[t]he adoption of Act 42 ... has resulted in the creation of at least two of seven districts which are devoid of any commonality of interest ... geographical compactness, eontiguousness [sic], consistency with existing political, societal, governmental or economic districts or jurisdictional boundaries, other than the racial designation of the majority therein.” Similarly, the plaintiffs allege that “Act 42 created two amorphous districts which embody a scheme for segregation of voters by race in order to meet a racial quota for representation of the State of Louisiana in the United States House of Representatives.” And again, Plaintiffs allege that “[t}he defendants enacted and intend to implement Act 42 with the intent to create a Congressional Plan concentrating voters of a particular race in designated districts____” We find that these statements clearly amount to allegations that the State of Louisiana “adopted a reapportionment scheme so irrational on its face that it can be understood only as an effort to segregate voters into separate voting districts because of their race.”⁹ We conclude that the Plaintiffs have stated an Equal Protection Clause claim essentially identical to the claim alleged by the plaintiffs in Shaw.

Moreover, our 1992 Order, denying Plaintiffs’ state and federal constitution claims, worked a temporary dismissal of Plaintiffs’ equal protection'- claim, not a final one. It was clearly an interlocutory ruling.

In an order dated June 29, 1993 (1993 Order) we requested supplemental briefing to clarify the relevance of Shaw to the instant case and to appear at the Evidentiary Hearing to elucidate whether the Plan was “narrowly tailored to further a compelling government interest,” as required by the Court in Shaw. Of necessity our 1993 Order modified our 1992 Order and reactivated the Plaintiffs’ equal protection claim in light of Shaw. As such a resuscitation is specifically contemplated and authorized by Rule 54(b) of the Federal Rules of Civil Procedure,¹⁰ the Plaintiffs’ equal protection claim persists and demands adjudication by this court.

Ill

FINDINGS OF FACT AND CONCLUSIONS OF - LAW

As the findings of fact and conclusions of law in this case are inextricably intertwined, we do not present them in separate sections. Such separate presentation would increase the length and redundancy of our discussion. Rather, our language will indicate whether we find a particular observation to be a finding of fact or a conclusion of law. To the extent that a finding of fact is also a conclusion of law, we adopt it as both a finding of fact and a conclusion of law. To the extent that a conclusion of law is also a finding of fact, we also embrace it as both a conclusion of law and a finding of fact.

A. Racial Gerrymandering: General

Plaintiffs allege that the Plan is a product of racial gerrymandering. Consequently, as noted above, this case falls squarely within the ambit of Shaw v. Reno. In Shaw, the Court held that plaintiffs state a justiciable claim -under the Equal Protection Clause by alleging that a redistricting scheme is “so irrational on its face that it can be understood only as an effort to segregate voters into separate voting districts because of their race____¹¹ ” As Shaw was an appeal of the district court’s dismissal of the plaintiffs’ case for failure to state a claim, however, the Court did not have to resolve that claim; it had only to recognize it. Thus, the roadmap sketched by the Court — as helpful as it is— leaves some questions to be answered in cases such as this.

1. Racial Gerrymandering Defined

A legislature creates a racially-gerrymandered districting plan when it intentionally draws one or more districts along racial lines or otherwise intentionally segregates citizens into voting districts based on their race.¹² Thus, “racial gerrymandering” refers to the intentional, not the accidental, segregation of voters on the basis of race.¹³

2. Racially Gerrymandered Plans are Subject to Strict Scrutiny

The bedrock principle underlying the Court’s decision in Shaw is that racially gerrymandered redistricting plans are subject to the same strict scrutiny that applies to other state legislation classifying citizens on the basis of race.¹⁴ As such, racially gerrymandered plans violate the Equal Protection Clause of the' Fourteenth Amendment unless

they are narrowly tailored to further a compelling governmental interest.¹⁵

Such plans receive “careful scrutiny under the Equal Protection Clause regardless of the motivations underlying their adoption.” ¹⁶ This intense scrutiny is justified by the grave danger that is posed to our constitutional order and national community when a state creates and administers laws based on the race of its citizens.¹⁷

3.Racial Gerrymandering May be Proved Either Inferentially or Directly

We have already noted the narrow holding of Shaw: ' a plaintiff may state a claim under the Equal Protection Clause by alleging that the reapportionment scheme adopted by his state is so irrational on its face “that it can only be understood as an effort to segregate voters into separate vot> ing districts because of their race....”¹⁸ Shaw primarily deals with the problem of proving racial gerrymandering indirectly or inferentially. Facial gerrymandering — says the Court in Shaw — can be inferred when districts are so bizarrely shaped that they presumptively bespeak an impermissible purpose.

But racial gerrymandering may — a fortiori — also be proved by direct evidence that a legislature enacted a districting plan with the specific intent of segregating citizens into voting districts based on their race. If everyone — or nearly everyone — involved in the design and passage of a redistricting plan asserts or concedes that design of the plan was driven by race, then racial gerrymandering may be found without resorting to the inferential approach approved by the Court in Shaw.¹⁹ The Court recognized in Shaw that “[n]o inquiry into legislative purpose is necessary when the racial classification appears on the face of the statute.”²⁰ The same is equally true when virtually unanimous, essentially uncontroverted direct trial evidence establishes racial classification, as it did here. In this case, we find overwhelming evidence — both indirect and direct — that the Plan is a product of racial gerrymandering.

4. Intent Distinguished from Motive

In a brief aside, we draw on the familiar crime of homicide as a didactic analogy to clarify the important distinction between intent and motive for purposes of this ease. By definition, one who knowingly cocks, aims, and fires a loaded gun at another has the intent to kill or cause great bodily harm. That is the purpose for shooting. But the shooter may have any number of motives for intentionally shooting the victim: to eliminate a romantic rival; to collect insurance proceeds; to avenge some actual or perceived wrong; to repel aggression; to prevent the victim from perpetrating a crime or misdeed; and on and on. The applicable motive is the goal sought to be accomplished through the intentional killing of the victim.

We belabor the distinction between intent and motive because it provides a more recognizable way of looking at two separate aspects of this case: 1) the intent of the Legislature in creating the Plan on the basis of race (racial gerrymandering); 2) the motives of the Legislature or of individual or groups of legislators — whether compelling or not— for engaging in racial gerrymandering. Whatever the motivations of the Legislature or of the individual legislators who passed the Plan, the evidence overwhelmingly indicates that the specific intent of the Legislature — as an independent, collective organism — was indisputably to enact a plan that included two black and five white majority districts.

At the Evidentiary Hearing so much of the testimony purporting to discuss the intent of the Legislature, or those who sought to influence Louisiana’s 1992 redistricting, confounded and confused intent or purpose with motive or goal. More telling (and more candid) was the testimony at the Trial, a year earlier, concerning the motive for creating a second majority-black congressional district. That testimony differed markedly from its counterpart at the Evidentiary Hearing. In both proceedings the testimony on motive was considerably less uniform than was the virtually unanimous testimony regarding the Legislature’s intent to create a second safe, black majority district. At least by implication, however, four immutable elements provided a common point of departure for all witnesses: 1) Louisiana had an unavoidable legal obligation to reapportion its congressional delegation; 2) reapportionment had to be in strict compliance with the constitutional imperative of one-person, one-vote, meaning that each of Louisiana’s seven congressional districts had to contain roughly 603,000 residents; 3) given its failure to adopt a reapportionment plan in 1991, the Legislature would have to adopt a plan in its 1992 session that would be certain to receive immediate preclearance so that the congressional election, scheduled for the fall of that year, could be held; and 4) to obtain timely preclearance, i.e., voluntarily and not by court decree, any plan would have to include two safe, black majority districts.²¹

B. Shaw v. Reno: Inferential Proof of Racial Gerrymandering

Shaw deals primarily with proving racial gerrymandering inferentially. We proceed to examine the indirect or circumstantial evidence of racial gerrymandering in the instant ease.

1. The Shifting Evidentiary Burden

The Court in Shaw apparently intended to establish an evidentiary “minuet” for racial gerrymandering eases analogous to the one established by McDonnell Douglas v. Green and Texas Department of Community Affairs v. Burdine in the Title VII context.²² After a citizen establishes .a cause of action by alleging that the reapportionment scheme adopted by his state is so irrational on its face that it can only be understood as an effort to segregate voters on the basis of race — thereby creating a presumption of unconstitutionality²³ — the burden shifts to the state to proffer a legitimate, non-racial explanation for the irrationally shaped districts in its plan.²⁴ If, per chance, the state should answer,' generally denying gerrymandering, and move for summary judgment, the plaintiff presumably would need to support such a presumption with summary judgment evidence, e.g., affidavits and depositions demonstrating legislative intent, violation of redistricting principles, or the like. But if the state advances a legitimate, non-racial explanation for the plan’s irregularity, then the factfinder must — as always-^weigh the evidence on both sides and decide whether the plaintiff has met his burden of demonstrating that the plan’s irrational shape reflects racial gerrymandering.²⁵

Although the Court in Shaw does not discuss the respective burdens borne by the parties in this shifting protocol, Supreme Court precedent clearly indicates that plaintiffs have the ultimate burden of proving — by a preponderance of the evidence — that the irregularity of the challenged districts re-.' fleets racial gerrymandering.²⁶ Proof on the. merits would likely focus on the pretext of the state’s proffered non-discriminatory reasons. Nevertheless, placing the burden of proof on the plaintiffs reflects a basic tenet of equal protection analysis: there can be no violation of the Equal Protection Clause unless those who complain demonstrate that the state has engaged in purposeful discrimination.²⁷ Thus, in this case, Plaintiffs have the burden of demonstrating by a preponderance of the evidence that the Legislature was “motivated by racial considerations” when it adopted Act 42.²⁸ We find that the plaintiffs have met that burden — comfortably.

2. Plaintiffs’ Cause of Action

In addition to stating a cause of action by alleging extreme irregularity of shape, a plaintiff may, under Shaw, strengthen the inference that the state engaged in constitutionally suspect racial gerrymandering by demonstrating the state’s disregard of traditional districting principles.²⁹ As discussed above in section II B, the Plaintiffs have clearly alleged both that the Plan— particularly District 4 — is highly irregular on its face, and that the Plan disregards traditional districting principles.³⁰ We therefore conclude that the Plaintiffs established a cause of action essentially identical to that alleged by the plaintiffs in Shaw. It consequently devolved upon the Defendants either to refute the allegations directly or to provide legitimate, non-raeial explanations for the Plan’s irrational shape and its disregard of traditional districting criteria.³¹

3. Defendants’ Proffered Justifications

Apparently recognizing the impossibility of directly refuting high irregularity and violation of traditional districting criteria, the Defendants attempted to justify the Plan. During the entire Trial, not one witness ever suggested that the Legislature’s intent in creating District 4 was anything other than to create a second district with a super-majority of black voters by drawing its boundaries along racial lines. Indeed, most of the testimony of the Defendants’ witnesses concerned how numerically large a black majority was needed to ensure the effectiveness of the Plan, which everyone conceded to be a product of racial gerrymandering.

At the Evidentiary Hearing, however, following as it did on the heels of the Supreme Court’s decision in Shaw, the Defendants— with the benefit of hindsight — attempted to deny the racial gerrymandering that they so candidly proclaimed and avidly defended a year earlier at the Trial. Specifically, the Defendants now suggest that two non-raeial factors played key roles in the creation of the Plan: partisan/incumbent politics and socioeconomic commonalities. As discussed in the following section, we find the Defendants’ explanations wholly unconvincing and in many respects disingenuous.

4. Choosing Between the Competing Inferences

Because of the procedural posture of Shaw, the Court did not have to go beyond reversing the district court’s grant of the defendants’ motion to dismiss. Following a full merits trial here, however, we are obliged to decide whether the Plaintiffs have carried their burden of proving — by a preponderance of the evidence — that the Plan’s irregularity and its disregard of traditional redistricting principles reflect racial gerrymandering. As we find that the only sensible explanation for the Plan’s extreme facial irregularity and its flagrant deviation from traditional districting criteria is that the Legislature intentionally segregated voters into congressional'districts based on their race, we conclude that the Plaintiffs have overwhelmingly satisfied their burden of proving racial gerrymandering.

a. Plaintiffs’ Inferential Evidence

1. The Plan is Highly Irregular

The most cursory inspection of the districts established by Act 42 reveals several of them to be irregular, and District 4 to be highly irregular.³² Like the fictional swordsman Zorro, when making his signature mark, District 4 slashes a giant but somewhat shaky “Z” across the state, as it cuts a swath through much of Louisiana.³³ It begins north of Shreveport — in the northwestern corner of Louisiana, just east of the Texas border and flush against the Arkansas border — and sweeps east along that border, periodically extending pseudopods southward to engulf small pockets of black voters, all the way to the Mississippi River. The district then turns south and meanders down the west bank of the Mississippi River in a narrow band, gobbling up more and more black voters as it goes. As it nears Baton Rouge, the. district juts abruptly east to swallow predominantly black portions of several more parishes. Simultaneously, it hooks in a northwesterly arc, appropriating still more black voters on its way to Alexandria, where it selectively includes only predominantly black residential neighborhoods. Finally, at its southern extremity, the district extends yet another projection — this one westward towards Lafayette — adding still more concentrations of black residents. On the basis of District 4’s physiognomy alone, the Plan is thus highly irregular, suggesting strongly that the Legislature engaged in racial gerrymandering.³⁴

2. The Plan Violates Traditional Redistricting Principles

In Shaw, the Court reiterates the notion that states are not constitutionally required to adhere to the traditional redistricting principles of compactness, contiguity, respect for established political subdivisions, and commonality of interests.³⁵ The Court also observes that a state’s adherence to traditional districting criteria “may serve to defeat a claim that a district has been gerrymandered on racial lines.”³⁶ Yet a state’s disregard of such- criteria, emphasizes the Shaw opinion, may be evidence of constitutionally-suspect racial gerrymandering.³⁷

Indeed, the Court in Shaw refers to a situation “in which a State concentrated a dispersed minority population in a single district by disregarding traditional districting principles ...” as a circumstance in which racial gerrymandering can be easily inferred, precisely because the state ignored such principles.³⁸ As a state’s decision to disregard traditional redistricting criteria is probative'of constitutionally-suspect racial gerrymandering, we closely examine the Plan’s adherence (or, more accurately, non-adherence) to those principles; and we find that, for the most part, the Plan cavalierly disregards them.

a. Compactness

District 4 snakes narrowly across Louisiana soil from end to end for more than 600 miles.³⁹ A rectangle superimposed on the Z-shaped figure formed by District 4 would overlay two-thirds of the State. Additionally, as it winds along its erratic path, District 4 projects myriad diverticulae to encapsulate small sacs of otherwise widely dispersed black voters. No one could claim that District 4 is compact, at least not with a straight face.

b. Contiguity

District 4 was confected to satisfy the traditional districting criterion of contiguity, but only hypertechnically and thus cynically. When displayed on a map of the State, the district’s boundaries seem several times to narrow to a single point. This impression reflects reality, for at some places along its attenuated path, District 4 is no more than 80 feet wide. Such tokenism mocks the traditional criterion of contiguity.⁴⁰

e. Respect For Political Subdivisions

As one witness explained at the Evidentiary Hearing, there is no more fundamental unit of societal organization in the history of Louisiana than the parish.⁴¹ Whereas the previous congressional plan divided only seven of modern Louisiana’s 64 parishes, scattering fractions, of the same parish in more than one congressional district, Act 42 splits and scatters 28. District 4 is particularly aggressive in violating the boundaries of these traditional political and governmental units of the State: of the 28 parishes touched by District 4, only four whole parishes are included; but the district annexes only shards of 24 additional parishes, usually incorporating none but the predominantly black fragments of those shattered regions. Additionally, for the first time in Louisiana history, with the advent of Act 42 a districting plan violates the boundaries of nearly all major municipalities in the State. With the exception of Lake Charles, the Plan fragments all major municipalities into more than one congressional district, thereby destroying the common representation historically enjoyed by residents of the same municipality.⁴²

d. Commonality of Interests

Within its irregular boundaries, District 4 subsumes bits of every religious, ethnic, economic, social, and topographical type found in Louisiana.

(i)Religion and Ethnicity

District 4 violates the traditional north-south ethno-religious division of the State. Along its circuitous route, this new district combines English-Scotch-Irish, mainline Protestants, traditional rural black Protestants, South Louisiana black Catholics, Continental Freneh-Spanish-German Roman Catholics, sui generic Creoles, and thoroughly mixed polyglots, each from an historically discrete and distinctive region of Louisiana, as never heretofore so extensively agglomerated.

(ii)Economic Base

Cotton and soybean plantations, centers of petrochemical production, urban manufacturing complexes, timberlands, saw mills and paper mills, river barge depots, and rice and sugarcane fields are strung together to form the eclectic and incoherent industrial base of District 4. These diverse segments of the State economy have little in common. Indeed, their interests more often conflict than harmonize.

(iii)Geography and Topography

Red clay hills and pinelands, hardwood bottomlands and forests, alluvial floodlands, coastal plains, marshes, swamps and wetlands — all are present in District 4, which— as noted above — stretches more than half a thousand miles from end to end. What did the Legislature intend when it created District 4 — this non-traditional, little-in-common “un-district”? The indirect evidence overwhelmingly indicates that the Legislature specifically intended to create a plan with at least two majority black districts. Given the pre-existence of District 2, we find beyond cavil that the Legislature accomplished this by parceling voters into the remaining six districts on the basis of race.

b. Defendants’ Counter-Proof

In contrast, we have been shown no credible evidence supporting the defense witnesses’ proffered motivations of party and incumbency protection and socioeconomic commonality. Their explanations ring hollow. We find them to be no more than disingenuous, post hoc rationalizations.

In particular, we find that neither partisan nor incumbency politics was a significant factor in the core decision intentionally to create a plan containing a second black majority district. At the Trial, the Defendants never suggested that partisan or incumbent politics played a role in the determination to create District 4.⁴³ Indeed, Defendants’ counsel objected to Plaintiffs’ counsel’s questions aimed at eliciting testimony about the Black Caucus-Republiean Caucus alliance' that supported Act 42, arguing that such testimony was political and therefore irrelevant to the case. Moreover, even though party and incumbency protection play some role — large or small — in virtually everything done in the legislative branches (and executive branches, for that matter) of our contemporary governments, this universal truism does not negate the compelling inference that Act 42 in general, and District 4 in particular, are products of racial gerrymandering. Additionally, without for a moment granting that incumbency polities played a significant role in the decision to create a second majority black district, we question whether the Defendants could prevail in this litigation even if it had.

The Defendants seem to believe that they can defeat a claim of racial gerrymandering under Shaw if any factor other than race played any cognizable role in the creation of a challenged redistricting plan. Although we need not correct such a misconstruction of Shaw to find racial gerrymandering in this case, we briefly address this legal issue — obiter dictum — in hopes of shedding a little additional light on this difficult area of the law.

The Defendants evidently base their belief — that the presence of any non-racial motivating factor will excuse racial gerrymandering — on language found at the end of the Shaw opinion. There the Court indicates that a plaintiff states a claim under the Equal Protection Clause by alleging that a reapportionment plan is so irrational on its face “that it can be understood only as an effort to segregate voters----”⁴⁴ This emphatically does not mean that if any other factor influenced the legislature the plaintiff is unable to establish a racial gerrymander. Rather, it means that if the contours and content of a redistricting plan can be wholly explained to be the product of one or more factors other than race, then the defendants have created a competing inference. The court must then weigh the competing inferences — as indeed it usually must — to decide whether the plaintiff has proved his inference by a preponderance of the evidence.⁴⁵ Thus, accurately stated, the question posed by Shaw is whether a redistricting plan can be reasonably conceived as the product of nonracial factors. In this case the Plan cannot.⁴⁶

No one claims that the contours of District 4 can be wholly and alternately explained as a product of partisan or incumbency politics. Every single witness who addressed the issue either proclaimed or acknowledged that the creation of a second black majority district was the primary factor — or at least a substantial and important factor — in the creation of the Plan. We agree. There is no way that a rational fact-finder — looking at the map and reviewing the credible evidence with care — could conclude that Act 42 can be explained entirely without reference to racial gerrymandering.

At the Evidentiary Hearing, Dr. Alan Liehtman, an expert witness for the Defendants, opined that District 4 — which appears so violative of traditional redistrieting principles — actually possesses socioeconomic commonality and coherence. Dr. Liehtman contended that District 4 was distinguishable from other districts because its residents were relatively poor, relatively under-educated, and owned fewer telephones and automobiles than did the residents of other districts established by the Plan. The Defendants offered this evidence to suggest “a rational basis [for District 4] other than race.” ⁴⁷ We do not gainsay the conclusions of this witness, but we disagree that they have significance.

Any able statistician who looks at enough statistical characteristics (multivariate analysis) can find something distinctive about any district. In this case District 4 was found to be fairly poor, although poverty is not'particularly distinctive in any region of Louisiana. Had District 4 been a fairly wealthy district, the Defendants’ expert could have opined that the citizens of District 4 shared the common interest of wealth. And had District 4 fallen right in the middle, he could have explained that District 4 was distinctive in being the most solidly middle class. But all these observations are irrelevant because we find them to rise to no level higher than post hoc rationalization ⁴⁸

The Defendants admit that the socioeconomic profiles of the Plan’s districts were not actually used by the Legislature: the census data used in Defendants’ statistical analyses were not even available to the Legislature when it passed Act 42. More specifically, freshman Louisiana State Senator Tom Greene testified that no socioeconomic data was submitted with the various redistricting plans when they were considered by the Legislature. Thus, the allegedly distinctive socioeconomic profile of District 4 is factually unconvincing, methodologically flawed, irrelevant, and unquestionably a hindsight rationalization of a plan that everyone understands to have been principally designed to create two majority-black districts. More simply, the socioeconomic profile of District 4 is an effect of District 4’s design, not a cause.

We see, then, that the Defendants’ proffer of protection of incumbent politicians and distinctive socioeconomic profiles as alternative explanations for the peculiar contours of the Plan simply do not ring true. Faced with competing inferences, this court — as fact finder — must determine the credibility of witnesses, weigh the evidence, and choose between those inferences.⁴⁹ Concluding that the facts and inferences overwhelmingly favor the Plaintiffs, we find that the Plan is undeniably a child of racial gerrymandering.⁵⁰ Defendants must therefore demon-state that Act 42 is narrowly tailored to satisfy a compelling state interest, as required by Shaw and other applicable Equal Protection Clause cases.⁵¹

C. Direct Proof of Racial Gerrymandering

We need not even consider the kind of indirect or inferential proof approbated in Shaw to reach the same point — a finding of racial gerrymandering. In this case, we also reach that junction when we consider the great weight of the direct evidence elicited at both the Trial and the Evidentiary Hearing.⁵² Regardless of whether we reach a finding of racial gerrymandering by the inferential approach elaborated in Shaw, or by direct testimony and documentary evidence, the Defendants have the burden of justifying that gerrymandering.

In this case, direct evidence clearly and forcefully demonstrates that the Plan is a product of racial gerrymandering. Virtually every witness who testified at the Trial (all without the benefit of a retrospective, self-serving view of Shaw) either affirmatively stated or accepted as gospel that the Plan was drawn with the specific intent of ensuring the creation of a second, safe, black majority congressional district: namely, District 4. The Defendants’ witnesses either stated or conceded that the districts created by Act 42 were racially gerrymandered. Indeed, those witnesses, both lay and expert, spent most of their time at the Trial discussing how large the percentage of registered black voters needed to be in the new majority black district to guarantee the efficacy of their racial gerrymander — an efficacy they viewed as the sine qua non of preclearance.

In response to this court’s query whether the Legislature had created a racial gerrymander, Defendants’ counsel — Mr. Mon-grue — said, “[a]nd [racial gerrymandering]^ exactly what [the Legislature] can do____” Similarly, Mr. Willie Hunter, a black state legislator and a fact witness for the Defendants, testified that the Legislature’s intention “was to create a district where there was an excellent possibility of having another black elected ..., and that to accomplish this they [the legislators] “looked at numbers [of black voters] period.” Professor Engstrom, an expert witness for the defense, stated unabashedly that “race drove ... [the creation of] ... the Districts.”

Not surprisingly, the Plaintiffs’ witnesses were at least equally convinced that considerations of race motivated the creation of Act 42. State Representative Adley, a white legislator from Northwest Louisiana, testified that “the only issue presented to us was a racial issue” and stated that District 4 was created “for the sole purpose of making sure that an additional black district got created regardless of what it looked like and what parishes it ran through ...” Dr. Gary Stokley agreed that “this plan is [based on] race.”

At the more recent Evidentiary Hearing, however, the Defendants attempted to recast their arguments in light of Shaw and to gainsay the racial gerrymandering that they so readily approbated during the pre-Shaw Trial. But even at this latter hearing the witnesses agreed that race was the overarching factor that drove the actual creation of the Plan, and that the Legislature had specifically intended to assort voters into districts based on race. Senator Marc Morial, a black legislator from New Orleans who appeared on behalf of the Defendants, testified that “[i]t was the intent of the Legislature to create ... [a] ... second majority black district.” This is perhaps the clearest, most direct post-Shaw statement of intent, and the best illustration of the difference between intent and motive. Similarly, United States Congressman James A. Hayes (D. La) stated that the politics of race “was the major element” that drove the creation of the Fourth District. Dr. Lawrence N. Powell, one of Defendants’ expert witnesses, agreed that “the primary determinant” of the shapes of Districts 2 and 4 was race.

During four full days of testimony, two in 1992 and two more in 1993, some witnesses stated that race was the only factor, while others said that race was the primary factor. One witness — apparently uncomfortable with saying race was the primary factor — admitted that race was a very important factor. Not one witness, Plaintiffs’ or Defendants’, testified that Act 42 was not largely a product of racial gerrymandering — not one.

Harking back to our homicide analogy, we note that although witnesses at the Evidentiary Hearing voiced various altruistic motives — or accused others of various ulterior motives — for intentionally employing racial gerrymandering to, create voting districts on the basis of race,⁵³ everyone agreed that the intent of the Legislature — analyzed as a whole rather than from the point of view of its constituent members or caucuses — was to create a redistrieting plan with a second majority black district.

The evidence showed that the Plan passed the Legislature by virtue of an uncommon alliance of legislators: Some who supported the Plan wanted a second super-majority black district to increase the number of black representatives in Louisiana’s congressional delegation. Other supporters of the Plan perceived various benefits in the correlative whitening of some districts that attends the intentional segregation of black voters into other adjacent districts.⁵⁴ But there is absolutely no doubt that the immediate intent of the Legislature as a whole was to enact a plan containing two black majority districts, essentially without regard to any other considerations and interests.

Viewed in any light, the direct evidence in this ease proves the presence of racial gerrymandering. Thus, even in the event that we may have somehow misconstrued the Court’s opinion in Shaw, or misapplied the inferential minuet established therein, the direct evidence that Act 42 was a product of racial gerrymandering is overwhelming. Two independent evidentiary bases (inferential and direct) thus support our finding that Act 42 reflects racial gerrymandering. Each is sufficient on its own to subject Act 42 to strict scrutiny, and each is consequently sufficient to require the, Defendants to demonstrate that Act 42 is “narrowly tailored,to further a compelling governmental interest,” as required by Shaw and other applicable Equal Protection Clause precedents.⁵⁵

D. Strict Scrutiny of the Plan

Again, the core principle underlying the Supreme Court’s decision in Shaw is that racially gerrymandered redistricting plans are subject to the same strict scrutiny that applies to other state legislation classifying citizens on the basis of race.⁵⁶ To survive such scrutiny, racially gerrymandered redistricting plans must be narrowly tailored to further a compelling governmental interest.⁵⁷

1. Compelling Governmental Interest

Defendants advance four possible compelling state interests to justify their racial gerrymandering: , (1) conformity with Section 2 of the Voting Rights Act, (2) conformity with Section 5 of the Voting Rights Act, (3) proportional representation of Louisiana blacks in Congress,⁵⁸ and (4) remedying the effects of past racial discrimination. Witnesses also made oblique references to various other admirable but nebulous — and often question-begging — motives, like promoting racial harmony and ensuring fairness. As we conclude, however, that Act 42 is not narrowly tailored to further these or any other compelling state interests, we need not decide here whether any one or more of them — properly clarified — is such an interest. For the sake of judicial economy, then, we do not analyze them in detail. Rather, we assume — without granting — that one or all of them constitutes a compelling state interest.

2. Narrowly Tailored

By thus assuming — again, without granting — that the Defendants have articulated one or more compelling state interests that the Plan might further, we have shifted the focus of our strict scrutiny to the final aspect of Shaw; whether the Plan is narrowly tailored. In our final illustrative comparison with homicide, we note the parallelism between the hypothetical criminal defendant’s burden — at common law — of establishing an affirmative defense — such as justifiable homicide — and the State’s burden here of establishing the affirmative justification of a compelling state interest.⁵⁹ But even such affirmative defenses contain crucial limiting elements: the requirement of applying only reasonable force in the self-defense context, and the requirement of narrow tailoring in the compelling state interest context.

A homicide defendant, relying for acquittal on self-defense or justifiable homicide, must show not only that he had a reasonable fear of imminent harm from the aggressor/victim, but also that he used no more force than was reasonably necessary under the circumstances. If a single, deterrent gunshot to the aggressor/victim’s leg happens to cause death from uncontrollable hemorrhaging, the homicide may well be justified; but if the defendant fires two or three or four additional, immediately-fatal shots into vital areas of that same aggressor/victim’s body after he has been neutralized by the first shot in the leg, the defendant is likely to find that the affirmative defense is unavailing.

In close parallel, Shaw tells us that, even in retrogression cases under Section 5, the State does not have carte blanche to engage in racial gerrymandering; its reapportionment plan must not go “beyond what [is] reasonably necessary” to further the compelling governmental interest.⁶⁰ That is the essence of narrow tailoring in the redistricting context: just as a homicide defendant may not use excessive force to stop an aggressor, neither may a state burden the rights and interests of its citizens more than is reasonably necessary to further the compelling governmental interest advanced by the :state.

In this case, uneontroverted evidence from both the Trial and the Evidentiary Hearing convinces us that the Plan is not narrowly tailored to satisfy any of the supposedly compelling state interests advanced by the Defendants.⁶¹ We reach that conclusion first because the Plan entails considerably more segregation than is necessary to satisfy the need for a second black majority district— even assuming arguendo that such a second district were itself justified — and second because the Plan excessively burdens a variety of third party interests — dramatically so.

In its Shaw opinion, the Supreme Court provides one example to illustrate how to apply the requirement that a racially gerrymandered plan be narrowly tailored to satisfy one or more compelling state interests: “A reapportionment plan would not be narrowly tailored to avoiding the goal of retrogression if the state went beyond what was reasonably necessary to avoid retrogression.”⁶² Thus, if providing a single majority-minority district satisfies the nonretrogression requirement imposed by Section 5 of the Voting Rights Act,⁶³ then a racially gerrymandered redistricting plan comprising more than one minority-majority district is not narrowly tailored to satisfy the compelling state interest of comporting with Section 5.⁶⁴ Stated more broadly, as voters have an equal protection right not to be segregated by their state legislatures or local governments into various voting districts on the basis of race, only a plan that segregates to no greater extent than is reasonably necessary to further a compelling governmental interest can survive constitutional scrutiny. The same can be said for a plan that supersaturates a majority-minority district, while concomitantly depleting adjacent majority-majority districts of minority voters.

In this case, we find that the Plan entails more constitutionally suspect segregation than necessary, and is therefore not narrowly tailored. Continuing to assume arguendo that some state interest had been identified which could justify the creation of a second black-majority district, this Plan-would have to be rejected as insufficiently narrowly tailored. It packs more black voters into a District 4 than are reasonably necessary to give blacks a realistic chance to determine the outcome of elections there, providing that they exercise their right to vote. Also, the boundaries of the district violate traditional districting principles to a substantially greater extent than is reasonably necessary.

District 4 contains 63% registered black voters, significantly more than are needed to elect representatives of their choice. To greater or lesser degrees, all expert witnesses acknowledged — some only reluctantly under cross examination — that 63% black voting age population was well in excess of the percentage needed for reasonable comfort in creating a safe voting-age majority-minority district. For further confirmation that this is so we need only ask rhetorically “If 54% is sufficient in District 2, why must District 4 be supersaturated with 63% black voting age citizens?” No evidence was adduced by the Defendants to demonstrate a substantial difference between the voting patterns in the areas covered respectively by Districts 2 and 4 — either in white crossover or black block voting — to rationalize the 54%-63% disparity-in black voting age populations.

Although the witnesses at trial disagreed on the exact percentage of net white crossover votes for black candidates, we find that the evidence supported an average, net white cross-over vote in non-judicial elections of between 10% and 25%.⁶⁵ The evidence at trial also indicated that minority voter registration is now comparable to white registration. Clearly, District 4 need not contain a black voting age population of 63% to satisfy the interest of a second district in which black officials can be readily elected. And, although it is not the province of this court in this case to establish the demographic specifications of the State’s congressional redistricting plan, we find on the basis of the credible testimony and documentary evidence, that in this instance a district with a black voting age population of not more than 55% — and probably less — would have been adequate to ensure that blacks could elect a candidate of their choice, assuming they chose to exercise their franchise and assuming the candidate of their choice had more than a modicum of appeal for non-black voters.

Additional confirmation that the Plan is not narrowly tailored lies in its excessive disregard of traditional redistrieting criteria and its derogation from third-party interests. Supreme Court precedent indicates that a variety of factors, both relative and absolute, are germane to analyzing whether a government measure is narrowly tailored. These factors include (1) the necessity of the measure, (2) the efficacy of alternative race-neutral measures, (3) the availability of more narrowly tailored (less intrusive) measures, (4) the flexibility and duration of the measure, and (5) the impact of the measure on the rights of third parties.⁶⁶ Relevant Supreme Court jurisprudence thus suggests that — in essence — a plan is not narrowly tailored if it adversely affects more interests, if it generally wreaks more havoc, than it reasonably must to accomplish the articulated compelling state interest. We find that the Plan thus offends.

Whether under a relative or a comparative analysis, the evidence adduced at Trial and at the Evidentiary Hearing undeniably established that, even in the face of the black population’s wide dispersion in Louisiana (outside New Orleans), a second black majority district could have been drawn that would have done substantially less violence to traditional redistrieting principles. For example, Marc Morial stated that “[t]here were alternatives which would have created a more compact black district that were not enacted____” Similarly, Congressman Hayes admitted that a more compact plan could have been enacted. Freshman United States Congressman Cleo Fields — the former state senator and incumbent congressman from new District 4, who testified for the Defendants— also acknowledged that “it would have been possible to fulfill the desire of a second [black] majority district ... and pay more attention than this District [District 4] does to both compactness and contiguity.”

Dr. Alan Lichtman — one of the Defendants’ expert witnesses at the Evidentiary Hearing — likewise testified that, were he to sit down with Dr. Ronald Weber (the expert on the other side) they “could create a [second black majority] district that looked better than this one [district 4].” And Dr. Weber — the Plaintiffs’ expert witness at both the trial and the evidentiary hearing — testified that a plan could be devised that would include a second black majority district, yet still be significantly more compact and comport much more closely with other traditional redistrieting principles.

The testimony of freshman state Senator Tom Greene at the Evidentiary Hearing was perhaps the most enlightening on this point. He stated that, with the neutral assistance of the Legislature’s redistrieting computer technician, he (Greene) actually created a plan that evinced greater respect for Louisiana’s traditional parish boundaries than the Plan, and still included a second majority-black district. Finally, Mr. Glenn Koepp, who is the Assistant Secretary of the Senate and the technician in charge of reapportionment activity in Louisiana since 1981, testified that computer-supported mathematical modeling programs gave the State the tools to create several plans — including several with a second black district — that would respect traditional redistrieting criteria to a much greater degree than does the Plan. Thus, the great weight of the credible evidence indicates that the Legislature could have developed and adopted a redistrieting plan — even one with a second majority black district — that reflected greater respect for traditional redistrieting criteria and that was less disruptive to the traditional political, social, economic, ethnic, geographical, and religious organization of the State. In short, substantially less extreme racially gerrymandered plans are readily available; plans without such obviously overbroad, overarching overkill.

In summary, we hold that the Plan is not narrowly tailored, either relatively or absolutely. This is so because it embraces considerably more racial gerrymandering — and thus more segregation — than is needed to satisfy any advanced state interest, and because the Plan unnecessarily violates a host of historically important redistrieting principles, thereby adversely affecting countless third party interests. These several and varied interests — some constitutionally protected and others merely important — may not be callously sacrificed on the altar of political expediency, particularly when less broadly tailored plans are conceivable.⁶⁷

. IV

CONCLUSION

We find that the Plan in general and Louisiana’s Congressional District 4 in particular are the products of racial gerrymandering and are not narrowly tailored to further any compelling governmental interest. - We are therefore constrained to conclude that the Plaintiffs’ right to equal protection as guaranteed by the United States Constitution is violated by the Plan. Consequently, we declare Act 42 of 1992 to be unconstitutional and the redistrieting plan embodied therein to be null and void; and we enjoin the State of Louisiana from holding any future congressional elections based on the Plan. We do not, however, invalidate the 1992 congressional elections held thereunder; but we do hold that the term of office of each member of the United States House of Representatives from Louisiana who represents a district created under the Plan — and each district thus created — shall expire, ipso facto, at noon on the 3rd day of January, 1995, such terms of office and such districts not to be extended or carried over into the next Congress in any manner whatsoever.

APPENDIX A

The Seven Districts Created by Act 42

[[Image here]]

APPENDIX B

1 . A technical distinction can be drawn between the terms "apportionment” and "reapportionment,” on the one hand, and "districting” and "redistrieting" on the other:

apportionment and reapportionment involve the allocation [by Congress] of a finite number of representatives among a fixed number of preestablished areas. Districting and redistrieting ... refer to the process by which the lines separating legislative districts are drawn [by states].

See Major v. Treen, 574 F.Supp. 325, 328 (E.D.La.1983) (quoting Backstrom et al., Issues in Gerrymandering: An Exploratory Measure of Partisan Gerrymandering Applied to Minnesota, 62 Minn.L.Rev. 1121, 1121 n. 1 (1978). For simplicity, however, we use both terms interchangeably.

2 . Shaw v. Reno, 508 U.S. -, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993).

3 . See U.S. Const. amend. XX, § 1.

4 .See Table, Appendix A.

5 . Id. 508 U.S. at -, 113 S.Ct. at 2829, 125 L.Ed.2d at 531 (citing Richmond v. J.A. Croson Co., 488 U.S. 469, 494, 109 S.Ct. 706, 722, 102 L.Ed.2d 854 (1989) (internal quotations omitted).

6 . Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 290-91, 98 S.Ct 2733, 2748, 57 L.Ed.2d 750 (1978). Indeed, laws that entail overt racial classifications are subject to strict scrutiny, even when they burden or benefit different races equally. See, e.g., Powers v. Ohio, 499 U.S. 400, 111 S.Ct. 1364, 113 L.Ed.2d 411 (1991). "[R]acial classifications do not become legitimate on the assumption that all persons suffer them in equal degree.” Powers, 499 U.S. at 410, 111 S.Ct. at 1370, 113 L.Ed.2d at 425.

7 . George Orwell, Animal Farm 123 (Penguin Books 1972) (1946).

8 . Shaw, 508 U.S. at-, 113 S.Ct. at 2832, 125 L.Ed.2d at 536 (indicating that irregular shapes may imply impermissible racial gerrymandering). See also id. at---, 113 S.Ct. at 2826-27, 528-29.

9 . Shaw, 508 U.S. at-, 113 S.Ct. at 2832, 125 L.Ed.2d at 536.

10 . Rule 54(b) of the Federal Rules of Civil Procedure reads in pertinent part: "[A]ny order or other form of decision, however designated, which adjudicates fewer than all the claims ... shall not terminate the action as to any of the claims ... and the order ... is subject to revision at any time before the entry of judgment adjudicating all the claims ...” Fed.R.Civ.P. 54(b).

11 . Shaw v. Reno, 508 U.S. at -, 113 S.Ct. at 2832, 125 L.Ed.2d. at 536.

12 . See, e.g., Wright v. Rockefeller, 376 U.S. 52, 66-67, 84 S.Ct. 603, 610-11, 11 L.Ed.2d 512 (1964) (in. which the Court examined whether the plaintiffs had sustained their burden of proving "that the New York legislature was either motivated by racial considerations or in fact drew districts on racial lines”). See also Shaw, 508 U.S. -, 113 S.Ct. 2816, 125 L.Ed.2d 511 (in which ‘‘racially gerrymandering” and "intentional segregation of voters into separate voting districts” are used interchangeably throughout).

13 . For example, if a legislature devises a redistricting plan that separates river districts from mountain districts, and more blacks happen to live along the rivers, and more whites happen to live in the mountains, then the plan is not a product of racial gerrymandering, even though it entails some coincidental segregation. In short, de facto or accidental segregation is not constitutionally suspect, but state-sponsored intentional segregation is, irrespective, of the legislature's underlying motives.

14 . Shaw v. Reno, 508 U.S.-, 113 S.Ct. 2816, 125 L.Ed.2d 511; accord City of Mobile v. Bolden, 446 U.S. 55, 67, 100 S.Ct. 1490, 1499, 64 L.Ed.2d 47 (1980) (indicating that claims of racial discrimination in the reapportionment context are resolved in the same way as other claims of racial discrimination); Washington v. Davis, 426 U.S. 229, 240, 96 S.Ct. 2040, 2047, 48 L.Ed.2d 597 (1976) (referring to Wright v. Rockefeller, 376 U.S. at 58, 84 S.Ct. at 606, and pointing out that every member of the Court believed that plaintiffs stated a claim by alleging that “boundaries were purposefully drawn on racial lines”). In its supplemental brief to this court, the United States (which entered this case as amicus curiae) agrees that racially gerrymandered redistricting plans must be strictly scrutinized.

15 . See Shaw, 508 U.S. at-, 113 S.Ct. at 2826, 125 L.Ed.2d at 528.

16 . Id. at --, 113 S.Ct. at 2826, 125 L.Ed.2d at 527. Thus, even if benign or benevolent motives underlie a legislature's decision to racially gerrymander a redistricting plan, that plan is still subject to strict judicial' scrutiny. As discussed below, good motives may allow a plan to survive strict scrutiny, if they rise to the level of a compelling state interest, and if the plan is narrowly tailored to further such an interest. But such motives — however unspotted — do not automatically exempt the plan from what amounts to a presumption of unconstitutionality.

17 . In Wright v. Rockefeller, Justice Douglas discusses the dangers of racial gerrymandering. Wright v. Rockefeller, 376 U.S. 52, 66-67, 84 S.Ct. 603, 610-11 (Douglas, J., dissenting) ("When racial ... lines are drawn by the State, the multiracial ... communities that our Constitution seeks to weld together as one become separatist; antagonisms that relate to race ... rather than to political issues are generated; communities seek not the best representative but the best racial ... partisan."). In Shaw, Justice O’Connor echoes these same concerns. Shaw, 508 U.S. at.-, 113 S.Ct. at 2827, 125 L.Ed.2d. at 529 (Redistricting plans that segregate citizens primarily on the basis of race bear "an uncomfortable resemblance to political apartheid.”).

18 . Shaw, 508 U.S. at-, 113 S.Ct. at 2832, 125 L.Ed.2d at 536; see id. at ----, 113 S.Ct. at 2826-27, 125 L.Ed.2d at 528-29 (indicating that irregular shapes may reflect impermissible racial gerrymandering).

19 . See id. at---, --, 113 S.Ct. at 2826-27, 2832, 125 L.Ed.2d at 528-29, 536 (regardless of how a racial gerrymander is created, it should receive strict scrutiny); accord City of Mobile v. Bolden, 446 U.S. 55, 67, 100 S.Ct. 1490, 1499, 64 L.Ed.2d 47 (1980).

20 . 508 U.S. at -, 113 S.Ct. at 2824, 125 L.Ed.2d at 525 (citing Personnel Administrator of Massachusetts v. Feeney, 442 U.S. 256, 272, 99 S.Ct. 2282, 2292, 60 L.Ed.2d 870 (1979)).

21 . Even-though the Legislature put itself under severe time constraints by failing to redistrict in 1991, the United States Department of Justice— not a party to this litigation — must be regarded as an active player in this case. As Louisiana is covered by Section 5 of the Voting Rights Act, the Legislature must either (1) have any proposed plan precleared by the Department of Justice, or (2) seek a judgment from the United States District Court for the District of Columbia declaring that the plan "does not have the purpose and will not have the effect of denying or abridging the right to vote on account of race or color____” Voting Rights Act of 1965, 42 U.S.C. § 1973c. The testimony at the Trial and at the Evidentiary Hearing reflected overwhelmingly that the Attorney General’s Office (AGO) had let it be known that preclearance would not be forthcoming for any plan that did not include at least two "safe” black districts out of seven. But neither Section 2, nor Section 5 of the Voting Rights Act justifies the AGO’s insistence upon two black districts.

To challenge a redistricting plan under Section 2 successfully, plaintiffs who are members of a cognizable racial group must demonstrate that their group is numerous enough and geography catty compact enough to be a majority in a district. Growe v. Emison, 507 U.S.-,-, 113 S.Ct. 1075, 1083, 122 L.Ed.2d 388, 403 (1993) (extending the threshold requirements elaborated in Thornburg v. Gingles, 478 U.S. 30, 46-47, 106 S.Ct. 2752, 2764, 92 L.Ed.2d 25 (1986), to single member, as well as multimember, districts). Ob.viously, however, black voters in District 4 do not constitute a geographically compact body as required by Gingles. Neither would any district resembling District 4 be geographically compact. Consequently, the AGO cannot legitimately claim that Section 2 of the Voting Rights compels the creation of any redistricting plan that contains a tortured district like District 4. Unless a second geographically compact black majority district is possible, the State’s failure to create one does not violate Section 2. Growe, 507 U.S. at --- -, 113 S.Ct. at 1084-85, 122 L.Ed.2d at 404. This is not to say that Section 2 forbids the creation of such a plan; simply that Section 2 does not require it. Voinovich v. Quilter, 507 U.S. -,---, 113 S.Ct. 1149, 1156-58, 122 L.Ed.2d 500, 513-14 (1993). Consequently, the AGO cannot rely on Section 2 of the Voting Rights Act to force a state to adopt a plan containing an additional majority-minority district with a geographically dispersed black majority.

Additionally, Section 2 of the Voting Rights Act expressly states that "[njothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.” Voting Rights Act of 1965, Section 2, 42 U.S.C. § 1973. Section 2 refuses to transform the individual right to vote into a group right to elect. Thus, Section 2 emphatically does not authorize the AGO to reject plans that fail to "give full effect” to minority voters, or otherwise to create a de facto requirement of proportional representation by rejecting all plans that fail to maximize the concentration of minority voters. The text of Section 2 of the Voting Rights Act expressly declares that proportional representation is not required.

Neither can Section 5 of the Voting Rights Act- — on the basis of the 1990 Census — -justify the AGO’s insistence that Louisiana adopt a congressional redistricting plan with two safe, black-majority districts. In Beer v. United States, the Supreme Court stated that "the purpose of Section 5 has always been to insure that no voting-procedure changes would be made that would lead to a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise.” Beer v. United States, 425 U.S. 130, 141, 96 S.Ct. 1357, 1364, 47 L.Ed.2d 629 (1976) (emphasis added); accord Shaw, 508 U.S. at-, 113 S.Ct. at 2830, 125 L.Ed.2d at 533. In this case, however, a redistricting Plan for the State of Louisiana that provided for one black majority district would have satisfied the requirement of nonretrogression.

As a result of the 1990 Census, Louisiana lost a seat in the United States House of Representatives: previously it had eight seats, -now it has seven. Thus, even if the Legislature had adopted a plan with a single black majority district, that plan would have satisfied the Section 5 nonretrogression principle: whereas before black majority districts comprised one-eighth of all districts, under a new plan with one black majority district, such districts would constitute one-seventh of all districts, an increase of nearly two percent. Thus, the Section 5 nonretrogression principle does not, on the instant facts, require Louisiana to adopt a redistricting plan with a second black-majority district. Any suggestion by the AGO to the contrary was incorrect.

In summary, neither Section 2 nor Section 5 of the Voting Rights Act justify the AGO's insistence that Louisiana adopt a plan with two safe, black majority districts. Yet members of the Legislature uniformly believed that they needed to create such a redistricting plan to secure preclearance. The letters sent by the Office of the Assistant Attorney General (AAGO) in response to redistricting plans for the Louisiana Senate and the Board of Elementary and Secondary Education (BESE) show how Louisiana legislators got this mistaken impression.

In these letters, the AAGO acknowledged that the plans "appeared] to have no retrogressive effect” and met "in large part ... Section 5 preclearance requirements.” Yet the AAGO refused to preclear the plans because — in its apparent judgment — they could have been drawn “in a manner that would more effectively provide to black voters an equal opportunity ... to elect candidates of their choice," there a euphemism for black candidates. But, again, the Voting Rights Act does not require that a plan be drawn to maximize the efficacy of the black vote, and we perceive the AAGO’s insistence upon such a result to be tantamount to an insistence upon proportional representation: an insistence that is expressly forbidden by Section 2 of the Voting Rights Act.

In one letter the AAGO went so far as to suggest how the plan should be drawn to secure early preclearance by combining "significant concentrations of black voters in northeastern Louisiana and in the parishes bordering the State of Mississippi, both along the river and the state's southern border.” Here — in a nutshell— was a blueprint for the highly irregular district known as District 4: a district that runs through the length and breadth of the State, a district that ignores traditional geographical and political boundaries, yet a district that the AAGO inappropriately hinted would have to be created if Louisiana was to secure preclearance. What was the authority for the AAGO's insistence? The answer presents itself: none.

When State authorities wrote back, explaining the historical, cultural, political, economic, and religious significance of the north-south divide in Louisiana, as a means of explaining and defending its proposed plan, the AAGO dismissed the explanation and concluded that lumping black voters together in a district "transcends the distinction between northern and southern parishes." What was the authority for this judgment: again, none! To reiterate, neither Section 2 nor Section 5 of the Voting Rights Act requires that geographically dispersed black voters be lumped together to maximize the efficacy of their vote. That is nothing more than an AAGO “gloss” on the Voting Rights Act — a gloss unapproved by Congress and unsanctioned by the courts.

The Department of Justice did not conclude that Louisiana’s concern with preserving its historic parishes was pretextual; it decided that creating black districts trumped traditional redistricting criteria. What was the authority for this policy decision? Yet again, none! As the AAGO concluded that the State BESE plan was not retrogressive with' respect to the black vote, it necessarily admitted that the plan had no discriminatory effect. See Beer v. United States, 425 U.S. 130, 140, 96 S.Ct. 1357, 1363, 47 L.Ed.2d 629, 639. Thus, the AAGO could legally reject the plan only if it determined — and presumably for non-arbitrary reasons — that the plan reflected discriminatory motives. Georgia v. United States, 411 U.S. 526, 93 S.Ct. 1702, 36 L.Ed.2d 472 (1973). But the AAGO did not so conclude.

Rather, the AAGO arrogated the power to use Section 5 preclearance as a sword to implement forcibly its own redistricting policies, rather than as a shield to prevent lamentable historical abuses. The Attorney General’s Office has no authority to withhold preclearance for reasons outside the ambit of the Voting Rights Act. Indeed, because any plan that entails more racial gerrymandering than is absolutely necessary to pass Voting Rights Act muster is potentially unconstitutional, Shaw, 508 U.S. at -, 113 S.Ct. at 2831, 125 L.Ed.2d at 534, the AGO will risk encouraging unconstitutional conduct if it pressures states to gerrymander their districts to maximize the concentration of minority voters.

22 . McDonnell Douglas v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973); Texas Dep’t of Community Affairs v. Burdine, 450 U.S. 248, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981). The three-step McDonnell-Douglas-Burdine "minuet” structures the basic allocation of burdens and order of presentation of proof in the Title VII context. Burdine, 450 U.S. at 252-53, 101 S.Ct. at 1093. This minuet requires (1) the plaintiff to prove — by a preponderance of the evidence — a prima facie case of disparate treatment that contravenes Title VII of the Civil Rights Act of 1964, (2) the defendant to proffer a legitimate, nondiscriminatoiy reason for the employee’s rejection, and (3) the plaintiff to demonstrate that the defendant's proffered explanation is pretextual. Id. at 253, 101 S.Ct. at 1093. At the end of the day, however, the plaintiff has the burden of proving that a violation of Title VII occurred. See St. Mary’s Honor Center v. Hicks, 509 U.S. --, 113 S.Ct. 2742, 125 L.Ed.2d 407 (1993).

23 . Shaw, 508 U.S. at-, 113 S.Ct. at 2832, 125 L.Ed.2d at 536; See also id. at---, 113 S.Ct. at 2826-27, 125 L.Ed.2d at 528-29 (indicating that irregular shapes may imply impermissible racial gerrymandering).

24 . Id

25 . See, e.g., Wright, 376 U.S. 52, 84 S.Ct. 603 (in which the Court held that the plaintiffs had not “sustained their burden of proving that ... [the plan] ... segregates eligible voters by race and place of origin in violation of the Equal Protection and Due Process Clauses____"); accord Whitcomb v. Chavis, 403 U.S. 124, 91 S.Ct. 1858, 29 L.Ed.2d 363 (1971). To clarify this minuet further, if a plaintiff comes into court with a map bearing hideously contorted districts and evidence that the state legislature, drew those districts on the basis of race, and if the plaintiff complains that those districts lack a non-racial explanation — i.e., cannot be explained or understood without hypothesizing racial gerrymandering — then the plaintiff has stated a prima facie case under Shaw. If the state then introduces evidence that tends to show that the legislature was actuated by other motives that can explain the bizarre contours of the districts without resorting to race, the state has created a competing inference. The factfinder must then decide, on the basis of all available evidence, who is right.

26 . Id.

27 . City of Mobile v. Bolden, 446 U.S. 55, 66, 100 S.Ct. 1490, 1499, 64 L.Ed.2d 47 (1980) (citing Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976)); Plaintiffs also have the burden of proving the invalidity of a reapportionment plan under Section 2 of the Voting Rights Act. Voinovich v. Quilter, 507 U.S. -, -, 113 S.Ct. 1149, 1156, 122 L.Ed.2d 500, 512 (1993).

28 . Wright, 376 U.S. at 55, 84 S.Ct. at 604.

29 . Shaw, 508 U.S. at-, 113 S.Ct. at 2826, 125 L.Ed.2d at 528.

30 . For example, Plaintiffs allege that "[t]he adoption of Act 42 ... has resulted in the creation of at least two of seven districts which are devoid of any commonality of interest ... geographical compactness, contiguousness [sic], consistency with existing political, societal, governmental or economic districts or jurisdictional boundaries, other than the racial designation of the majority therein." See discussion in section II B of this Memorandum Opinion for other examples of Plaintiffs’ allegations.

31 .Recalling our pedagogic comparison with homicide, when the state attempts to convict a defendant of homicide using circumstantial evidence — essentially arguing that the facts are consistent with homicide, the defendant may attempt to show that the evidence -is consistent with a far more innocent hypothesis: for example, an accident. The Defendants' opportunity, to' provide legitimate, non-racial explanations for the Plan's irrational shape is analogous.

32 . See Map, Appendix B.

33 . Id.

34 . See generally Shaw, 508 U.S. --, 113 S.Ct. 2816, 125 L.Ed.2d 511.

35 . Shaw, 508 U.S. at---, 113 S.Ct. at 2826-27, 125 L.Ed.2d at 528-29 (citing Gaffney v. Cummings, 412 U.S. 735, 752, n. 18, 93 S.Ct. 2321, 2331, n. 18, 37 L.Ed.2d 298 (1973)). The seminal case of Major v. Treen also discusses these traditional redistricting criteria and applies them to the Louisiana context. Major v. Treen, 574 F.Supp. 325 (E.D.La.1983).

36 . Id. 508 U.S. at --, 113 S.Ct. at 2827, 125 L.Ed.2d at 529 (citing Karcher v. Daggett, 462 U.S. 725, 755, 103 S.Ct. 2653, 2672, 77 L.Ed.2d 133 (1983)).

37 . Id. at --, 113 S.Ct. at 2826, 125 L.Ed.2d at 528.

38 . Id.

39 . See Map, Appendix B.

40 . In complementary fashion, District 6 maintains contiguity only by employing gossamer connections at several junctions.

41 . A parish is analogous to a county, but it has its roots in, and takes its name from, the ministrations of the Catholic Church in early Louisiana history.

42 . Disregarding New Orleans, which comprises the majority of District 2, the major municipalities of Shreveport, Baton Rouge, Lafayette, Monroe, Alexandria, and Ruston are all rent asunder in this manner.

43 . At the Evidentiary Hearing, Defendants introduced evidence of the role of incumbency, but it proved irrelevant to the issue of intentional segregation of voters by race. Such evidence showed, at most, that incumbency affected only the general location of the gerrymander that is District 4, i.e., the effective efforts of Congressional incumbents in "pushing” the pre-ordained second black majority district generally as far north and east as possible. Thus it follows largely the northern boundary of Louisiana with Arkansas and the eastern boundary of Louisiana, i.e., the Mississippi River. But, as Louisiana had already lost one of its eight seats in the House of Representatives, creating a game of incumbents’ "musical chairs,” none should be surprised by the interesting but irrelevant fact that, once the decision to create a second minority-majority district had been made, each incumbent did what he could to keep his traditional geographical base and see that the new district was located anywhere but in his.

44 . Shaw, 508 U.S. at-113 S.Ct. at 2832, 125 L.Ed.2d at 536 (emphasis added).

45 . Wright v. Rockefeller, 376 U.S. 52, 57, 84 S.Ct. 603, 605, 11 L.Ed.2d 512 (1964).

46 . At the risk of flaying a dead horse, we point out that logic demands that the Defendants' reading of Shaw be rejected. Suppose a state legislature came right out and admitted that it -racially gerrymandered districts in a reapportionment plan, thereby obviating the plaintiff's need to show racial gerrymandering inferentially. Could the state then say, "but we also drew a few district lines to protect incumbents, or to follow a river, or to put a state wildlife park in the district of a congressman who likes migratory birds,” and thereby defeat plaintiff's claim because race was not the only factor involved in the delineation of districts? Race will never be literally the only factor. So Shaw requires only that race be an important factor. If the plaintiff's evidence shows that the plan cannot be understood without postulating that the legislature was significantly actuated by racial motives, then the plaintiff has made out a case of racial gerrymandering.

The cases of Wright v. Rockefeller and Arlington Heights v. Metropolitan Hous. Dev. Corp. support this interpretation. In Wright every member of the court accepted that the plaintiffs. stated a claim by alleging that the "New York Legislature was either motivated by racial considerations, or in fact drew the districts on racial lines.” 376 U.S. at 54, 84 S.Ct. at 604. Similarly, the Court in Arlington Heights noted that invidious discriminatory intent need not be the legislature's dominant purpose: proof that such discriminatory intent was "a motivating factor” in the legislation’is sufficient. 429 U.S. 252, 265, 97 S.Ct. 555, 563, 50 L.Ed.2d 450 (1977).

47 . The Defendants also offered this evidence to argue (1) that Act 42 served a compelling state interest by enhancing the representation for poor people — especially poor black people — and (2) that the Plan adhered to the traditional redistricting principle of "commonality of interests” and thus was narrowly tailored.

48 . An additional criticism of Mr. Lichtman's statistical legerdemain is that the socioeconomic characteristics he analyzes are themselves strongly correlated with race, a classic chicken- or-egg fallacy. At this moment in history black people in the South (and generally in America) are — on the average — poorer and less well-educated than their white counterparts. Moreover, blacks in largely segregated communities are probably poorer — on the average — than blacks in more integrated communities. Consequently, racially gerrymandered plans, which seek to draw boundaries around various concentrations of black persons, will inevitably tend to concentrate the poorer, less well-educated blacks.

The Defendants’ conclusions therefore have a tautological quality: to prove that factors other than race can explain District 4, the Defendants analyze socioeconomic factors that correlate strongly with race. Of course District 4 is relatively poor and relatively uneducated: it was intentionally filled with relatively poor and uneducated minorities. To use statistical parlance, Dr. Lichtman's conclusions are spurious: they tell us nothing that we could not have predicted based on the Legislature's decision to pack District 4 with black voters. For a more scientific discussion of such spurious correlations-, see Hubert M. Blalock, Jr., Social Statistics 443-48 (McGraw-Hill 1972).

49 . Wright v. Rockefeller, 376 U.S. 52, 57, 84 S.Ct. 603, 605, 11 L.Ed.2d 512 (1964).

50 . In finding that racial gerrymandering exists in this case, we feel confident that we are faithfully following the Supreme Court’s temper in Shaw. In that case, the Court gives two examples of fact patterns in which proving racial gerrymandering "will not be difficult at all." Shaw, 508 U.S. at -, 113 S.Ct. at 2826, 125 L.Ed.2d at 528. One example is when "a State concentrate[s] a dispersed minority population in a single district by disregarding traditional districting principles such as compactness, contiguity, and respect for subdivisions.” Id. Those facts squarely conform to this case. In the instant case, the Louisiana State Legislature created a serpentine belt 600 miles long to engulf enough blacks (actually, more than enough) to create a single super-majority district. According to the Court, this case presents a textbook example of racial gerrymandering that can be easily proved by application of the Shaw inferential minuet.

51 . Shaw, 508 U.S. at-, 113 S.Ct. at 2826, 125 L.Ed.2d at 528; see also supra note 13.

52 . See Shaw, 508 U.S. at-, 113 S.Ct. at 2826, 125 L.Ed.2d at 528 (implying that however a racial gerrymander is established, it should receive strict scrutiny). See also supra note 17 and accompanying text.

53 . Among the positive goals thus voiced were promoting racial harmony, increasing racial fairness, eradicating vestiges of past de jure segregation, and "obeying the law" in the sense of complying with the Voting Rights Act in a way that would justify preclearance, to name a few.

54 . Testimony at the trial revealed that Act 42 was passed by a legislative alliance between the Black and the Republican Caucuses, historically uncommon bedfellows but, according to expert testimony, a phenomenon occurring with increasing frequency across the country.

55 . Shaw, 508 U.S. at -, 113 S.Ct. at 2826, 125 L.Ed.2d at 528.

56 . See supra note 13.

57 . Shaw, 508 U.S.-at-, 113 S.Ct. at 2826, 125 L.Ed.2d at 528.

58 . At the Trial, Legislator Willie Hunter strongly advocated a second black majority district for the expressed goal of achieving black proportionality in congress. Without commenting — one way or another — on proportional representation’s viability as a "compelling governmental interest,” we foresee serious constitutional problems in accomplishing that goal through creation of a second black majority district. Evidence adduced here demonstrates that, of Louisiana black voting age population outside New Orleans, even the overloaded 63% black District 4 would produce the election of the [black] candidate of choice of fewer than 40% of those blacks whom the proponents of proportional representation want to be thus represented in Congress. Consequently, the Plan would constitute state action that, in effect, grants voting “proxies” from the black voters residing in Districts 1, 3, 5, -6 and 7 to the black voters of District 4 — clearly a disenfranchisement implicating equal protection and possibly due process as well.

59 . We realize that many states choose to make the absence of justification an element of the actus reus, thereby imposing upon the prosecution the burden of proving that the criminal defendant did not act in self-defense. As the Supreme Court has lately declined to review whether the Due Process Clause requires the burden of persuasion in self-defense cases to be placed on the prosecution, see, e.g., Moran v. Ohio, 469 U.S. 948, 105 S.Ct. 350, 83 L.Ed.2d 285 (1984) (Brennan, J., dissenting), state legislatures remain free to allocate burdens of proof "by labeling as affirmative defenses” aspects of a case that could alternatively be considered elements of a crime. Patterson v. New York, 432 U.S. 197, 210, 97 S.Ct. 2319, 2327, 53 L.Ed.2d 281 (1977); see also McElroy v. Holloway, 451 U.S. 1028, 1029-31, 101 S.Ct. 3019, 3019-21, 69 L.Ed.2d 398, 399-400 (Rehnquist, L, dissenting). For our purposes, however, such subtleties are irrelevant; as noted above, we are merely using classic common law homicide/self-defense as an instructive analogy.

60 . Shaw, 508 U.S. at -, 113 S.Ct. at 2831, 125 L.Ed.2d at 534.

61 . As noted above, in this section we assume— arguendo — that the State has demonstrated a compelling state interest, although, in fact, we have not so found. Neither could we on the evidence here presented. That is a judgment for the Legislature to make and the courts to review.

62 . Shaw, 508 U.S. at-, 113 S.Ct. at 2831, 125 L.Ed.2d at 534.

63 . Beer v. United States, 425 U.S. 130, 141, 96 S.Ct. 1357, 1364 (interpreting Section 5 as prohibiting voting-procedure changes that “lead to a retrogression in the position of racial minorities____").

64 . The Court’s example clearly indicates that the State of Louisiana may not justify the Plan, or any other racially gerrymandered plan with more than one black majority district, by reference to the need to comply with the Section 5 nonretrogression principle. As noted above, because Louisiana lost a seat in the United States House of Representatives, a plan with one black majority district would satisfy the Section 5 nonretrogression requirement given no more dramatic shift in the statewide ratio of white-to-black voters than is demonstrated by a comparison of the 1980 and 1990 census figures. See supra note 21.

65 . In other words, on the average — for non-judicial elections in Louisiana — black candidates will gain more votes from white cross-over voters (whites voting for black candidates) than they will lose from black cross-over voters (blacks voting for white candidates): thus, there is a positive net white cross-over vote. To put it another way, black voters are — again on the average — more racially coherentytonscious than white voters.

66 . See, e.g., Richmond v. J.A. Croson Co., 488 U.S. 469, 510-511, 109 S.Ct. 706, 730-31, 102 L.Ed.2d 854 (1989); United States v. Paradise, 480 U.S. 149, 171, 107 S.Ct. 1053, 1066, 94 L.Ed.2d 203 (1987); Fullilove v. Klutznick, 448 U.S. 448, 510-15, 100 S.Ct. 2758, 2791-93, 65 L.Ed.2d 902 (1980).

67 . We offer no opinion as to how much more narrowly tailored a plan would have to be to survive strict scrutiny. Indeed, we cannot say whether it is even possible — based on the 1990 Census — to devise a plan that would have two majority-minority districts and still be narrowly tailored vel non.

Even a plan that under the circumstances is as narrowly tailored as possible — in a relative sense (i.e., the least invasive and damaging plan available to further a compelling state interest) — may fail to be sufficiently narrowly tailored — on an absolute basis — to pass constitutional muster. For example, when the minority population is spread so evenly throughout a state that a majority-minority district cannot be drawn without dramatically impairing the constitutional rights of the citizens of that state, there may simply be no constitutionally permissible way to draw such a district, no matter how defensible the legislature’s motives for wishing to do so, or how bona fide its efforts to tailor the plan narrowly.